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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1098 - (show annotations)
Tue Oct 16 15:56:18 2012 UTC (7 years ago) by chpe
File size: 121962 byte(s)
Error occurred while calculating annotation data.
pcre32: utf: Reject all non-characters and not just 0xfffe
1 .TH PCREAPI 3 "07 September 2012" "PCRE 8.32"
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 pcre_jit_stack *pcre_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
99 .PP
100 .B void pcre_jit_stack_free(pcre_jit_stack *\fIstack\fP);
101 .PP
102 .B void pcre_assign_jit_stack(pcre_extra *\fIextra\fP,
103 .ti +5n
104 .B pcre_jit_callback \fIcallback\fP, void *\fIdata\fP);
105 .PP
106 .B const unsigned char *pcre_maketables(void);
107 .PP
108 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
109 .ti +5n
110 .B int \fIwhat\fP, void *\fIwhere\fP);
111 .PP
112 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
113 .PP
114 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
115 .PP
116 .B const char *pcre_version(void);
117 .PP
118 .B int pcre_pattern_to_host_byte_order(pcre *\fIcode\fP,
119 .ti +5n
120 .B pcre_extra *\fIextra\fP, const unsigned char *\fItables\fP);
121 .
122 .
123 .SH "PCRE NATIVE API INDIRECTED FUNCTIONS"
124 .rs
125 .sp
126 .B void *(*pcre_malloc)(size_t);
127 .PP
128 .B void (*pcre_free)(void *);
129 .PP
130 .B void *(*pcre_stack_malloc)(size_t);
131 .PP
132 .B void (*pcre_stack_free)(void *);
133 .PP
134 .B int (*pcre_callout)(pcre_callout_block *);
135 .
136 .
137 .SH "PCRE 8-BIT, 16-BIT AND 32-BIT LIBRARIES"
138 .rs
139 .sp
140 From release 8.30, PCRE can be compiled as a library for handling 16-bit
141 character strings as well as, or instead of, the original library that handles
142 8-bit character strings. From release 8.FIXME, PCRE can also be compiled as a
143 library for handling 32-bit character strings. To avoid too much complication,
144 this document describes the 8-bit versions of the functions, with only
145 occasional references to the 16-bit and 32-bit libraries.
146 .P
147 The 16-bit functions operate in the same way as their 8-bit counterparts; they
148 just use different data types for their arguments and results, and their names
149 start with \fBpcre16_\fP instead of \fBpcre_\fP. For every option that has UTF8
150 in its name (for example, PCRE_UTF8), there is a corresponding 16-bit name with
151 UTF8 replaced by UTF16. This facility is in fact just cosmetic; the 16-bit
152 option names define the same bit values.
153 .P
154 The 32-bit functions operate in the same way as their 8-bit counterparts; they
155 just use different data types for their arguments and results, and their names
156 start with \fBpcre32_\fP instead of \fBpcre_\fP. For every option that has UTF8
157 in its name (for example, PCRE_UTF8), there is a corresponding 32-bit name with
158 UTF8 replaced by UTF32. This facility is in fact just cosmetic; the 32-bit
159 option names define the same bit values.
160 .P
161 References to bytes and UTF-8 in this document should be read as references to
162 16-bit data quantities and UTF-16 when using the 16-bit library, unless
163 specified otherwise. More details of the specific differences for the 16-bit
164 library are given in the
165 .\" HREF
166 \fBpcre16\fP
167 .\"
168 page.
169 .
170 .P
171 References to bytes and UTF-8 in this document should be read as references to
172 32-bit data quantities and UTF-32 when using the 32-bit library, unless
173 specified otherwise. More details of the specific differences for the 32-bit
174 library are given in the
175 .\" HREF
176 \fBpcre32\fP
177 .\"
178 page.
179 .
180 .
181 .SH "PCRE API OVERVIEW"
182 .rs
183 .sp
184 PCRE has its own native API, which is described in this document. There are
185 also some wrapper functions (for the 8-bit library only) that correspond to the
186 POSIX regular expression API, but they do not give access to all the
187 functionality. They are described in the
188 .\" HREF
189 \fBpcreposix\fP
190 .\"
191 documentation. Both of these APIs define a set of C function calls. A C++
192 wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
193 documented in the
194 .\" HREF
195 \fBpcrecpp\fP
196 .\"
197 page.
198 .P
199 The native API C function prototypes are defined in the header file
200 \fBpcre.h\fP, and on Unix-like systems the (8-bit) library itself is called
201 \fBlibpcre\fP. It can normally be accessed by adding \fB-lpcre\fP to the
202 command for linking an application that uses PCRE. The header file defines the
203 macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
204 for the library. Applications can use these to include support for different
205 releases of PCRE.
206 .P
207 In a Windows environment, if you want to statically link an application program
208 against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
209 including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
210 \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
211 \fB__declspec(dllimport)\fP, with unwanted results.
212 .P
213 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
214 and \fBpcre_exec()\fP are used for compiling and matching regular expressions
215 in a Perl-compatible manner. A sample program that demonstrates the simplest
216 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
217 source distribution. A listing of this program is given in the
218 .\" HREF
219 \fBpcredemo\fP
220 .\"
221 documentation, and the
222 .\" HREF
223 \fBpcresample\fP
224 .\"
225 documentation describes how to compile and run it.
226 .P
227 Just-in-time compiler support is an optional feature of PCRE that can be built
228 in appropriate hardware environments. It greatly speeds up the matching
229 performance of many patterns. Simple programs can easily request that it be
230 used if available, by setting an option that is ignored when it is not
231 relevant. More complicated programs might need to make use of the functions
232 \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
233 \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
234 These functions are discussed in the
235 .\" HREF
236 \fBpcrejit\fP
237 .\"
238 documentation.
239 .P
240 A second matching function, \fBpcre_dfa_exec()\fP, which is not
241 Perl-compatible, is also provided. This uses a different algorithm for the
242 matching. The alternative algorithm finds all possible matches (at a given
243 point in the subject), and scans the subject just once (unless there are
244 lookbehind assertions). However, this algorithm does not return captured
245 substrings. A description of the two matching algorithms and their advantages
246 and disadvantages is given in the
247 .\" HREF
248 \fBpcrematching\fP
249 .\"
250 documentation.
251 .P
252 In addition to the main compiling and matching functions, there are convenience
253 functions for extracting captured substrings from a subject string that is
254 matched by \fBpcre_exec()\fP. They are:
255 .sp
256 \fBpcre_copy_substring()\fP
257 \fBpcre_copy_named_substring()\fP
258 \fBpcre_get_substring()\fP
259 \fBpcre_get_named_substring()\fP
260 \fBpcre_get_substring_list()\fP
261 \fBpcre_get_stringnumber()\fP
262 \fBpcre_get_stringtable_entries()\fP
263 .sp
264 \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
265 provided, to free the memory used for extracted strings.
266 .P
267 The function \fBpcre_maketables()\fP is used to build a set of character tables
268 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
269 or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
270 specialist use. Most commonly, no special tables are passed, in which case
271 internal tables that are generated when PCRE is built are used.
272 .P
273 The function \fBpcre_fullinfo()\fP is used to find out information about a
274 compiled pattern. The function \fBpcre_version()\fP returns a pointer to a
275 string containing the version of PCRE and its date of release.
276 .P
277 The function \fBpcre_refcount()\fP maintains a reference count in a data block
278 containing a compiled pattern. This is provided for the benefit of
279 object-oriented applications.
280 .P
281 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
282 the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
283 respectively. PCRE calls the memory management functions via these variables,
284 so a calling program can replace them if it wishes to intercept the calls. This
285 should be done before calling any PCRE functions.
286 .P
287 The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
288 indirections to memory management functions. These special functions are used
289 only when PCRE is compiled to use the heap for remembering data, instead of
290 recursive function calls, when running the \fBpcre_exec()\fP function. See the
291 .\" HREF
292 \fBpcrebuild\fP
293 .\"
294 documentation for details of how to do this. It is a non-standard way of
295 building PCRE, for use in environments that have limited stacks. Because of the
296 greater use of memory management, it runs more slowly. Separate functions are
297 provided so that special-purpose external code can be used for this case. When
298 used, these functions are always called in a stack-like manner (last obtained,
299 first freed), and always for memory blocks of the same size. There is a
300 discussion about PCRE's stack usage in the
301 .\" HREF
302 \fBpcrestack\fP
303 .\"
304 documentation.
305 .P
306 The global variable \fBpcre_callout\fP initially contains NULL. It can be set
307 by the caller to a "callout" function, which PCRE will then call at specified
308 points during a matching operation. Details are given in the
309 .\" HREF
310 \fBpcrecallout\fP
311 .\"
312 documentation.
313 .
314 .
315 .\" HTML <a name="newlines"></a>
316 .SH NEWLINES
317 .rs
318 .sp
319 PCRE supports five different conventions for indicating line breaks in
320 strings: a single CR (carriage return) character, a single LF (linefeed)
321 character, the two-character sequence CRLF, any of the three preceding, or any
322 Unicode newline sequence. The Unicode newline sequences are the three just
323 mentioned, plus the single characters VT (vertical tab, U+000B), FF (form feed,
324 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
325 (paragraph separator, U+2029).
326 .P
327 Each of the first three conventions is used by at least one operating system as
328 its standard newline sequence. When PCRE is built, a default can be specified.
329 The default default is LF, which is the Unix standard. When PCRE is run, the
330 default can be overridden, either when a pattern is compiled, or when it is
331 matched.
332 .P
333 At compile time, the newline convention can be specified by the \fIoptions\fP
334 argument of \fBpcre_compile()\fP, or it can be specified by special text at the
335 start of the pattern itself; this overrides any other settings. See the
336 .\" HREF
337 \fBpcrepattern\fP
338 .\"
339 page for details of the special character sequences.
340 .P
341 In the PCRE documentation the word "newline" is used to mean "the character or
342 pair of characters that indicate a line break". The choice of newline
343 convention affects the handling of the dot, circumflex, and dollar
344 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
345 recognized line ending sequence, the match position advancement for a
346 non-anchored pattern. There is more detail about this in the
347 .\" HTML <a href="#execoptions">
348 .\" </a>
349 section on \fBpcre_exec()\fP options
350 .\"
351 below.
352 .P
353 The choice of newline convention does not affect the interpretation of
354 the \en or \er escape sequences, nor does it affect what \eR matches, which is
355 controlled in a similar way, but by separate options.
356 .
357 .
358 .SH MULTITHREADING
359 .rs
360 .sp
361 The PCRE functions can be used in multi-threading applications, with the
362 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
363 \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
364 callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
365 .P
366 The compiled form of a regular expression is not altered during matching, so
367 the same compiled pattern can safely be used by several threads at once.
368 .P
369 If the just-in-time optimization feature is being used, it needs separate
370 memory stack areas for each thread. See the
371 .\" HREF
372 \fBpcrejit\fP
373 .\"
374 documentation for more details.
375 .
376 .
377 .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
378 .rs
379 .sp
380 The compiled form of a regular expression can be saved and re-used at a later
381 time, possibly by a different program, and even on a host other than the one on
382 which it was compiled. Details are given in the
383 .\" HREF
384 \fBpcreprecompile\fP
385 .\"
386 documentation, which includes a description of the
387 \fBpcre_pattern_to_host_byte_order()\fP function. However, compiling a regular
388 expression with one version of PCRE for use with a different version is not
389 guaranteed to work and may cause crashes.
390 .
391 .
392 .SH "CHECKING BUILD-TIME OPTIONS"
393 .rs
394 .sp
395 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
396 .PP
397 The function \fBpcre_config()\fP makes it possible for a PCRE client to
398 discover which optional features have been compiled into the PCRE library. The
399 .\" HREF
400 \fBpcrebuild\fP
401 .\"
402 documentation has more details about these optional features.
403 .P
404 The first argument for \fBpcre_config()\fP is an integer, specifying which
405 information is required; the second argument is a pointer to a variable into
406 which the information is placed. The returned value is zero on success, or the
407 negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
408 not recognized. The following information is available:
409 .sp
410 PCRE_CONFIG_UTF8
411 .sp
412 The output is an integer that is set to one if UTF-8 support is available;
413 otherwise it is set to zero. This value should normally be given to the 8-bit
414 version of this function, \fBpcre_config()\fP. If it is given to the 16-bit
415 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
416 .sp
417 PCRE_CONFIG_UTF16
418 .sp
419 The output is an integer that is set to one if UTF-16 support is available;
420 otherwise it is set to zero. This value should normally be given to the 16-bit
421 version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
422 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
423 .sp
424 PCRE_CONFIG_UTF32
425 .sp
426 The output is an integer that is set to one if UTF-32 support is available;
427 otherwise it is set to zero. This value should normally be given to the 32-bit
428 version of this function, \fBpcre32_config()\fP. If it is given to the 8-bit
429 or 16-bit version of this function, the result is PCRE_ERROR_BADOPTION.
430 .sp
431 PCRE_CONFIG_UNICODE_PROPERTIES
432 .sp
433 The output is an integer that is set to one if support for Unicode character
434 properties is available; otherwise it is set to zero.
435 .sp
436 PCRE_CONFIG_JIT
437 .sp
438 The output is an integer that is set to one if support for just-in-time
439 compiling is available; otherwise it is set to zero.
440 .sp
441 PCRE_CONFIG_JITTARGET
442 .sp
443 The output is a pointer to a zero-terminated "const char *" string. If JIT
444 support is available, the string contains the name of the architecture for
445 which the JIT compiler is configured, for example "x86 32bit (little endian +
446 unaligned)". If JIT support is not available, the result is NULL.
447 .sp
448 PCRE_CONFIG_NEWLINE
449 .sp
450 The output is an integer whose value specifies the default character sequence
451 that is recognized as meaning "newline". The values that are supported in
452 ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for
453 ANYCRLF, and -1 for ANY. In EBCDIC environments, CR, ANYCRLF, and ANY yield the
454 same values. However, the value for LF is normally 21, though some EBCDIC
455 environments use 37. The corresponding values for CRLF are 3349 and 3365. The
456 default should normally correspond to the standard sequence for your operating
457 system.
458 .sp
459 PCRE_CONFIG_BSR
460 .sp
461 The output is an integer whose value indicates what character sequences the \eR
462 escape sequence matches by default. A value of 0 means that \eR matches any
463 Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
464 or CRLF. The default can be overridden when a pattern is compiled or matched.
465 .sp
466 PCRE_CONFIG_LINK_SIZE
467 .sp
468 The output is an integer that contains the number of bytes used for internal
469 linkage in compiled regular expressions. For the 8-bit library, the value can
470 be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
471 a number of bytes. For the 32-bit library, the value is either 2 or 4 and is
472 still a number of bytes. The default value of 2 is sufficient for all but the
473 most massive patterns, since it allows the compiled pattern to be up to 64K in
474 size. Larger values allow larger regular expressions to be compiled, at the
475 expense of slower matching.
476 .sp
477 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
478 .sp
479 The output is an integer that contains the threshold above which the POSIX
480 interface uses \fBmalloc()\fP for output vectors. Further details are given in
481 the
482 .\" HREF
483 \fBpcreposix\fP
484 .\"
485 documentation.
486 .sp
487 PCRE_CONFIG_MATCH_LIMIT
488 .sp
489 The output is a long integer that gives the default limit for the number of
490 internal matching function calls in a \fBpcre_exec()\fP execution. Further
491 details are given with \fBpcre_exec()\fP below.
492 .sp
493 PCRE_CONFIG_MATCH_LIMIT_RECURSION
494 .sp
495 The output is a long integer that gives the default limit for the depth of
496 recursion when calling the internal matching function in a \fBpcre_exec()\fP
497 execution. Further details are given with \fBpcre_exec()\fP below.
498 .sp
499 PCRE_CONFIG_STACKRECURSE
500 .sp
501 The output is an integer that is set to one if internal recursion when running
502 \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
503 to remember their state. This is the usual way that PCRE is compiled. The
504 output is zero if PCRE was compiled to use blocks of data on the heap instead
505 of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
506 \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
507 avoiding the use of the stack.
508 .
509 .
510 .SH "COMPILING A PATTERN"
511 .rs
512 .sp
513 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
514 .ti +5n
515 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
516 .ti +5n
517 .B const unsigned char *\fItableptr\fP);
518 .sp
519 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
520 .ti +5n
521 .B int *\fIerrorcodeptr\fP,
522 .ti +5n
523 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
524 .ti +5n
525 .B const unsigned char *\fItableptr\fP);
526 .P
527 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
528 called to compile a pattern into an internal form. The only difference between
529 the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
530 \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
531 too much repetition, we refer just to \fBpcre_compile()\fP below, but the
532 information applies equally to \fBpcre_compile2()\fP.
533 .P
534 The pattern is a C string terminated by a binary zero, and is passed in the
535 \fIpattern\fP argument. A pointer to a single block of memory that is obtained
536 via \fBpcre_malloc\fP is returned. This contains the compiled code and related
537 data. The \fBpcre\fP type is defined for the returned block; this is a typedef
538 for a structure whose contents are not externally defined. It is up to the
539 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
540 .P
541 Although the compiled code of a PCRE regex is relocatable, that is, it does not
542 depend on memory location, the complete \fBpcre\fP data block is not
543 fully relocatable, because it may contain a copy of the \fItableptr\fP
544 argument, which is an address (see below).
545 .P
546 The \fIoptions\fP argument contains various bit settings that affect the
547 compilation. It should be zero if no options are required. The available
548 options are described below. Some of them (in particular, those that are
549 compatible with Perl, but some others as well) can also be set and unset from
550 within the pattern (see the detailed description in the
551 .\" HREF
552 \fBpcrepattern\fP
553 .\"
554 documentation). For those options that can be different in different parts of
555 the pattern, the contents of the \fIoptions\fP argument specifies their
556 settings at the start of compilation and execution. The PCRE_ANCHORED,
557 PCRE_BSR_\fIxxx\fP, PCRE_NEWLINE_\fIxxx\fP, PCRE_NO_UTF8_CHECK, and
558 PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
559 compile time.
560 .P
561 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
562 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
563 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
564 error message. This is a static string that is part of the library. You must
565 not try to free it. Normally, the offset from the start of the pattern to the
566 byte that was being processed when the error was discovered is placed in the
567 variable pointed to by \fIerroffset\fP, which must not be NULL (if it is, an
568 immediate error is given). However, for an invalid UTF-8 string, the offset is
569 that of the first byte of the failing character.
570 .P
571 Some errors are not detected until the whole pattern has been scanned; in these
572 cases, the offset passed back is the length of the pattern. Note that the
573 offset is in bytes, not characters, even in UTF-8 mode. It may sometimes point
574 into the middle of a UTF-8 character.
575 .P
576 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
577 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
578 returned via this argument in the event of an error. This is in addition to the
579 textual error message. Error codes and messages are listed below.
580 .P
581 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
582 character tables that are built when PCRE is compiled, using the default C
583 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
584 call to \fBpcre_maketables()\fP. This value is stored with the compiled
585 pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
586 passed to it. For more discussion, see the section on locale support below.
587 .P
588 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
589 .sp
590 pcre *re;
591 const char *error;
592 int erroffset;
593 re = pcre_compile(
594 "^A.*Z", /* the pattern */
595 0, /* default options */
596 &error, /* for error message */
597 &erroffset, /* for error offset */
598 NULL); /* use default character tables */
599 .sp
600 The following names for option bits are defined in the \fBpcre.h\fP header
601 file:
602 .sp
603 PCRE_ANCHORED
604 .sp
605 If this bit is set, the pattern is forced to be "anchored", that is, it is
606 constrained to match only at the first matching point in the string that is
607 being searched (the "subject string"). This effect can also be achieved by
608 appropriate constructs in the pattern itself, which is the only way to do it in
609 Perl.
610 .sp
611 PCRE_AUTO_CALLOUT
612 .sp
613 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
614 all with number 255, before each pattern item. For discussion of the callout
615 facility, see the
616 .\" HREF
617 \fBpcrecallout\fP
618 .\"
619 documentation.
620 .sp
621 PCRE_BSR_ANYCRLF
622 PCRE_BSR_UNICODE
623 .sp
624 These options (which are mutually exclusive) control what the \eR escape
625 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
626 match any Unicode newline sequence. The default is specified when PCRE is
627 built. It can be overridden from within the pattern, or by setting an option
628 when a compiled pattern is matched.
629 .sp
630 PCRE_CASELESS
631 .sp
632 If this bit is set, letters in the pattern match both upper and lower case
633 letters. It is equivalent to Perl's /i option, and it can be changed within a
634 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
635 concept of case for characters whose values are less than 128, so caseless
636 matching is always possible. For characters with higher values, the concept of
637 case is supported if PCRE is compiled with Unicode property support, but not
638 otherwise. If you want to use caseless matching for characters 128 and above,
639 you must ensure that PCRE is compiled with Unicode property support as well as
640 with UTF-8 support.
641 .sp
642 PCRE_DOLLAR_ENDONLY
643 .sp
644 If this bit is set, a dollar metacharacter in the pattern matches only at the
645 end of the subject string. Without this option, a dollar also matches
646 immediately before a newline at the end of the string (but not before any other
647 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
648 There is no equivalent to this option in Perl, and no way to set it within a
649 pattern.
650 .sp
651 PCRE_DOTALL
652 .sp
653 If this bit is set, a dot metacharacter in the pattern matches a character of
654 any value, including one that indicates a newline. However, it only ever
655 matches one character, even if newlines are coded as CRLF. Without this option,
656 a dot does not match when the current position is at a newline. This option is
657 equivalent to Perl's /s option, and it can be changed within a pattern by a
658 (?s) option setting. A negative class such as [^a] always matches newline
659 characters, independent of the setting of this option.
660 .sp
661 PCRE_DUPNAMES
662 .sp
663 If this bit is set, names used to identify capturing subpatterns need not be
664 unique. This can be helpful for certain types of pattern when it is known that
665 only one instance of the named subpattern can ever be matched. There are more
666 details of named subpatterns below; see also the
667 .\" HREF
668 \fBpcrepattern\fP
669 .\"
670 documentation.
671 .sp
672 PCRE_EXTENDED
673 .sp
674 If this bit is set, white space data characters in the pattern are totally
675 ignored except when escaped or inside a character class. White space does not
676 include the VT character (code 11). In addition, characters between an
677 unescaped # outside a character class and the next newline, inclusive, are also
678 ignored. This is equivalent to Perl's /x option, and it can be changed within a
679 pattern by a (?x) option setting.
680 .P
681 Which characters are interpreted as newlines is controlled by the options
682 passed to \fBpcre_compile()\fP or by a special sequence at the start of the
683 pattern, as described in the section entitled
684 .\" HTML <a href="pcrepattern.html#newlines">
685 .\" </a>
686 "Newline conventions"
687 .\"
688 in the \fBpcrepattern\fP documentation. Note that the end of this type of
689 comment is a literal newline sequence in the pattern; escape sequences that
690 happen to represent a newline do not count.
691 .P
692 This option makes it possible to include comments inside complicated patterns.
693 Note, however, that this applies only to data characters. White space characters
694 may never appear within special character sequences in a pattern, for example
695 within the sequence (?( that introduces a conditional subpattern.
696 .sp
697 PCRE_EXTRA
698 .sp
699 This option was invented in order to turn on additional functionality of PCRE
700 that is incompatible with Perl, but it is currently of very little use. When
701 set, any backslash in a pattern that is followed by a letter that has no
702 special meaning causes an error, thus reserving these combinations for future
703 expansion. By default, as in Perl, a backslash followed by a letter with no
704 special meaning is treated as a literal. (Perl can, however, be persuaded to
705 give an error for this, by running it with the -w option.) There are at present
706 no other features controlled by this option. It can also be set by a (?X)
707 option setting within a pattern.
708 .sp
709 PCRE_FIRSTLINE
710 .sp
711 If this option is set, an unanchored pattern is required to match before or at
712 the first newline in the subject string, though the matched text may continue
713 over the newline.
714 .sp
715 PCRE_JAVASCRIPT_COMPAT
716 .sp
717 If this option is set, PCRE's behaviour is changed in some ways so that it is
718 compatible with JavaScript rather than Perl. The changes are as follows:
719 .P
720 (1) A lone closing square bracket in a pattern causes a compile-time error,
721 because this is illegal in JavaScript (by default it is treated as a data
722 character). Thus, the pattern AB]CD becomes illegal when this option is set.
723 .P
724 (2) At run time, a back reference to an unset subpattern group matches an empty
725 string (by default this causes the current matching alternative to fail). A
726 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
727 an "a" in the subject), whereas it fails by default, for Perl compatibility.
728 .P
729 (3) \eU matches an upper case "U" character; by default \eU causes a compile
730 time error (Perl uses \eU to upper case subsequent characters).
731 .P
732 (4) \eu matches a lower case "u" character unless it is followed by four
733 hexadecimal digits, in which case the hexadecimal number defines the code point
734 to match. By default, \eu causes a compile time error (Perl uses it to upper
735 case the following character).
736 .P
737 (5) \ex matches a lower case "x" character unless it is followed by two
738 hexadecimal digits, in which case the hexadecimal number defines the code point
739 to match. By default, as in Perl, a hexadecimal number is always expected after
740 \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
741 binary zero character followed by z).
742 .sp
743 PCRE_MULTILINE
744 .sp
745 By default, PCRE treats the subject string as consisting of a single line of
746 characters (even if it actually contains newlines). The "start of line"
747 metacharacter (^) matches only at the start of the string, while the "end of
748 line" metacharacter ($) matches only at the end of the string, or before a
749 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
750 Perl.
751 .P
752 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
753 match immediately following or immediately before internal newlines in the
754 subject string, respectively, as well as at the very start and end. This is
755 equivalent to Perl's /m option, and it can be changed within a pattern by a
756 (?m) option setting. If there are no newlines in a subject string, or no
757 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
758 .sp
759 PCRE_NEWLINE_CR
760 PCRE_NEWLINE_LF
761 PCRE_NEWLINE_CRLF
762 PCRE_NEWLINE_ANYCRLF
763 PCRE_NEWLINE_ANY
764 .sp
765 These options override the default newline definition that was chosen when PCRE
766 was built. Setting the first or the second specifies that a newline is
767 indicated by a single character (CR or LF, respectively). Setting
768 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
769 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
770 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
771 that any Unicode newline sequence should be recognized.
772 .P
773 In an ASCII/Unicode environment, the Unicode newline sequences are the three
774 just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
775 feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
776 (paragraph separator, U+2029). For the 8-bit library, the last two are
777 recognized only in UTF-8 mode.
778 .P
779 When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
780 CR is 0x0d, the same as ASCII. However, the character code for LF is normally
781 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
782 not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
783 less than 256. For more details, see the
784 .\" HREF
785 \fBpcrebuild\fP
786 .\"
787 documentation.
788 .P
789 The newline setting in the options word uses three bits that are treated
790 as a number, giving eight possibilities. Currently only six are used (default
791 plus the five values above). This means that if you set more than one newline
792 option, the combination may or may not be sensible. For example,
793 PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
794 other combinations may yield unused numbers and cause an error.
795 .P
796 The only time that a line break in a pattern is specially recognized when
797 compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
798 and so are ignored in this mode. Also, an unescaped # outside a character class
799 indicates a comment that lasts until after the next line break sequence. In
800 other circumstances, line break sequences in patterns are treated as literal
801 data.
802 .P
803 The newline option that is set at compile time becomes the default that is used
804 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
805 .sp
806 PCRE_NO_AUTO_CAPTURE
807 .sp
808 If this option is set, it disables the use of numbered capturing parentheses in
809 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
810 were followed by ?: but named parentheses can still be used for capturing (and
811 they acquire numbers in the usual way). There is no equivalent of this option
812 in Perl.
813 .sp
814 NO_START_OPTIMIZE
815 .sp
816 This is an option that acts at matching time; that is, it is really an option
817 for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
818 it is remembered with the compiled pattern and assumed at matching time. For
819 details see the discussion of PCRE_NO_START_OPTIMIZE
820 .\" HTML <a href="#execoptions">
821 .\" </a>
822 below.
823 .\"
824 .sp
825 PCRE_UCP
826 .sp
827 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
828 \ew, and some of the POSIX character classes. By default, only ASCII characters
829 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
830 classify characters. More details are given in the section on
831 .\" HTML <a href="pcre.html#genericchartypes">
832 .\" </a>
833 generic character types
834 .\"
835 in the
836 .\" HREF
837 \fBpcrepattern\fP
838 .\"
839 page. If you set PCRE_UCP, matching one of the items it affects takes much
840 longer. The option is available only if PCRE has been compiled with Unicode
841 property support.
842 .sp
843 PCRE_UNGREEDY
844 .sp
845 This option inverts the "greediness" of the quantifiers so that they are not
846 greedy by default, but become greedy if followed by "?". It is not compatible
847 with Perl. It can also be set by a (?U) option setting within the pattern.
848 .sp
849 PCRE_UTF8
850 .sp
851 This option causes PCRE to regard both the pattern and the subject as strings
852 of UTF-8 characters instead of single-byte strings. However, it is available
853 only when PCRE is built to include UTF support. If not, the use of this option
854 provokes an error. Details of how this option changes the behaviour of PCRE are
855 given in the
856 .\" HREF
857 \fBpcreunicode\fP
858 .\"
859 page.
860 .sp
861 PCRE_NO_UTF8_CHECK
862 .sp
863 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8
864 string is automatically checked. There is a discussion about the
865 .\" HTML <a href="pcreunicode.html#utf8strings">
866 .\" </a>
867 validity of UTF-8 strings
868 .\"
869 in the
870 .\" HREF
871 \fBpcreunicode\fP
872 .\"
873 page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
874 error. If you already know that your pattern is valid, and you want to skip
875 this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
876 When it is set, the effect of passing an invalid UTF-8 string as a pattern is
877 undefined. It may cause your program to crash. Note that this option can also
878 be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress the
879 validity checking of subject strings.
880 .
881 .
882 .SH "COMPILATION ERROR CODES"
883 .rs
884 .sp
885 The following table lists the error codes than may be returned by
886 \fBpcre_compile2()\fP, along with the error messages that may be returned by
887 both compiling functions. Note that error messages are always 8-bit ASCII
888 strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
889 have fallen out of use. To avoid confusion, they have not been re-used.
890 .sp
891 0 no error
892 1 \e at end of pattern
893 2 \ec at end of pattern
894 3 unrecognized character follows \e
895 4 numbers out of order in {} quantifier
896 5 number too big in {} quantifier
897 6 missing terminating ] for character class
898 7 invalid escape sequence in character class
899 8 range out of order in character class
900 9 nothing to repeat
901 10 [this code is not in use]
902 11 internal error: unexpected repeat
903 12 unrecognized character after (? or (?-
904 13 POSIX named classes are supported only within a class
905 14 missing )
906 15 reference to non-existent subpattern
907 16 erroffset passed as NULL
908 17 unknown option bit(s) set
909 18 missing ) after comment
910 19 [this code is not in use]
911 20 regular expression is too large
912 21 failed to get memory
913 22 unmatched parentheses
914 23 internal error: code overflow
915 24 unrecognized character after (?<
916 25 lookbehind assertion is not fixed length
917 26 malformed number or name after (?(
918 27 conditional group contains more than two branches
919 28 assertion expected after (?(
920 29 (?R or (?[+-]digits must be followed by )
921 30 unknown POSIX class name
922 31 POSIX collating elements are not supported
923 32 this version of PCRE is compiled without UTF support
924 33 [this code is not in use]
925 34 character value in \ex{...} sequence is too large
926 35 invalid condition (?(0)
927 36 \eC not allowed in lookbehind assertion
928 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
929 38 number after (?C is > 255
930 39 closing ) for (?C expected
931 40 recursive call could loop indefinitely
932 41 unrecognized character after (?P
933 42 syntax error in subpattern name (missing terminator)
934 43 two named subpatterns have the same name
935 44 invalid UTF-8 string (specifically UTF-8)
936 45 support for \eP, \ep, and \eX has not been compiled
937 46 malformed \eP or \ep sequence
938 47 unknown property name after \eP or \ep
939 48 subpattern name is too long (maximum 32 characters)
940 49 too many named subpatterns (maximum 10000)
941 50 [this code is not in use]
942 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
943 52 internal error: overran compiling workspace
944 53 internal error: previously-checked referenced subpattern
945 not found
946 54 DEFINE group contains more than one branch
947 55 repeating a DEFINE group is not allowed
948 56 inconsistent NEWLINE options
949 57 \eg is not followed by a braced, angle-bracketed, or quoted
950 name/number or by a plain number
951 58 a numbered reference must not be zero
952 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
953 60 (*VERB) not recognized
954 61 number is too big
955 62 subpattern name expected
956 63 digit expected after (?+
957 64 ] is an invalid data character in JavaScript compatibility mode
958 65 different names for subpatterns of the same number are
959 not allowed
960 66 (*MARK) must have an argument
961 67 this version of PCRE is not compiled with Unicode property
962 support
963 68 \ec must be followed by an ASCII character
964 69 \ek is not followed by a braced, angle-bracketed, or quoted name
965 70 internal error: unknown opcode in find_fixedlength()
966 71 \eN is not supported in a class
967 72 too many forward references
968 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
969 74 invalid UTF-16 string (specifically UTF-16)
970 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
971 76 character value in \eu.... sequence is too large
972 77 invalid UTF-32 string (specifically UTF-32)
973 .sp
974 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
975 be used if the limits were changed when PCRE was built.
976 .
977 .
978 .\" HTML <a name="studyingapattern"></a>
979 .SH "STUDYING A PATTERN"
980 .rs
981 .sp
982 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
983 .ti +5n
984 .B const char **\fIerrptr\fP);
985 .PP
986 If a compiled pattern is going to be used several times, it is worth spending
987 more time analyzing it in order to speed up the time taken for matching. The
988 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
989 argument. If studying the pattern produces additional information that will
990 help speed up matching, \fBpcre_study()\fP returns a pointer to a
991 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
992 results of the study.
993 .P
994 The returned value from \fBpcre_study()\fP can be passed directly to
995 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
996 also contains other fields that can be set by the caller before the block is
997 passed; these are described
998 .\" HTML <a href="#extradata">
999 .\" </a>
1000 below
1001 .\"
1002 in the section on matching a pattern.
1003 .P
1004 If studying the pattern does not produce any useful information,
1005 \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
1006 calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
1007 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
1008 if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
1009 returns a \fBpcre_extra\fP block even if studying did not find any additional
1010 information. It may still return NULL, however, if an error occurs in
1011 \fBpcre_study()\fP.
1012 .P
1013 The second argument of \fBpcre_study()\fP contains option bits. There are three
1014 further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1015 .sp
1016 PCRE_STUDY_JIT_COMPILE
1017 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1018 PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1019 .sp
1020 If any of these are set, and the just-in-time compiler is available, the
1021 pattern is further compiled into machine code that executes much faster than
1022 the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
1023 compiler is not available, these options are ignored. All undefined bits in the
1024 \fIoptions\fP argument must be zero.
1025 .P
1026 JIT compilation is a heavyweight optimization. It can take some time for
1027 patterns to be analyzed, and for one-off matches and simple patterns the
1028 benefit of faster execution might be offset by a much slower study time.
1029 Not all patterns can be optimized by the JIT compiler. For those that cannot be
1030 handled, matching automatically falls back to the \fBpcre_exec()\fP
1031 interpreter. For more details, see the
1032 .\" HREF
1033 \fBpcrejit\fP
1034 .\"
1035 documentation.
1036 .P
1037 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
1038 studying succeeds (even if no data is returned), the variable it points to is
1039 set to NULL. Otherwise it is set to point to a textual error message. This is a
1040 static string that is part of the library. You must not try to free it. You
1041 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
1042 sure that it has run successfully.
1043 .P
1044 When you are finished with a pattern, you can free the memory used for the
1045 study data by calling \fBpcre_free_study()\fP. This function was added to the
1046 API for release 8.20. For earlier versions, the memory could be freed with
1047 \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1048 where JIT optimization is not used, but it is advisable to change to the new
1049 function when convenient.
1050 .P
1051 This is a typical way in which \fBpcre_study\fP() is used (except that in a
1052 real application there should be tests for errors):
1053 .sp
1054 int rc;
1055 pcre *re;
1056 pcre_extra *sd;
1057 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1058 sd = pcre_study(
1059 re, /* result of pcre_compile() */
1060 0, /* no options */
1061 &error); /* set to NULL or points to a message */
1062 rc = pcre_exec( /* see below for details of pcre_exec() options */
1063 re, sd, "subject", 7, 0, 0, ovector, 30);
1064 ...
1065 pcre_free_study(sd);
1066 pcre_free(re);
1067 .sp
1068 Studying a pattern does two things: first, a lower bound for the length of
1069 subject string that is needed to match the pattern is computed. This does not
1070 mean that there are any strings of that length that match, but it does
1071 guarantee that no shorter strings match. The value is used to avoid wasting
1072 time by trying to match strings that are shorter than the lower bound. You can
1073 find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
1074 .P
1075 Studying a pattern is also useful for non-anchored patterns that do not have a
1076 single fixed starting character. A bitmap of possible starting bytes is
1077 created. This speeds up finding a position in the subject at which to start
1078 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1079 In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1080 .P
1081 These two optimizations apply to both \fBpcre_exec()\fP and
1082 \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1083 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option
1084 when calling \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP, but if this is done,
1085 JIT execution is also disabled. You might want to do this if your pattern
1086 contains callouts or (*MARK) and you want to make use of these facilities in
1087 cases where matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
1088 .\" HTML <a href="#execoptions">
1089 .\" </a>
1090 below.
1091 .\"
1092 .
1093 .
1094 .\" HTML <a name="localesupport"></a>
1095 .SH "LOCALE SUPPORT"
1096 .rs
1097 .sp
1098 PCRE handles caseless matching, and determines whether characters are letters,
1099 digits, or whatever, by reference to a set of tables, indexed by character
1100 value. When running in UTF-8 mode, this applies only to characters
1101 with codes less than 128. By default, higher-valued codes never match escapes
1102 such as \ew or \ed, but they can be tested with \ep if PCRE is built with
1103 Unicode character property support. Alternatively, the PCRE_UCP option can be
1104 set at compile time; this causes \ew and friends to use Unicode property
1105 support instead of built-in tables. The use of locales with Unicode is
1106 discouraged. If you are handling characters with codes greater than 128, you
1107 should either use UTF-8 and Unicode, or use locales, but not try to mix the
1108 two.
1109 .P
1110 PCRE contains an internal set of tables that are used when the final argument
1111 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1112 Normally, the internal tables recognize only ASCII characters. However, when
1113 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1114 default "C" locale of the local system, which may cause them to be different.
1115 .P
1116 The internal tables can always be overridden by tables supplied by the
1117 application that calls PCRE. These may be created in a different locale from
1118 the default. As more and more applications change to using Unicode, the need
1119 for this locale support is expected to die away.
1120 .P
1121 External tables are built by calling the \fBpcre_maketables()\fP function,
1122 which has no arguments, in the relevant locale. The result can then be passed
1123 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
1124 example, to build and use tables that are appropriate for the French locale
1125 (where accented characters with values greater than 128 are treated as letters),
1126 the following code could be used:
1127 .sp
1128 setlocale(LC_CTYPE, "fr_FR");
1129 tables = pcre_maketables();
1130 re = pcre_compile(..., tables);
1131 .sp
1132 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1133 are using Windows, the name for the French locale is "french".
1134 .P
1135 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1136 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1137 that the memory containing the tables remains available for as long as it is
1138 needed.
1139 .P
1140 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1141 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1142 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
1143 pattern, compilation, studying and matching all happen in the same locale, but
1144 different patterns can be compiled in different locales.
1145 .P
1146 It is possible to pass a table pointer or NULL (indicating the use of the
1147 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
1148 this facility could be used to match a pattern in a different locale from the
1149 one in which it was compiled. Passing table pointers at run time is discussed
1150 below in the section on matching a pattern.
1151 .
1152 .
1153 .\" HTML <a name="infoaboutpattern"></a>
1154 .SH "INFORMATION ABOUT A PATTERN"
1155 .rs
1156 .sp
1157 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1158 .ti +5n
1159 .B int \fIwhat\fP, void *\fIwhere\fP);
1160 .PP
1161 The \fBpcre_fullinfo()\fP function returns information about a compiled
1162 pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1163 library at version 8.30, after more than 10 years of obsolescence.
1164 .P
1165 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1166 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1167 the pattern was not studied. The third argument specifies which piece of
1168 information is required, and the fourth argument is a pointer to a variable
1169 to receive the data. The yield of the function is zero for success, or one of
1170 the following negative numbers:
1171 .sp
1172 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1173 the argument \fIwhere\fP was NULL
1174 PCRE_ERROR_BADMAGIC the "magic number" was not found
1175 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1176 endianness
1177 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1178 .sp
1179 The "magic number" is placed at the start of each compiled pattern as an simple
1180 check against passing an arbitrary memory pointer. The endianness error can
1181 occur if a compiled pattern is saved and reloaded on a different host. Here is
1182 a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1183 pattern:
1184 .sp
1185 int rc;
1186 size_t length;
1187 rc = pcre_fullinfo(
1188 re, /* result of pcre_compile() */
1189 sd, /* result of pcre_study(), or NULL */
1190 PCRE_INFO_SIZE, /* what is required */
1191 &length); /* where to put the data */
1192 .sp
1193 The possible values for the third argument are defined in \fBpcre.h\fP, and are
1194 as follows:
1195 .sp
1196 PCRE_INFO_BACKREFMAX
1197 .sp
1198 Return the number of the highest back reference in the pattern. The fourth
1199 argument should point to an \fBint\fP variable. Zero is returned if there are
1200 no back references.
1201 .sp
1202 PCRE_INFO_CAPTURECOUNT
1203 .sp
1204 Return the number of capturing subpatterns in the pattern. The fourth argument
1205 should point to an \fBint\fP variable.
1206 .sp
1207 PCRE_INFO_DEFAULT_TABLES
1208 .sp
1209 Return a pointer to the internal default character tables within PCRE. The
1210 fourth argument should point to an \fBunsigned char *\fP variable. This
1211 information call is provided for internal use by the \fBpcre_study()\fP
1212 function. External callers can cause PCRE to use its internal tables by passing
1213 a NULL table pointer.
1214 .sp
1215 PCRE_INFO_FIRSTBYTE
1216 .sp
1217 Return information about the first data unit of any matched string, for a
1218 non-anchored pattern. (The name of this option refers to the 8-bit library,
1219 where data units are bytes.) The fourth argument should point to an \fBint\fP
1220 variable.
1221 .P
1222 If there is a fixed first value, for example, the letter "c" from a pattern
1223 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1224 value is always less than 256. In the 16-bit library the value can be up to
1225 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1226 .P
1227 If there is no fixed first value, and if either
1228 .sp
1229 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1230 starts with "^", or
1231 .sp
1232 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1233 (if it were set, the pattern would be anchored),
1234 .sp
1235 -1 is returned, indicating that the pattern matches only at the start of a
1236 subject string or after any newline within the string. Otherwise -2 is
1237 returned. For anchored patterns, -2 is returned.
1238 .P
1239 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1240 to return the full 32-bit range of the character, this value is deprecated;
1241 instead the PCRE_INFO_FIRSTLITERALSET and PCRE_INFO_FIRSTLITERAL values should
1242 be used.
1243 .sp
1244 PCRE_INFO_FIRSTTABLE
1245 .sp
1246 If the pattern was studied, and this resulted in the construction of a 256-bit
1247 table indicating a fixed set of values for the first data unit in any matching
1248 string, a pointer to the table is returned. Otherwise NULL is returned. The
1249 fourth argument should point to an \fBunsigned char *\fP variable.
1250 .sp
1251 PCRE_INFO_HASCRORLF
1252 .sp
1253 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1254 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1255 explicit match is either a literal CR or LF character, or \er or \en.
1256 .sp
1257 PCRE_INFO_JCHANGED
1258 .sp
1259 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1260 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1261 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1262 .sp
1263 PCRE_INFO_JIT
1264 .sp
1265 Return 1 if the pattern was studied with one of the JIT options, and
1266 just-in-time compiling was successful. The fourth argument should point to an
1267 \fBint\fP variable. A return value of 0 means that JIT support is not available
1268 in this version of PCRE, or that the pattern was not studied with a JIT option,
1269 or that the JIT compiler could not handle this particular pattern. See the
1270 .\" HREF
1271 \fBpcrejit\fP
1272 .\"
1273 documentation for details of what can and cannot be handled.
1274 .sp
1275 PCRE_INFO_JITSIZE
1276 .sp
1277 If the pattern was successfully studied with a JIT option, return the size of
1278 the JIT compiled code, otherwise return zero. The fourth argument should point
1279 to a \fBsize_t\fP variable.
1280 .sp
1281 PCRE_INFO_LASTLITERAL
1282 .sp
1283 Return the value of the rightmost literal data unit that must exist in any
1284 matched string, other than at its start, if such a value has been recorded. The
1285 fourth argument should point to an \fBint\fP variable. If there is no such
1286 value, -1 is returned. For anchored patterns, a last literal value is recorded
1287 only if it follows something of variable length. For example, for the pattern
1288 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1289 is -1.
1290 .P
1291 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1292 to return the full 32-bit range of the character, this value is deprecated;
1293 instead the PCRE_INFO_LASTLITERAL2SET and PCRE_INFO_LASTLITERAL2 values should
1294 be used.
1295 .sp
1296 PCRE_INFO_MAXLOOKBEHIND
1297 .sp
1298 Return the number of characters (NB not bytes) in the longest lookbehind
1299 assertion in the pattern. Note that the simple assertions \eb and \eB require a
1300 one-character lookbehind. This information is useful when doing multi-segment
1301 matching using the partial matching facilities.
1302 .sp
1303 PCRE_INFO_MINLENGTH
1304 .sp
1305 If the pattern was studied and a minimum length for matching subject strings
1306 was computed, its value is returned. Otherwise the returned value is -1. The
1307 value is a number of characters, which in UTF-8 mode may be different from the
1308 number of bytes. The fourth argument should point to an \fBint\fP variable. A
1309 non-negative value is a lower bound to the length of any matching string. There
1310 may not be any strings of that length that do actually match, but every string
1311 that does match is at least that long.
1312 .sp
1313 PCRE_INFO_NAMECOUNT
1314 PCRE_INFO_NAMEENTRYSIZE
1315 PCRE_INFO_NAMETABLE
1316 .sp
1317 PCRE supports the use of named as well as numbered capturing parentheses. The
1318 names are just an additional way of identifying the parentheses, which still
1319 acquire numbers. Several convenience functions such as
1320 \fBpcre_get_named_substring()\fP are provided for extracting captured
1321 substrings by name. It is also possible to extract the data directly, by first
1322 converting the name to a number in order to access the correct pointers in the
1323 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1324 you need to use the name-to-number map, which is described by these three
1325 values.
1326 .P
1327 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1328 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1329 entry; both of these return an \fBint\fP value. The entry size depends on the
1330 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1331 entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1332 the first two bytes of each entry are the number of the capturing parenthesis,
1333 most significant byte first. In the 16-bit library, the pointer points to
1334 16-bit data units, the first of which contains the parenthesis number.
1335 In the 32-bit library, the pointer points to 32-bit data units, the first of
1336 which contains the parenthesis number. The rest
1337 of the entry is the corresponding name, zero terminated.
1338 .P
1339 The names are in alphabetical order. Duplicate names may appear if (?| is used
1340 to create multiple groups with the same number, as described in the
1341 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1342 .\" </a>
1343 section on duplicate subpattern numbers
1344 .\"
1345 in the
1346 .\" HREF
1347 \fBpcrepattern\fP
1348 .\"
1349 page. Duplicate names for subpatterns with different numbers are permitted only
1350 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1351 table in the order in which they were found in the pattern. In the absence of
1352 (?| this is the order of increasing number; when (?| is used this is not
1353 necessarily the case because later subpatterns may have lower numbers.
1354 .P
1355 As a simple example of the name/number table, consider the following pattern
1356 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1357 space - including newlines - is ignored):
1358 .sp
1359 .\" JOIN
1360 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1361 (?<month>\ed\ed) - (?<day>\ed\ed) )
1362 .sp
1363 There are four named subpatterns, so the table has four entries, and each entry
1364 in the table is eight bytes long. The table is as follows, with non-printing
1365 bytes shows in hexadecimal, and undefined bytes shown as ??:
1366 .sp
1367 00 01 d a t e 00 ??
1368 00 05 d a y 00 ?? ??
1369 00 04 m o n t h 00
1370 00 02 y e a r 00 ??
1371 .sp
1372 When writing code to extract data from named subpatterns using the
1373 name-to-number map, remember that the length of the entries is likely to be
1374 different for each compiled pattern.
1375 .sp
1376 PCRE_INFO_OKPARTIAL
1377 .sp
1378 Return 1 if the pattern can be used for partial matching with
1379 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1380 \fBint\fP variable. From release 8.00, this always returns 1, because the
1381 restrictions that previously applied to partial matching have been lifted. The
1382 .\" HREF
1383 \fBpcrepartial\fP
1384 .\"
1385 documentation gives details of partial matching.
1386 .sp
1387 PCRE_INFO_OPTIONS
1388 .sp
1389 Return a copy of the options with which the pattern was compiled. The fourth
1390 argument should point to an \fBunsigned long int\fP variable. These option bits
1391 are those specified in the call to \fBpcre_compile()\fP, modified by any
1392 top-level option settings at the start of the pattern itself. In other words,
1393 they are the options that will be in force when matching starts. For example,
1394 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1395 result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1396 .P
1397 A pattern is automatically anchored by PCRE if all of its top-level
1398 alternatives begin with one of the following:
1399 .sp
1400 ^ unless PCRE_MULTILINE is set
1401 \eA always
1402 \eG always
1403 .\" JOIN
1404 .* if PCRE_DOTALL is set and there are no back
1405 references to the subpattern in which .* appears
1406 .sp
1407 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1408 \fBpcre_fullinfo()\fP.
1409 .sp
1410 PCRE_INFO_SIZE
1411 .sp
1412 Return the size of the compiled pattern in bytes (for both libraries). The
1413 fourth argument should point to a \fBsize_t\fP variable. This value does not
1414 include the size of the \fBpcre\fP structure that is returned by
1415 \fBpcre_compile()\fP. The value that is passed as the argument to
1416 \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1417 place the compiled data is the value returned by this option plus the size of
1418 the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1419 does not alter the value returned by this option.
1420 .sp
1421 PCRE_INFO_STUDYSIZE
1422 .sp
1423 Return the size in bytes of the data block pointed to by the \fIstudy_data\fP
1424 field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP is NULL, or there is no
1425 study data, zero is returned. The fourth argument should point to a
1426 \fBsize_t\fP variable. The \fIstudy_data\fP field is set by \fBpcre_study()\fP
1427 to record information that will speed up matching (see the section entitled
1428 .\" HTML <a href="#studyingapattern">
1429 .\" </a>
1430 "Studying a pattern"
1431 .\"
1432 above). The format of the \fIstudy_data\fP block is private, but its length
1433 is made available via this option so that it can be saved and restored (see the
1434 .\" HREF
1435 \fBpcreprecompile\fP
1436 .\"
1437 documentation for details).
1438 .sp
1439 PCRE_INFO_FIRSTLITERALSET
1440 .sp
1441 Return information about the first data unit of any matched string, for a
1442 non-anchored pattern. The fourth argument should point to an \fBint\fP
1443 variable.
1444 .P
1445 If there is a fixed first value, for example, the letter "c" from a pattern
1446 such as (cat|cow|coyote), 1 is returned, and the character value can be
1447 retrieved using PCRE_INFO_FIRSTLITERAL.
1448 .P
1449 If there is no fixed first value, and if either
1450 .sp
1451 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1452 starts with "^", or
1453 .sp
1454 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1455 (if it were set, the pattern would be anchored),
1456 .sp
1457 2 is returned, indicating that the pattern matches only at the start of a
1458 subject string or after any newline within the string. Otherwise 0 is
1459 returned. For anchored patterns, 0 is returned.
1460 .sp
1461 PCRE_INFO_FIRSTLITERAL
1462 .sp
1463 Return the fixed first character value, if PCRE_INFO_FIRSTLITERALSET returned 1;
1464 otherwise returns 0. The fourth argument should point to an \fBuint_t\fP
1465 variable.
1466 .P
1467 In the 8-bit library, the value is always less than 256. In the 16-bit library
1468 the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1469 can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1470 .P
1471 If there is no fixed first value, and if either
1472 .sp
1473 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1474 starts with "^", or
1475 .sp
1476 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1477 (if it were set, the pattern would be anchored),
1478 .sp
1479 -1 is returned, indicating that the pattern matches only at the start of a
1480 subject string or after any newline within the string. Otherwise -2 is
1481 returned. For anchored patterns, -2 is returned.
1482 .sp
1483 PCRE_INFO_LASTLITERAL2SET
1484 .sp
1485 Returns 1 if there is a rightmost literal data unit that must exist in any matched
1486 string, other than at its start. The fourth argument should point to an \fBint\fP
1487 variable. If there is no such value, 0 is returned. If returning 1, the character
1488 value itself can be retrieved using PCRE_INFO_LASTLITERAL2.
1489 .P
1490 For anchored patterns, a last literal value is recorded only if it follows something
1491 of variable length. For example, for the pattern /^a\ed+z\ed+/ the returned value
1492 1 (with "z" returned from PCRE_INFO_LASTLITERAL2), but for /^a\edz\ed/ the returned
1493 value is 0.
1494 .sp
1495 PCRE_INFO_LASTLITERAL2
1496 .sp
1497 Return the value of the rightmost literal data unit that must exist in any
1498 matched string, other than at its start, if such a value has been recorded. The
1499 fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1500 value, 0 is returned.
1501 .
1502 .
1503 .SH "REFERENCE COUNTS"
1504 .rs
1505 .sp
1506 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1507 .PP
1508 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1509 data block that contains a compiled pattern. It is provided for the benefit of
1510 applications that operate in an object-oriented manner, where different parts
1511 of the application may be using the same compiled pattern, but you want to free
1512 the block when they are all done.
1513 .P
1514 When a pattern is compiled, the reference count field is initialized to zero.
1515 It is changed only by calling this function, whose action is to add the
1516 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1517 function is the new value. However, the value of the count is constrained to
1518 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1519 it is forced to the appropriate limit value.
1520 .P
1521 Except when it is zero, the reference count is not correctly preserved if a
1522 pattern is compiled on one host and then transferred to a host whose byte-order
1523 is different. (This seems a highly unlikely scenario.)
1524 .
1525 .
1526 .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1527 .rs
1528 .sp
1529 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1530 .ti +5n
1531 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1532 .ti +5n
1533 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1534 .P
1535 The function \fBpcre_exec()\fP is called to match a subject string against a
1536 compiled pattern, which is passed in the \fIcode\fP argument. If the
1537 pattern was studied, the result of the study should be passed in the
1538 \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1539 and \fIextra\fP arguments as many times as you like, in order to match
1540 different subject strings with the same pattern.
1541 .P
1542 This function is the main matching facility of the library, and it operates in
1543 a Perl-like manner. For specialist use there is also an alternative matching
1544 function, which is described
1545 .\" HTML <a href="#dfamatch">
1546 .\" </a>
1547 below
1548 .\"
1549 in the section about the \fBpcre_dfa_exec()\fP function.
1550 .P
1551 In most applications, the pattern will have been compiled (and optionally
1552 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1553 possible to save compiled patterns and study data, and then use them later
1554 in different processes, possibly even on different hosts. For a discussion
1555 about this, see the
1556 .\" HREF
1557 \fBpcreprecompile\fP
1558 .\"
1559 documentation.
1560 .P
1561 Here is an example of a simple call to \fBpcre_exec()\fP:
1562 .sp
1563 int rc;
1564 int ovector[30];
1565 rc = pcre_exec(
1566 re, /* result of pcre_compile() */
1567 NULL, /* we didn't study the pattern */
1568 "some string", /* the subject string */
1569 11, /* the length of the subject string */
1570 0, /* start at offset 0 in the subject */
1571 0, /* default options */
1572 ovector, /* vector of integers for substring information */
1573 30); /* number of elements (NOT size in bytes) */
1574 .
1575 .
1576 .\" HTML <a name="extradata"></a>
1577 .SS "Extra data for \fBpcre_exec()\fR"
1578 .rs
1579 .sp
1580 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1581 data block. The \fBpcre_study()\fP function returns such a block (when it
1582 doesn't return NULL), but you can also create one for yourself, and pass
1583 additional information in it. The \fBpcre_extra\fP block contains the following
1584 fields (not necessarily in this order):
1585 .sp
1586 unsigned long int \fIflags\fP;
1587 void *\fIstudy_data\fP;
1588 void *\fIexecutable_jit\fP;
1589 unsigned long int \fImatch_limit\fP;
1590 unsigned long int \fImatch_limit_recursion\fP;
1591 void *\fIcallout_data\fP;
1592 const unsigned char *\fItables\fP;
1593 unsigned char **\fImark\fP;
1594 .sp
1595 In the 16-bit version of this structure, the \fImark\fP field has type
1596 "PCRE_UCHAR16 **".
1597 .sp
1598 In the 32-bit version of this structure, the \fImark\fP field has type
1599 "PCRE_UCHAR32 **".
1600 .P
1601 The \fIflags\fP field is used to specify which of the other fields are set. The
1602 flag bits are:
1603 .sp
1604 PCRE_EXTRA_CALLOUT_DATA
1605 PCRE_EXTRA_EXECUTABLE_JIT
1606 PCRE_EXTRA_MARK
1607 PCRE_EXTRA_MATCH_LIMIT
1608 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1609 PCRE_EXTRA_STUDY_DATA
1610 PCRE_EXTRA_TABLES
1611 .sp
1612 Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1613 the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1614 returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1615 should not set these yourself, but you may add to the block by setting other
1616 fields and their corresponding flag bits.
1617 .P
1618 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1619 vast amount of resources when running patterns that are not going to match,
1620 but which have a very large number of possibilities in their search trees. The
1621 classic example is a pattern that uses nested unlimited repeats.
1622 .P
1623 Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1624 calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1625 imposed on the number of times this function is called during a match, which
1626 has the effect of limiting the amount of backtracking that can take place. For
1627 patterns that are not anchored, the count restarts from zero for each position
1628 in the subject string.
1629 .P
1630 When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1631 with a JIT option, the way that the matching is executed is entirely different.
1632 However, there is still the possibility of runaway matching that goes on for a
1633 very long time, and so the \fImatch_limit\fP value is also used in this case
1634 (but in a different way) to limit how long the matching can continue.
1635 .P
1636 The default value for the limit can be set when PCRE is built; the default
1637 default is 10 million, which handles all but the most extreme cases. You can
1638 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1639 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1640 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1641 PCRE_ERROR_MATCHLIMIT.
1642 .P
1643 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1644 instead of limiting the total number of times that \fBmatch()\fP is called, it
1645 limits the depth of recursion. The recursion depth is a smaller number than the
1646 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1647 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1648 .P
1649 Limiting the recursion depth limits the amount of machine stack that can be
1650 used, or, when PCRE has been compiled to use memory on the heap instead of the
1651 stack, the amount of heap memory that can be used. This limit is not relevant,
1652 and is ignored, when matching is done using JIT compiled code.
1653 .P
1654 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1655 built; the default default is the same value as the default for
1656 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1657 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1658 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1659 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1660 .P
1661 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1662 and is described in the
1663 .\" HREF
1664 \fBpcrecallout\fP
1665 .\"
1666 documentation.
1667 .P
1668 The \fItables\fP field is used to pass a character tables pointer to
1669 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1670 pattern. A non-NULL value is stored with the compiled pattern only if custom
1671 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1672 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1673 internal tables to be used. This facility is helpful when re-using patterns
1674 that have been saved after compiling with an external set of tables, because
1675 the external tables might be at a different address when \fBpcre_exec()\fP is
1676 called. See the
1677 .\" HREF
1678 \fBpcreprecompile\fP
1679 .\"
1680 documentation for a discussion of saving compiled patterns for later use.
1681 .P
1682 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1683 be set to point to a suitable variable. If the pattern contains any
1684 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1685 a name to pass back, a pointer to the name string (zero terminated) is placed
1686 in the variable pointed to by the \fImark\fP field. The names are within the
1687 compiled pattern; if you wish to retain such a name you must copy it before
1688 freeing the memory of a compiled pattern. If there is no name to pass back, the
1689 variable pointed to by the \fImark\fP field is set to NULL. For details of the
1690 backtracking control verbs, see the section entitled
1691 .\" HTML <a href="pcrepattern#backtrackcontrol">
1692 .\" </a>
1693 "Backtracking control"
1694 .\"
1695 in the
1696 .\" HREF
1697 \fBpcrepattern\fP
1698 .\"
1699 documentation.
1700 .
1701 .
1702 .\" HTML <a name="execoptions"></a>
1703 .SS "Option bits for \fBpcre_exec()\fP"
1704 .rs
1705 .sp
1706 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1707 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1708 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1709 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1710 PCRE_PARTIAL_SOFT.
1711 .P
1712 If the pattern was successfully studied with one of the just-in-time (JIT)
1713 compile options, the only supported options for JIT execution are
1714 PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1715 PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1716 unsupported option is used, JIT execution is disabled and the normal
1717 interpretive code in \fBpcre_exec()\fP is run.
1718 .sp
1719 PCRE_ANCHORED
1720 .sp
1721 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1722 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1723 to be anchored by virtue of its contents, it cannot be made unachored at
1724 matching time.
1725 .sp
1726 PCRE_BSR_ANYCRLF
1727 PCRE_BSR_UNICODE
1728 .sp
1729 These options (which are mutually exclusive) control what the \eR escape
1730 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1731 match any Unicode newline sequence. These options override the choice that was
1732 made or defaulted when the pattern was compiled.
1733 .sp
1734 PCRE_NEWLINE_CR
1735 PCRE_NEWLINE_LF
1736 PCRE_NEWLINE_CRLF
1737 PCRE_NEWLINE_ANYCRLF
1738 PCRE_NEWLINE_ANY
1739 .sp
1740 These options override the newline definition that was chosen or defaulted when
1741 the pattern was compiled. For details, see the description of
1742 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1743 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1744 the way the match position is advanced after a match failure for an unanchored
1745 pattern.
1746 .P
1747 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1748 match attempt for an unanchored pattern fails when the current position is at a
1749 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1750 characters, the match position is advanced by two characters instead of one, in
1751 other words, to after the CRLF.
1752 .P
1753 The above rule is a compromise that makes the most common cases work as
1754 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1755 set), it does not match the string "\er\enA" because, after failing at the
1756 start, it skips both the CR and the LF before retrying. However, the pattern
1757 [\er\en]A does match that string, because it contains an explicit CR or LF
1758 reference, and so advances only by one character after the first failure.
1759 .P
1760 An explicit match for CR of LF is either a literal appearance of one of those
1761 characters, or one of the \er or \en escape sequences. Implicit matches such as
1762 [^X] do not count, nor does \es (which includes CR and LF in the characters
1763 that it matches).
1764 .P
1765 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1766 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1767 .sp
1768 PCRE_NOTBOL
1769 .sp
1770 This option specifies that first character of the subject string is not the
1771 beginning of a line, so the circumflex metacharacter should not match before
1772 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1773 never to match. This option affects only the behaviour of the circumflex
1774 metacharacter. It does not affect \eA.
1775 .sp
1776 PCRE_NOTEOL
1777 .sp
1778 This option specifies that the end of the subject string is not the end of a
1779 line, so the dollar metacharacter should not match it nor (except in multiline
1780 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1781 compile time) causes dollar never to match. This option affects only the
1782 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1783 .sp
1784 PCRE_NOTEMPTY
1785 .sp
1786 An empty string is not considered to be a valid match if this option is set. If
1787 there are alternatives in the pattern, they are tried. If all the alternatives
1788 match the empty string, the entire match fails. For example, if the pattern
1789 .sp
1790 a?b?
1791 .sp
1792 is applied to a string not beginning with "a" or "b", it matches an empty
1793 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1794 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1795 .sp
1796 PCRE_NOTEMPTY_ATSTART
1797 .sp
1798 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1799 the start of the subject is permitted. If the pattern is anchored, such a match
1800 can occur only if the pattern contains \eK.
1801 .P
1802 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1803 does make a special case of a pattern match of the empty string within its
1804 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1805 emulate Perl's behaviour after matching a null string by first trying the match
1806 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1807 if that fails, by advancing the starting offset (see below) and trying an
1808 ordinary match again. There is some code that demonstrates how to do this in
1809 the
1810 .\" HREF
1811 \fBpcredemo\fP
1812 .\"
1813 sample program. In the most general case, you have to check to see if the
1814 newline convention recognizes CRLF as a newline, and if so, and the current
1815 character is CR followed by LF, advance the starting offset by two characters
1816 instead of one.
1817 .sp
1818 PCRE_NO_START_OPTIMIZE
1819 .sp
1820 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1821 a match, in order to speed up the process. For example, if it is known that an
1822 unanchored match must start with a specific character, it searches the subject
1823 for that character, and fails immediately if it cannot find it, without
1824 actually running the main matching function. This means that a special item
1825 such as (*COMMIT) at the start of a pattern is not considered until after a
1826 suitable starting point for the match has been found. When callouts or (*MARK)
1827 items are in use, these "start-up" optimizations can cause them to be skipped
1828 if the pattern is never actually used. The start-up optimizations are in effect
1829 a pre-scan of the subject that takes place before the pattern is run.
1830 .P
1831 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1832 causing performance to suffer, but ensuring that in cases where the result is
1833 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1834 are considered at every possible starting position in the subject string. If
1835 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1836 time. The use of PCRE_NO_START_OPTIMIZE disables JIT execution; when it is set,
1837 matching is always done using interpretively.
1838 .P
1839 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1840 Consider the pattern
1841 .sp
1842 (*COMMIT)ABC
1843 .sp
1844 When this is compiled, PCRE records the fact that a match must start with the
1845 character "A". Suppose the subject string is "DEFABC". The start-up
1846 optimization scans along the subject, finds "A" and runs the first match
1847 attempt from there. The (*COMMIT) item means that the pattern must match the
1848 current starting position, which in this case, it does. However, if the same
1849 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1850 subject string does not happen. The first match attempt is run starting from
1851 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1852 the overall result is "no match". If the pattern is studied, more start-up
1853 optimizations may be used. For example, a minimum length for the subject may be
1854 recorded. Consider the pattern
1855 .sp
1856 (*MARK:A)(X|Y)
1857 .sp
1858 The minimum length for a match is one character. If the subject is "ABC", there
1859 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1860 If the pattern is studied, the final attempt does not take place, because PCRE
1861 knows that the subject is too short, and so the (*MARK) is never encountered.
1862 In this case, studying the pattern does not affect the overall match result,
1863 which is still "no match", but it does affect the auxiliary information that is
1864 returned.
1865 .sp
1866 PCRE_NO_UTF8_CHECK
1867 .sp
1868 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1869 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1870 The entire string is checked before any other processing takes place. The value
1871 of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1872 UTF-8 character. There is a discussion about the
1873 .\" HTML <a href="pcreunicode.html#utf8strings">
1874 .\" </a>
1875 validity of UTF-8 strings
1876 .\"
1877 in the
1878 .\" HREF
1879 \fBpcreunicode\fP
1880 .\"
1881 page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1882 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1883 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1884 cases, information about the precise nature of the error may also be returned
1885 (see the descriptions of these errors in the section entitled \fIError return
1886 values from\fP \fBpcre_exec()\fP
1887 .\" HTML <a href="#errorlist">
1888 .\" </a>
1889 below).
1890 .\"
1891 If \fIstartoffset\fP contains a value that does not point to the start of a
1892 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1893 returned.
1894 .P
1895 If you already know that your subject is valid, and you want to skip these
1896 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1897 calling \fBpcre_exec()\fP. You might want to do this for the second and
1898 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1899 all the matches in a single subject string. However, you should be sure that
1900 the value of \fIstartoffset\fP points to the start of a character (or the end
1901 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1902 invalid string as a subject or an invalid value of \fIstartoffset\fP is
1903 undefined. Your program may crash.
1904 .sp
1905 PCRE_PARTIAL_HARD
1906 PCRE_PARTIAL_SOFT
1907 .sp
1908 These options turn on the partial matching feature. For backwards
1909 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1910 occurs if the end of the subject string is reached successfully, but there are
1911 not enough subject characters to complete the match. If this happens when
1912 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1913 testing any remaining alternatives. Only if no complete match can be found is
1914 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1915 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1916 but only if no complete match can be found.
1917 .P
1918 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1919 partial match is found, \fBpcre_exec()\fP immediately returns
1920 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1921 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1922 important that an alternative complete match.
1923 .P
1924 In both cases, the portion of the string that was inspected when the partial
1925 match was found is set as the first matching string. There is a more detailed
1926 discussion of partial and multi-segment matching, with examples, in the
1927 .\" HREF
1928 \fBpcrepartial\fP
1929 .\"
1930 documentation.
1931 .
1932 .
1933 .SS "The string to be matched by \fBpcre_exec()\fP"
1934 .rs
1935 .sp
1936 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1937 \fIsubject\fP, a length in bytes in \fIlength\fP, and a starting byte offset
1938 in \fIstartoffset\fP. If this is negative or greater than the length of the
1939 subject, \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting
1940 offset is zero, the search for a match starts at the beginning of the subject,
1941 and this is by far the most common case. In UTF-8 mode, the byte offset must
1942 point to the start of a UTF-8 character (or the end of the subject). Unlike the
1943 pattern string, the subject may contain binary zero bytes.
1944 .P
1945 A non-zero starting offset is useful when searching for another match in the
1946 same subject by calling \fBpcre_exec()\fP again after a previous success.
1947 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1948 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1949 lookbehind. For example, consider the pattern
1950 .sp
1951 \eBiss\eB
1952 .sp
1953 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1954 the current position in the subject is not a word boundary.) When applied to
1955 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1956 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1957 subject, namely "issipi", it does not match, because \eB is always false at the
1958 start of the subject, which is deemed to be a word boundary. However, if
1959 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1960 set to 4, it finds the second occurrence of "iss" because it is able to look
1961 behind the starting point to discover that it is preceded by a letter.
1962 .P
1963 Finding all the matches in a subject is tricky when the pattern can match an
1964 empty string. It is possible to emulate Perl's /g behaviour by first trying the
1965 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1966 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1967 and trying an ordinary match again. There is some code that demonstrates how to
1968 do this in the
1969 .\" HREF
1970 \fBpcredemo\fP
1971 .\"
1972 sample program. In the most general case, you have to check to see if the
1973 newline convention recognizes CRLF as a newline, and if so, and the current
1974 character is CR followed by LF, advance the starting offset by two characters
1975 instead of one.
1976 .P
1977 If a non-zero starting offset is passed when the pattern is anchored, one
1978 attempt to match at the given offset is made. This can only succeed if the
1979 pattern does not require the match to be at the start of the subject.
1980 .
1981 .
1982 .SS "How \fBpcre_exec()\fP returns captured substrings"
1983 .rs
1984 .sp
1985 In general, a pattern matches a certain portion of the subject, and in
1986 addition, further substrings from the subject may be picked out by parts of the
1987 pattern. Following the usage in Jeffrey Friedl's book, this is called
1988 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1989 a fragment of a pattern that picks out a substring. PCRE supports several other
1990 kinds of parenthesized subpattern that do not cause substrings to be captured.
1991 .P
1992 Captured substrings are returned to the caller via a vector of integers whose
1993 address is passed in \fIovector\fP. The number of elements in the vector is
1994 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
1995 argument is NOT the size of \fIovector\fP in bytes.
1996 .P
1997 The first two-thirds of the vector is used to pass back captured substrings,
1998 each substring using a pair of integers. The remaining third of the vector is
1999 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
2000 and is not available for passing back information. The number passed in
2001 \fIovecsize\fP should always be a multiple of three. If it is not, it is
2002 rounded down.
2003 .P
2004 When a match is successful, information about captured substrings is returned
2005 in pairs of integers, starting at the beginning of \fIovector\fP, and
2006 continuing up to two-thirds of its length at the most. The first element of
2007 each pair is set to the byte offset of the first character in a substring, and
2008 the second is set to the byte offset of the first character after the end of a
2009 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
2010 mode. They are not character counts.
2011 .P
2012 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
2013 portion of the subject string matched by the entire pattern. The next pair is
2014 used for the first capturing subpattern, and so on. The value returned by
2015 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
2016 For example, if two substrings have been captured, the returned value is 3. If
2017 there are no capturing subpatterns, the return value from a successful match is
2018 1, indicating that just the first pair of offsets has been set.
2019 .P
2020 If a capturing subpattern is matched repeatedly, it is the last portion of the
2021 string that it matched that is returned.
2022 .P
2023 If the vector is too small to hold all the captured substring offsets, it is
2024 used as far as possible (up to two-thirds of its length), and the function
2025 returns a value of zero. If neither the actual string matched nor any captured
2026 substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2027 passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2028 back references and the \fIovector\fP is not big enough to remember the related
2029 substrings, PCRE has to get additional memory for use during matching. Thus it
2030 is usually advisable to supply an \fIovector\fP of reasonable size.
2031 .P
2032 There are some cases where zero is returned (indicating vector overflow) when
2033 in fact the vector is exactly the right size for the final match. For example,
2034 consider the pattern
2035 .sp
2036 (a)(?:(b)c|bd)
2037 .sp
2038 If a vector of 6 elements (allowing for only 1 captured substring) is given
2039 with subject string "abd", \fBpcre_exec()\fP will try to set the second
2040 captured string, thereby recording a vector overflow, before failing to match
2041 "c" and backing up to try the second alternative. The zero return, however,
2042 does correctly indicate that the maximum number of slots (namely 2) have been
2043 filled. In similar cases where there is temporary overflow, but the final
2044 number of used slots is actually less than the maximum, a non-zero value is
2045 returned.
2046 .P
2047 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
2048 subpatterns there are in a compiled pattern. The smallest size for
2049 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
2050 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
2051 .P
2052 It is possible for capturing subpattern number \fIn+1\fP to match some part of
2053 the subject when subpattern \fIn\fP has not been used at all. For example, if
2054 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2055 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2056 happens, both values in the offset pairs corresponding to unused subpatterns
2057 are set to -1.
2058 .P
2059 Offset values that correspond to unused subpatterns at the end of the
2060 expression are also set to -1. For example, if the string "abc" is matched
2061 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2062 return from the function is 2, because the highest used capturing subpattern
2063 number is 1, and the offsets for for the second and third capturing subpatterns
2064 (assuming the vector is large enough, of course) are set to -1.
2065 .P
2066 \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
2067 correspond to capturing parentheses in the pattern are never changed. That is,
2068 if a pattern contains \fIn\fP capturing parentheses, no more than
2069 \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
2070 elements (in the first two-thirds) retain whatever values they previously had.
2071 .P
2072 Some convenience functions are provided for extracting the captured substrings
2073 as separate strings. These are described below.
2074 .
2075 .
2076 .\" HTML <a name="errorlist"></a>
2077 .SS "Error return values from \fBpcre_exec()\fP"
2078 .rs
2079 .sp
2080 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
2081 defined in the header file:
2082 .sp
2083 PCRE_ERROR_NOMATCH (-1)
2084 .sp
2085 The subject string did not match the pattern.
2086 .sp
2087 PCRE_ERROR_NULL (-2)
2088 .sp
2089 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
2090 NULL and \fIovecsize\fP was not zero.
2091 .sp
2092 PCRE_ERROR_BADOPTION (-3)
2093 .sp
2094 An unrecognized bit was set in the \fIoptions\fP argument.
2095 .sp
2096 PCRE_ERROR_BADMAGIC (-4)
2097 .sp
2098 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2099 the case when it is passed a junk pointer and to detect when a pattern that was
2100 compiled in an environment of one endianness is run in an environment with the
2101 other endianness. This is the error that PCRE gives when the magic number is
2102 not present.
2103 .sp
2104 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2105 .sp
2106 While running the pattern match, an unknown item was encountered in the
2107 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2108 of the compiled pattern.
2109 .sp
2110 PCRE_ERROR_NOMEMORY (-6)
2111 .sp
2112 If a pattern contains back references, but the \fIovector\fP that is passed to
2113 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2114 gets a block of memory at the start of matching to use for this purpose. If the
2115 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2116 automatically freed at the end of matching.
2117 .P
2118 This error is also given if \fBpcre_stack_malloc()\fP fails in
2119 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2120 \fB--disable-stack-for-recursion\fP.
2121 .sp
2122 PCRE_ERROR_NOSUBSTRING (-7)
2123 .sp
2124 This error is used by the \fBpcre_copy_substring()\fP,
2125 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2126 below). It is never returned by \fBpcre_exec()\fP.
2127 .sp
2128 PCRE_ERROR_MATCHLIMIT (-8)
2129 .sp
2130 The backtracking limit, as specified by the \fImatch_limit\fP field in a
2131 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2132 above.
2133 .sp
2134 PCRE_ERROR_CALLOUT (-9)
2135 .sp
2136 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2137 use by callout functions that want to yield a distinctive error code. See the
2138 .\" HREF
2139 \fBpcrecallout\fP
2140 .\"
2141 documentation for details.
2142 .sp
2143 PCRE_ERROR_BADUTF8 (-10)
2144 .sp
2145 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2146 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2147 (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2148 UTF-8 character is placed in the first element, and a reason code is placed in
2149 the second element. The reason codes are listed in the
2150 .\" HTML <a href="#badutf8reasons">
2151 .\" </a>
2152 following section.
2153 .\"
2154 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2155 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2156 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2157 .sp
2158 PCRE_ERROR_BADUTF8_OFFSET (-11)
2159 .sp
2160 The UTF-8 byte sequence that was passed as a subject was checked and found to
2161 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2162 \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2163 end of the subject.
2164 .sp
2165 PCRE_ERROR_PARTIAL (-12)
2166 .sp
2167 The subject string did not match, but it did match partially. See the
2168 .\" HREF
2169 \fBpcrepartial\fP
2170 .\"
2171 documentation for details of partial matching.
2172 .sp
2173 PCRE_ERROR_BADPARTIAL (-13)
2174 .sp
2175 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2176 option was used with a compiled pattern containing items that were not
2177 supported for partial matching. From release 8.00 onwards, there are no
2178 restrictions on partial matching.
2179 .sp
2180 PCRE_ERROR_INTERNAL (-14)
2181 .sp
2182 An unexpected internal error has occurred. This error could be caused by a bug
2183 in PCRE or by overwriting of the compiled pattern.
2184 .sp
2185 PCRE_ERROR_BADCOUNT (-15)
2186 .sp
2187 This error is given if the value of the \fIovecsize\fP argument is negative.
2188 .sp
2189 PCRE_ERROR_RECURSIONLIMIT (-21)
2190 .sp
2191 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2192 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2193 description above.
2194 .sp
2195 PCRE_ERROR_BADNEWLINE (-23)
2196 .sp
2197 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2198 .sp
2199 PCRE_ERROR_BADOFFSET (-24)
2200 .sp
2201 The value of \fIstartoffset\fP was negative or greater than the length of the
2202 subject, that is, the value in \fIlength\fP.
2203 .sp
2204 PCRE_ERROR_SHORTUTF8 (-25)
2205 .sp
2206 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2207 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2208 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2209 fact sufficient to detect this case, but this special error code for
2210 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2211 retained for backwards compatibility.
2212 .sp
2213 PCRE_ERROR_RECURSELOOP (-26)
2214 .sp
2215 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2216 the pattern. Specifically, it means that either the whole pattern or a
2217 subpattern has been called recursively for the second time at the same position
2218 in the subject string. Some simple patterns that might do this are detected and
2219 faulted at compile time, but more complicated cases, in particular mutual
2220 recursions between two different subpatterns, cannot be detected until run
2221 time.
2222 .sp
2223 PCRE_ERROR_JIT_STACKLIMIT (-27)
2224 .sp
2225 This error is returned when a pattern that was successfully studied using a
2226 JIT compile option is being matched, but the memory available for the
2227 just-in-time processing stack is not large enough. See the
2228 .\" HREF
2229 \fBpcrejit\fP
2230 .\"
2231 documentation for more details.
2232 .sp
2233 PCRE_ERROR_BADMODE (-28)
2234 .sp
2235 This error is given if a pattern that was compiled by the 8-bit library is
2236 passed to a 16-bit or 32-bit library function, or vice versa.
2237 .sp
2238 PCRE_ERROR_BADENDIANNESS (-29)
2239 .sp
2240 This error is given if a pattern that was compiled and saved is reloaded on a
2241 host with different endianness. The utility function
2242 \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2243 so that it runs on the new host.
2244 .P
2245 Error numbers -16 to -20, -22, and -30 are not used by \fBpcre_exec()\fP.
2246 .
2247 .
2248 .\" HTML <a name="badutf8reasons"></a>
2249 .SS "Reason codes for invalid UTF-8 strings"
2250 .rs
2251 .sp
2252 This section applies only to the 8-bit library. The corresponding information
2253 for the 16-bit library is given in the
2254 .\" HREF
2255 \fBpcre16\fP
2256 .\"
2257 page. The corresponding information for the 32-bit library is given in the
2258 .\" HREF
2259 \fBpcre32\fP
2260 .\"
2261 page.
2262 .P
2263 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2264 PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2265 least 2, the offset of the start of the invalid UTF-8 character is placed in
2266 the first output vector element (\fIovector[0]\fP) and a reason code is placed
2267 in the second element (\fIovector[1]\fP). The reason codes are given names in
2268 the \fBpcre.h\fP header file:
2269 .sp
2270 PCRE_UTF8_ERR1
2271 PCRE_UTF8_ERR2
2272 PCRE_UTF8_ERR3
2273 PCRE_UTF8_ERR4
2274 PCRE_UTF8_ERR5
2275 .sp
2276 The string ends with a truncated UTF-8 character; the code specifies how many
2277 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2278 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2279 allows for up to 6 bytes, and this is checked first; hence the possibility of
2280 4 or 5 missing bytes.
2281 .sp
2282 PCRE_UTF8_ERR6
2283 PCRE_UTF8_ERR7
2284 PCRE_UTF8_ERR8
2285 PCRE_UTF8_ERR9
2286 PCRE_UTF8_ERR10
2287 .sp
2288 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2289 character do not have the binary value 0b10 (that is, either the most
2290 significant bit is 0, or the next bit is 1).
2291 .sp
2292 PCRE_UTF8_ERR11
2293 PCRE_UTF8_ERR12
2294 .sp
2295 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2296 these code points are excluded by RFC 3629.
2297 .sp
2298 PCRE_UTF8_ERR13
2299 .sp
2300 A 4-byte character has a value greater than 0x10fff; these code points are
2301 excluded by RFC 3629.
2302 .sp
2303 PCRE_UTF8_ERR14
2304 .sp
2305 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2306 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2307 from UTF-8.
2308 .sp
2309 PCRE_UTF8_ERR15
2310 PCRE_UTF8_ERR16
2311 PCRE_UTF8_ERR17
2312 PCRE_UTF8_ERR18
2313 PCRE_UTF8_ERR19
2314 .sp
2315 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2316 value that can be represented by fewer bytes, which is invalid. For example,
2317 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2318 one byte.
2319 .sp
2320 PCRE_UTF8_ERR20
2321 .sp
2322 The two most significant bits of the first byte of a character have the binary
2323 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2324 byte can only validly occur as the second or subsequent byte of a multi-byte
2325 character.
2326 .sp
2327 PCRE_UTF8_ERR21
2328 .sp
2329 The first byte of a character has the value 0xfe or 0xff. These values can
2330 never occur in a valid UTF-8 string.
2331 .sp
2332 PCRE_UTF8_ERR2
2333 .sp
2334 Non-character. These are the last two characters in each plane (0xfffe, 0xffff,
2335 0x1fffe, 0x1ffff .. 0x10fffe, 0x10ffff), and the characters 0xfdd0..0xfdef.
2336 .
2337 .
2338 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2339 .rs
2340 .sp
2341 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2342 .ti +5n
2343 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
2344 .ti +5n
2345 .B int \fIbuffersize\fP);
2346 .PP
2347 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2348 .ti +5n
2349 .B int \fIstringcount\fP, int \fIstringnumber\fP,
2350 .ti +5n
2351 .B const char **\fIstringptr\fP);
2352 .PP
2353 .B int pcre_get_substring_list(const char *\fIsubject\fP,
2354 .ti +5n
2355 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
2356 .PP
2357 Captured substrings can be accessed directly by using the offsets returned by
2358 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2359 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2360 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2361 as new, separate, zero-terminated strings. These functions identify substrings
2362 by number. The next section describes functions for extracting named
2363 substrings.
2364 .P
2365 A substring that contains a binary zero is correctly extracted and has a
2366 further zero added on the end, but the result is not, of course, a C string.
2367 However, you can process such a string by referring to the length that is
2368 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2369 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2370 for handling strings containing binary zeros, because the end of the final
2371 string is not independently indicated.
2372 .P
2373 The first three arguments are the same for all three of these functions:
2374 \fIsubject\fP is the subject string that has just been successfully matched,
2375 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2376 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2377 captured by the match, including the substring that matched the entire regular
2378 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2379 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2380 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2381 number of elements in the vector divided by three.
2382 .P
2383 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2384 extract a single substring, whose number is given as \fIstringnumber\fP. A
2385 value of zero extracts the substring that matched the entire pattern, whereas
2386 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2387 the string is placed in \fIbuffer\fP, whose length is given by
2388 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2389 obtained via \fBpcre_malloc\fP, and its address is returned via
2390 \fIstringptr\fP. The yield of the function is the length of the string, not
2391 including the terminating zero, or one of these error codes:
2392 .sp
2393 PCRE_ERROR_NOMEMORY (-6)
2394 .sp
2395 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2396 memory failed for \fBpcre_get_substring()\fP.
2397 .sp
2398 PCRE_ERROR_NOSUBSTRING (-7)
2399 .sp
2400 There is no substring whose number is \fIstringnumber\fP.
2401 .P
2402 The \fBpcre_get_substring_list()\fP function extracts all available substrings
2403 and builds a list of pointers to them. All this is done in a single block of
2404 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2405 is returned via \fIlistptr\fP, which is also the start of the list of string
2406 pointers. The end of the list is marked by a NULL pointer. The yield of the
2407 function is zero if all went well, or the error code
2408 .sp
2409 PCRE_ERROR_NOMEMORY (-6)
2410 .sp
2411 if the attempt to get the memory block failed.
2412 .P
2413 When any of these functions encounter a substring that is unset, which can
2414 happen when capturing subpattern number \fIn+1\fP matches some part of the
2415 subject, but subpattern \fIn\fP has not been used at all, they return an empty
2416 string. This can be distinguished from a genuine zero-length substring by
2417 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2418 substrings.
2419 .P
2420 The two convenience functions \fBpcre_free_substring()\fP and
2421 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2422 a previous call of \fBpcre_get_substring()\fP or
2423 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2424 the function pointed to by \fBpcre_free\fP, which of course could be called
2425 directly from a C program. However, PCRE is used in some situations where it is
2426 linked via a special interface to another programming language that cannot use
2427 \fBpcre_free\fP directly; it is for these cases that the functions are
2428 provided.
2429 .
2430 .
2431 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2432 .rs
2433 .sp
2434 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2435 .ti +5n
2436 .B const char *\fIname\fP);
2437 .PP
2438 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2439 .ti +5n
2440 .B const char *\fIsubject\fP, int *\fIovector\fP,
2441 .ti +5n
2442 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2443 .ti +5n
2444 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
2445 .PP
2446 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2447 .ti +5n
2448 .B const char *\fIsubject\fP, int *\fIovector\fP,
2449 .ti +5n
2450 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2451 .ti +5n
2452 .B const char **\fIstringptr\fP);
2453 .PP
2454 To extract a substring by name, you first have to find associated number.
2455 For example, for this pattern
2456 .sp
2457 (a+)b(?<xxx>\ed+)...
2458 .sp
2459 the number of the subpattern called "xxx" is 2. If the name is known to be
2460 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2461 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2462 pattern, and the second is the name. The yield of the function is the
2463 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2464 that name.
2465 .P
2466 Given the number, you can extract the substring directly, or use one of the
2467 functions described in the previous section. For convenience, there are also
2468 two functions that do the whole job.
2469 .P
2470 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2471 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2472 functions that extract by number. As these are described in the previous
2473 section, they are not re-described here. There are just two differences:
2474 .P
2475 First, instead of a substring number, a substring name is given. Second, there
2476 is an extra argument, given at the start, which is a pointer to the compiled
2477 pattern. This is needed in order to gain access to the name-to-number
2478 translation table.
2479 .P
2480 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2481 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2482 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2483 the behaviour may not be what you want (see the next section).
2484 .P
2485 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2486 subpatterns with the same number, as described in the
2487 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2488 .\" </a>
2489 section on duplicate subpattern numbers
2490 .\"
2491 in the
2492 .\" HREF
2493 \fBpcrepattern\fP
2494 .\"
2495 page, you cannot use names to distinguish the different subpatterns, because
2496 names are not included in the compiled code. The matching process uses only
2497 numbers. For this reason, the use of different names for subpatterns of the
2498 same number causes an error at compile time.
2499 .
2500 .
2501 .SH "DUPLICATE SUBPATTERN NAMES"
2502 .rs
2503 .sp
2504 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2505 .ti +5n
2506 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
2507 .PP
2508 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2509 are not required to be unique. (Duplicate names are always allowed for
2510 subpatterns with the same number, created by using the (?| feature. Indeed, if
2511 such subpatterns are named, they are required to use the same names.)
2512 .P
2513 Normally, patterns with duplicate names are such that in any one match, only
2514 one of the named subpatterns participates. An example is shown in the
2515 .\" HREF
2516 \fBpcrepattern\fP
2517 .\"
2518 documentation.
2519 .P
2520 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2521 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2522 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2523 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2524 returns one of the numbers that are associated with the name, but it is not
2525 defined which it is.
2526 .P
2527 If you want to get full details of all captured substrings for a given name,
2528 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2529 argument is the compiled pattern, and the second is the name. The third and
2530 fourth are pointers to variables which are updated by the function. After it
2531 has run, they point to the first and last entries in the name-to-number table
2532 for the given name. The function itself returns the length of each entry, or
2533 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2534 described above in the section entitled \fIInformation about a pattern\fP
2535 .\" HTML <a href="#infoaboutpattern">
2536 .\" </a>
2537 above.
2538 .\"
2539 Given all the relevant entries for the name, you can extract each of their
2540 numbers, and hence the captured data, if any.
2541 .
2542 .
2543 .SH "FINDING ALL POSSIBLE MATCHES"
2544 .rs
2545 .sp
2546 The traditional matching function uses a similar algorithm to Perl, which stops
2547 when it finds the first match, starting at a given point in the subject. If you
2548 want to find all possible matches, or the longest possible match, consider
2549 using the alternative matching function (see below) instead. If you cannot use
2550 the alternative function, but still need to find all possible matches, you
2551 can kludge it up by making use of the callout facility, which is described in
2552 the
2553 .\" HREF
2554 \fBpcrecallout\fP
2555 .\"
2556 documentation.
2557 .P
2558 What you have to do is to insert a callout right at the end of the pattern.
2559 When your callout function is called, extract and save the current matched
2560 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2561 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2562 will yield PCRE_ERROR_NOMATCH.
2563 .
2564 .
2565 .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2566 .rs
2567 .sp
2568 Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2569 stack, which in certain environments can be rather limited in size. Some users
2570 find it helpful to have an estimate of the amount of stack that is used by
2571 \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2572 .\" HREF
2573 \fBpcrestack\fP
2574 .\"
2575 documentation. The estimate that is output by \fBpcretest\fP when called with
2576 the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2577 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2578 .P
2579 Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2580 the negative error code PCRE_ERROR_NULL, but with this special combination of
2581 arguments, it returns instead a negative number whose absolute value is the
2582 approximate stack frame size in bytes. (A negative number is used so that it is
2583 clear that no match has happened.) The value is approximate because in some
2584 cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2585 additional variables on the stack.
2586 .P
2587 If PCRE has been compiled to use the heap instead of the stack for recursion,
2588 the value returned is the size of each block that is obtained from the heap.
2589 .
2590 .
2591 .\" HTML <a name="dfamatch"></a>
2592 .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2593 .rs
2594 .sp
2595 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2596 .ti +5n
2597 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2598 .ti +5n
2599 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2600 .ti +5n
2601 .B int *\fIworkspace\fP, int \fIwscount\fP);
2602 .P
2603 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2604 a compiled pattern, using a matching algorithm that scans the subject string
2605 just once, and does not backtrack. This has different characteristics to the
2606 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2607 patterns are not supported. Nevertheless, there are times when this kind of
2608 matching can be useful. For a discussion of the two matching algorithms, and a
2609 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2610 .\" HREF
2611 \fBpcrematching\fP
2612 .\"
2613 documentation.
2614 .P
2615 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2616 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2617 different way, and this is described below. The other common arguments are used
2618 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2619 here.
2620 .P
2621 The two additional arguments provide workspace for the function. The workspace
2622 vector should contain at least 20 elements. It is used for keeping track of
2623 multiple paths through the pattern tree. More workspace will be needed for
2624 patterns and subjects where there are a lot of potential matches.
2625 .P
2626 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2627 .sp
2628 int rc;
2629 int ovector[10];
2630 int wspace[20];
2631 rc = pcre_dfa_exec(
2632 re, /* result of pcre_compile() */
2633 NULL, /* we didn't study the pattern */
2634 "some string", /* the subject string */
2635 11, /* the length of the subject string */
2636 0, /* start at offset 0 in the subject */
2637 0, /* default options */
2638 ovector, /* vector of integers for substring information */
2639 10, /* number of elements (NOT size in bytes) */
2640 wspace, /* working space vector */
2641 20); /* number of elements (NOT size in bytes) */
2642 .
2643 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2644 .rs
2645 .sp
2646 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2647 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2648 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2649 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2650 PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2651 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2652 so their description is not repeated here.
2653 .sp
2654 PCRE_PARTIAL_HARD
2655 PCRE_PARTIAL_SOFT
2656 .sp
2657 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2658 details are slightly different. When PCRE_PARTIAL_HARD is set for
2659 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2660 is reached and there is still at least one matching possibility that requires
2661 additional characters. This happens even if some complete matches have also
2662 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2663 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2664 there have been no complete matches, but there is still at least one matching
2665 possibility. The portion of the string that was inspected when the longest
2666 partial match was found is set as the first matching string in both cases.
2667 There is a more detailed discussion of partial and multi-segment matching, with
2668 examples, in the
2669 .\" HREF
2670 \fBpcrepartial\fP
2671 .\"
2672 documentation.
2673 .sp
2674 PCRE_DFA_SHORTEST
2675 .sp
2676 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2677 soon as it has found one match. Because of the way the alternative algorithm
2678 works, this is necessarily the shortest possible match at the first possible
2679 matching point in the subject string.
2680 .sp
2681 PCRE_DFA_RESTART
2682 .sp
2683 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2684 again, with additional subject characters, and have it continue with the same
2685 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2686 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2687 before because data about the match so far is left in them after a partial
2688 match. There is more discussion of this facility in the
2689 .\" HREF
2690 \fBpcrepartial\fP
2691 .\"
2692 documentation.
2693 .
2694 .
2695 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2696 .rs
2697 .sp
2698 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2699 substring in the subject. Note, however, that all the matches from one run of
2700 the function start at the same point in the subject. The shorter matches are
2701 all initial substrings of the longer matches. For example, if the pattern
2702 .sp
2703 <.*>
2704 .sp
2705 is matched against the string
2706 .sp
2707 This is <something> <something else> <something further> no more
2708 .sp
2709 the three matched strings are
2710 .sp
2711 <something>
2712 <something> <something else>
2713 <something> <something else> <something further>
2714 .sp
2715 On success, the yield of the function is a number greater than zero, which is
2716 the number of matched substrings. The substrings themselves are returned in
2717 \fIovector\fP. Each string uses two elements; the first is the offset to the
2718 start, and the second is the offset to the end. In fact, all the strings have
2719 the same start offset. (Space could have been saved by giving this only once,
2720 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2721 returns data, even though the meaning of the strings is different.)
2722 .P
2723 The strings are returned in reverse order of length; that is, the longest
2724 matching string is given first. If there were too many matches to fit into
2725 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2726 the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2727 the entire \fIovector\fP for returning matched strings.
2728 .
2729 .
2730 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2731 .rs
2732 .sp
2733 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2734 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2735 described
2736 .\" HTML <a href="#errorlist">
2737 .\" </a>
2738 above.
2739 .\"
2740 There are in addition the following errors that are specific to
2741 \fBpcre_dfa_exec()\fP:
2742 .sp
2743 PCRE_ERROR_DFA_UITEM (-16)
2744 .sp
2745 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2746 that it does not support, for instance, the use of \eC or a back reference.
2747 .sp
2748 PCRE_ERROR_DFA_UCOND (-17)
2749 .sp
2750 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2751 uses a back reference for the condition, or a test for recursion in a specific
2752 group. These are not supported.
2753 .sp
2754 PCRE_ERROR_DFA_UMLIMIT (-18)
2755 .sp
2756 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2757 block that contains a setting of the \fImatch_limit\fP or
2758 \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2759 meaningless for DFA matching).
2760 .sp
2761 PCRE_ERROR_DFA_WSSIZE (-19)
2762 .sp
2763 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2764 \fIworkspace\fP vector.
2765 .sp
2766 PCRE_ERROR_DFA_RECURSE (-20)
2767 .sp
2768 When a recursive subpattern is processed, the matching function calls itself
2769 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2770 error is given if the output vector is not large enough. This should be
2771 extremely rare, as a vector of size 1000 is used.
2772 .sp
2773 PCRE_ERROR_DFA_BADRESTART (-30)
2774 .sp
2775 When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2776 some plausibility checks are made on the contents of the workspace, which
2777 should contain data about the previous partial match. If any of these checks
2778 fail, this error is given.
2779 .
2780 .
2781 .SH "SEE ALSO"
2782 .rs
2783 .sp
2784 \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2785 \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2786 \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2787 \fBpcrestack\fP(3).
2788 .
2789 .
2790 .SH AUTHOR
2791 .rs
2792 .sp
2793 .nf
2794 Philip Hazel
2795 University Computing Service
2796 Cambridge CB2 3QH, England.
2797 .fi
2798 .
2799 .
2800 .SH REVISION
2801 .rs
2802 .sp
2803 .nf
2804 Last updated: 07 September 2012
2805 Copyright (c) 1997-2012 University of Cambridge.
2806 .fi

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5