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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1191 - (show annotations)
Tue Oct 30 16:50:57 2012 UTC (6 years, 8 months ago) by ph10
File size: 121856 byte(s)
Some documentation updates.
1 .TH PCREAPI 3 "29 October 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 8.32, 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 and 32-bit functions operate in the same way as their 8-bit
148 counterparts; they just use different data types for their arguments and
149 results, and their names start with \fBpcre16_\fP or \fBpcre32_\fP instead of
150 \fBpcre_\fP. For every option that has UTF8 in its name (for example,
151 PCRE_UTF8), there are corresponding 16-bit and 32-bit names with UTF8 replaced
152 by UTF16 or UTF32, respectively. This facility is in fact just cosmetic; the
153 16-bit and 32-bit option names define the same bit values.
154 .P
155 References to bytes and UTF-8 in this document should be read as references to
156 16-bit data quantities and UTF-16 when using the 16-bit library, or 32-bit data
157 quantities and UTF-32 when using the 32-bit library, unless specified
158 otherwise. More details of the specific differences for the 16-bit and 32-bit
159 libraries are given in the
160 .\" HREF
161 \fBpcre16\fP
162 .\"
163 and
164 .\" HREF
165 \fBpcre32\fP
166 .\"
167 pages.
168 .
169 .
170 .SH "PCRE API OVERVIEW"
171 .rs
172 .sp
173 PCRE has its own native API, which is described in this document. There are
174 also some wrapper functions (for the 8-bit library only) that correspond to the
175 POSIX regular expression API, but they do not give access to all the
176 functionality. They are described in the
177 .\" HREF
178 \fBpcreposix\fP
179 .\"
180 documentation. Both of these APIs define a set of C function calls. A C++
181 wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
182 documented in the
183 .\" HREF
184 \fBpcrecpp\fP
185 .\"
186 page.
187 .P
188 The native API C function prototypes are defined in the header file
189 \fBpcre.h\fP, and on Unix-like systems the (8-bit) library itself is called
190 \fBlibpcre\fP. It can normally be accessed by adding \fB-lpcre\fP to the
191 command for linking an application that uses PCRE. The header file defines the
192 macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
193 for the library. Applications can use these to include support for different
194 releases of PCRE.
195 .P
196 In a Windows environment, if you want to statically link an application program
197 against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
198 including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
199 \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
200 \fB__declspec(dllimport)\fP, with unwanted results.
201 .P
202 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
203 and \fBpcre_exec()\fP are used for compiling and matching regular expressions
204 in a Perl-compatible manner. A sample program that demonstrates the simplest
205 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
206 source distribution. A listing of this program is given in the
207 .\" HREF
208 \fBpcredemo\fP
209 .\"
210 documentation, and the
211 .\" HREF
212 \fBpcresample\fP
213 .\"
214 documentation describes how to compile and run it.
215 .P
216 Just-in-time compiler support is an optional feature of PCRE that can be built
217 in appropriate hardware environments. It greatly speeds up the matching
218 performance of many patterns. Simple programs can easily request that it be
219 used if available, by setting an option that is ignored when it is not
220 relevant. More complicated programs might need to make use of the functions
221 \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
222 \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
223 .P
224 From release 8.32 there is also a direct interface for JIT execution, which
225 gives improved performance. The JIT-specific functions are discussed in the
226 .\" HREF
227 \fBpcrejit\fP
228 .\"
229 documentation.
230 .P
231 A second matching function, \fBpcre_dfa_exec()\fP, which is not
232 Perl-compatible, is also provided. This uses a different algorithm for the
233 matching. The alternative algorithm finds all possible matches (at a given
234 point in the subject), and scans the subject just once (unless there are
235 lookbehind assertions). However, this algorithm does not return captured
236 substrings. A description of the two matching algorithms and their advantages
237 and disadvantages is given in the
238 .\" HREF
239 \fBpcrematching\fP
240 .\"
241 documentation.
242 .P
243 In addition to the main compiling and matching functions, there are convenience
244 functions for extracting captured substrings from a subject string that is
245 matched by \fBpcre_exec()\fP. They are:
246 .sp
247 \fBpcre_copy_substring()\fP
248 \fBpcre_copy_named_substring()\fP
249 \fBpcre_get_substring()\fP
250 \fBpcre_get_named_substring()\fP
251 \fBpcre_get_substring_list()\fP
252 \fBpcre_get_stringnumber()\fP
253 \fBpcre_get_stringtable_entries()\fP
254 .sp
255 \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
256 provided, to free the memory used for extracted strings.
257 .P
258 The function \fBpcre_maketables()\fP is used to build a set of character tables
259 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
260 or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
261 specialist use. Most commonly, no special tables are passed, in which case
262 internal tables that are generated when PCRE is built are used.
263 .P
264 The function \fBpcre_fullinfo()\fP is used to find out information about a
265 compiled pattern. The function \fBpcre_version()\fP returns a pointer to a
266 string containing the version of PCRE and its date of release.
267 .P
268 The function \fBpcre_refcount()\fP maintains a reference count in a data block
269 containing a compiled pattern. This is provided for the benefit of
270 object-oriented applications.
271 .P
272 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
273 the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
274 respectively. PCRE calls the memory management functions via these variables,
275 so a calling program can replace them if it wishes to intercept the calls. This
276 should be done before calling any PCRE functions.
277 .P
278 The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
279 indirections to memory management functions. These special functions are used
280 only when PCRE is compiled to use the heap for remembering data, instead of
281 recursive function calls, when running the \fBpcre_exec()\fP function. See the
282 .\" HREF
283 \fBpcrebuild\fP
284 .\"
285 documentation for details of how to do this. It is a non-standard way of
286 building PCRE, for use in environments that have limited stacks. Because of the
287 greater use of memory management, it runs more slowly. Separate functions are
288 provided so that special-purpose external code can be used for this case. When
289 used, these functions are always called in a stack-like manner (last obtained,
290 first freed), and always for memory blocks of the same size. There is a
291 discussion about PCRE's stack usage in the
292 .\" HREF
293 \fBpcrestack\fP
294 .\"
295 documentation.
296 .P
297 The global variable \fBpcre_callout\fP initially contains NULL. It can be set
298 by the caller to a "callout" function, which PCRE will then call at specified
299 points during a matching operation. Details are given in the
300 .\" HREF
301 \fBpcrecallout\fP
302 .\"
303 documentation.
304 .
305 .
306 .\" HTML <a name="newlines"></a>
307 .SH NEWLINES
308 .rs
309 .sp
310 PCRE supports five different conventions for indicating line breaks in
311 strings: a single CR (carriage return) character, a single LF (linefeed)
312 character, the two-character sequence CRLF, any of the three preceding, or any
313 Unicode newline sequence. The Unicode newline sequences are the three just
314 mentioned, plus the single characters VT (vertical tab, U+000B), FF (form feed,
315 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
316 (paragraph separator, U+2029).
317 .P
318 Each of the first three conventions is used by at least one operating system as
319 its standard newline sequence. When PCRE is built, a default can be specified.
320 The default default is LF, which is the Unix standard. When PCRE is run, the
321 default can be overridden, either when a pattern is compiled, or when it is
322 matched.
323 .P
324 At compile time, the newline convention can be specified by the \fIoptions\fP
325 argument of \fBpcre_compile()\fP, or it can be specified by special text at the
326 start of the pattern itself; this overrides any other settings. See the
327 .\" HREF
328 \fBpcrepattern\fP
329 .\"
330 page for details of the special character sequences.
331 .P
332 In the PCRE documentation the word "newline" is used to mean "the character or
333 pair of characters that indicate a line break". The choice of newline
334 convention affects the handling of the dot, circumflex, and dollar
335 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
336 recognized line ending sequence, the match position advancement for a
337 non-anchored pattern. There is more detail about this in the
338 .\" HTML <a href="#execoptions">
339 .\" </a>
340 section on \fBpcre_exec()\fP options
341 .\"
342 below.
343 .P
344 The choice of newline convention does not affect the interpretation of
345 the \en or \er escape sequences, nor does it affect what \eR matches, which is
346 controlled in a similar way, but by separate options.
347 .
348 .
349 .SH MULTITHREADING
350 .rs
351 .sp
352 The PCRE functions can be used in multi-threading applications, with the
353 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
354 \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
355 callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
356 .P
357 The compiled form of a regular expression is not altered during matching, so
358 the same compiled pattern can safely be used by several threads at once.
359 .P
360 If the just-in-time optimization feature is being used, it needs separate
361 memory stack areas for each thread. See the
362 .\" HREF
363 \fBpcrejit\fP
364 .\"
365 documentation for more details.
366 .
367 .
368 .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
369 .rs
370 .sp
371 The compiled form of a regular expression can be saved and re-used at a later
372 time, possibly by a different program, and even on a host other than the one on
373 which it was compiled. Details are given in the
374 .\" HREF
375 \fBpcreprecompile\fP
376 .\"
377 documentation, which includes a description of the
378 \fBpcre_pattern_to_host_byte_order()\fP function. However, compiling a regular
379 expression with one version of PCRE for use with a different version is not
380 guaranteed to work and may cause crashes.
381 .
382 .
383 .SH "CHECKING BUILD-TIME OPTIONS"
384 .rs
385 .sp
386 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
387 .PP
388 The function \fBpcre_config()\fP makes it possible for a PCRE client to
389 discover which optional features have been compiled into the PCRE library. The
390 .\" HREF
391 \fBpcrebuild\fP
392 .\"
393 documentation has more details about these optional features.
394 .P
395 The first argument for \fBpcre_config()\fP is an integer, specifying which
396 information is required; the second argument is a pointer to a variable into
397 which the information is placed. The returned value is zero on success, or the
398 negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
399 not recognized. The following information is available:
400 .sp
401 PCRE_CONFIG_UTF8
402 .sp
403 The output is an integer that is set to one if UTF-8 support is available;
404 otherwise it is set to zero. This value should normally be given to the 8-bit
405 version of this function, \fBpcre_config()\fP. If it is given to the 16-bit
406 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
407 .sp
408 PCRE_CONFIG_UTF16
409 .sp
410 The output is an integer that is set to one if UTF-16 support is available;
411 otherwise it is set to zero. This value should normally be given to the 16-bit
412 version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
413 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
414 .sp
415 PCRE_CONFIG_UTF32
416 .sp
417 The output is an integer that is set to one if UTF-32 support is available;
418 otherwise it is set to zero. This value should normally be given to the 32-bit
419 version of this function, \fBpcre32_config()\fP. If it is given to the 8-bit
420 or 16-bit version of this function, the result is PCRE_ERROR_BADOPTION.
421 .sp
422 PCRE_CONFIG_UNICODE_PROPERTIES
423 .sp
424 The output is an integer that is set to one if support for Unicode character
425 properties is available; otherwise it is set to zero.
426 .sp
427 PCRE_CONFIG_JIT
428 .sp
429 The output is an integer that is set to one if support for just-in-time
430 compiling is available; otherwise it is set to zero.
431 .sp
432 PCRE_CONFIG_JITTARGET
433 .sp
434 The output is a pointer to a zero-terminated "const char *" string. If JIT
435 support is available, the string contains the name of the architecture for
436 which the JIT compiler is configured, for example "x86 32bit (little endian +
437 unaligned)". If JIT support is not available, the result is NULL.
438 .sp
439 PCRE_CONFIG_NEWLINE
440 .sp
441 The output is an integer whose value specifies the default character sequence
442 that is recognized as meaning "newline". The values that are supported in
443 ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for
444 ANYCRLF, and -1 for ANY. In EBCDIC environments, CR, ANYCRLF, and ANY yield the
445 same values. However, the value for LF is normally 21, though some EBCDIC
446 environments use 37. The corresponding values for CRLF are 3349 and 3365. The
447 default should normally correspond to the standard sequence for your operating
448 system.
449 .sp
450 PCRE_CONFIG_BSR
451 .sp
452 The output is an integer whose value indicates what character sequences the \eR
453 escape sequence matches by default. A value of 0 means that \eR matches any
454 Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
455 or CRLF. The default can be overridden when a pattern is compiled or matched.
456 .sp
457 PCRE_CONFIG_LINK_SIZE
458 .sp
459 The output is an integer that contains the number of bytes used for internal
460 linkage in compiled regular expressions. For the 8-bit library, the value can
461 be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
462 a number of bytes. For the 32-bit library, the value is either 2 or 4 and is
463 still a number of bytes. The default value of 2 is sufficient for all but the
464 most massive patterns, since it allows the compiled pattern to be up to 64K in
465 size. Larger values allow larger regular expressions to be compiled, at the
466 expense of slower matching.
467 .sp
468 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
469 .sp
470 The output is an integer that contains the threshold above which the POSIX
471 interface uses \fBmalloc()\fP for output vectors. Further details are given in
472 the
473 .\" HREF
474 \fBpcreposix\fP
475 .\"
476 documentation.
477 .sp
478 PCRE_CONFIG_MATCH_LIMIT
479 .sp
480 The output is a long integer that gives the default limit for the number of
481 internal matching function calls in a \fBpcre_exec()\fP execution. Further
482 details are given with \fBpcre_exec()\fP below.
483 .sp
484 PCRE_CONFIG_MATCH_LIMIT_RECURSION
485 .sp
486 The output is a long integer that gives the default limit for the depth of
487 recursion when calling the internal matching function in a \fBpcre_exec()\fP
488 execution. Further details are given with \fBpcre_exec()\fP below.
489 .sp
490 PCRE_CONFIG_STACKRECURSE
491 .sp
492 The output is an integer that is set to one if internal recursion when running
493 \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
494 to remember their state. This is the usual way that PCRE is compiled. The
495 output is zero if PCRE was compiled to use blocks of data on the heap instead
496 of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
497 \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
498 avoiding the use of the stack.
499 .
500 .
501 .SH "COMPILING A PATTERN"
502 .rs
503 .sp
504 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
505 .ti +5n
506 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
507 .ti +5n
508 .B const unsigned char *\fItableptr\fP);
509 .sp
510 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
511 .ti +5n
512 .B int *\fIerrorcodeptr\fP,
513 .ti +5n
514 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
515 .ti +5n
516 .B const unsigned char *\fItableptr\fP);
517 .P
518 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
519 called to compile a pattern into an internal form. The only difference between
520 the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
521 \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
522 too much repetition, we refer just to \fBpcre_compile()\fP below, but the
523 information applies equally to \fBpcre_compile2()\fP.
524 .P
525 The pattern is a C string terminated by a binary zero, and is passed in the
526 \fIpattern\fP argument. A pointer to a single block of memory that is obtained
527 via \fBpcre_malloc\fP is returned. This contains the compiled code and related
528 data. The \fBpcre\fP type is defined for the returned block; this is a typedef
529 for a structure whose contents are not externally defined. It is up to the
530 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
531 .P
532 Although the compiled code of a PCRE regex is relocatable, that is, it does not
533 depend on memory location, the complete \fBpcre\fP data block is not
534 fully relocatable, because it may contain a copy of the \fItableptr\fP
535 argument, which is an address (see below).
536 .P
537 The \fIoptions\fP argument contains various bit settings that affect the
538 compilation. It should be zero if no options are required. The available
539 options are described below. Some of them (in particular, those that are
540 compatible with Perl, but some others as well) can also be set and unset from
541 within the pattern (see the detailed description in the
542 .\" HREF
543 \fBpcrepattern\fP
544 .\"
545 documentation). For those options that can be different in different parts of
546 the pattern, the contents of the \fIoptions\fP argument specifies their
547 settings at the start of compilation and execution. The PCRE_ANCHORED,
548 PCRE_BSR_\fIxxx\fP, PCRE_NEWLINE_\fIxxx\fP, PCRE_NO_UTF8_CHECK, and
549 PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
550 compile time.
551 .P
552 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
553 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
554 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
555 error message. This is a static string that is part of the library. You must
556 not try to free it. Normally, the offset from the start of the pattern to the
557 byte that was being processed when the error was discovered is placed in the
558 variable pointed to by \fIerroffset\fP, which must not be NULL (if it is, an
559 immediate error is given). However, for an invalid UTF-8 string, the offset is
560 that of the first byte of the failing character.
561 .P
562 Some errors are not detected until the whole pattern has been scanned; in these
563 cases, the offset passed back is the length of the pattern. Note that the
564 offset is in bytes, not characters, even in UTF-8 mode. It may sometimes point
565 into the middle of a UTF-8 character.
566 .P
567 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
568 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
569 returned via this argument in the event of an error. This is in addition to the
570 textual error message. Error codes and messages are listed below.
571 .P
572 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
573 character tables that are built when PCRE is compiled, using the default C
574 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
575 call to \fBpcre_maketables()\fP. This value is stored with the compiled
576 pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
577 passed to it. For more discussion, see the section on locale support below.
578 .P
579 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
580 .sp
581 pcre *re;
582 const char *error;
583 int erroffset;
584 re = pcre_compile(
585 "^A.*Z", /* the pattern */
586 0, /* default options */
587 &error, /* for error message */
588 &erroffset, /* for error offset */
589 NULL); /* use default character tables */
590 .sp
591 The following names for option bits are defined in the \fBpcre.h\fP header
592 file:
593 .sp
594 PCRE_ANCHORED
595 .sp
596 If this bit is set, the pattern is forced to be "anchored", that is, it is
597 constrained to match only at the first matching point in the string that is
598 being searched (the "subject string"). This effect can also be achieved by
599 appropriate constructs in the pattern itself, which is the only way to do it in
600 Perl.
601 .sp
602 PCRE_AUTO_CALLOUT
603 .sp
604 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
605 all with number 255, before each pattern item. For discussion of the callout
606 facility, see the
607 .\" HREF
608 \fBpcrecallout\fP
609 .\"
610 documentation.
611 .sp
612 PCRE_BSR_ANYCRLF
613 PCRE_BSR_UNICODE
614 .sp
615 These options (which are mutually exclusive) control what the \eR escape
616 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
617 match any Unicode newline sequence. The default is specified when PCRE is
618 built. It can be overridden from within the pattern, or by setting an option
619 when a compiled pattern is matched.
620 .sp
621 PCRE_CASELESS
622 .sp
623 If this bit is set, letters in the pattern match both upper and lower case
624 letters. It is equivalent to Perl's /i option, and it can be changed within a
625 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
626 concept of case for characters whose values are less than 128, so caseless
627 matching is always possible. For characters with higher values, the concept of
628 case is supported if PCRE is compiled with Unicode property support, but not
629 otherwise. If you want to use caseless matching for characters 128 and above,
630 you must ensure that PCRE is compiled with Unicode property support as well as
631 with UTF-8 support.
632 .sp
633 PCRE_DOLLAR_ENDONLY
634 .sp
635 If this bit is set, a dollar metacharacter in the pattern matches only at the
636 end of the subject string. Without this option, a dollar also matches
637 immediately before a newline at the end of the string (but not before any other
638 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
639 There is no equivalent to this option in Perl, and no way to set it within a
640 pattern.
641 .sp
642 PCRE_DOTALL
643 .sp
644 If this bit is set, a dot metacharacter in the pattern matches a character of
645 any value, including one that indicates a newline. However, it only ever
646 matches one character, even if newlines are coded as CRLF. Without this option,
647 a dot does not match when the current position is at a newline. This option is
648 equivalent to Perl's /s option, and it can be changed within a pattern by a
649 (?s) option setting. A negative class such as [^a] always matches newline
650 characters, independent of the setting of this option.
651 .sp
652 PCRE_DUPNAMES
653 .sp
654 If this bit is set, names used to identify capturing subpatterns need not be
655 unique. This can be helpful for certain types of pattern when it is known that
656 only one instance of the named subpattern can ever be matched. There are more
657 details of named subpatterns below; see also the
658 .\" HREF
659 \fBpcrepattern\fP
660 .\"
661 documentation.
662 .sp
663 PCRE_EXTENDED
664 .sp
665 If this bit is set, white space data characters in the pattern are totally
666 ignored except when escaped or inside a character class. White space does not
667 include the VT character (code 11). In addition, characters between an
668 unescaped # outside a character class and the next newline, inclusive, are also
669 ignored. This is equivalent to Perl's /x option, and it can be changed within a
670 pattern by a (?x) option setting.
671 .P
672 Which characters are interpreted as newlines is controlled by the options
673 passed to \fBpcre_compile()\fP or by a special sequence at the start of the
674 pattern, as described in the section entitled
675 .\" HTML <a href="pcrepattern.html#newlines">
676 .\" </a>
677 "Newline conventions"
678 .\"
679 in the \fBpcrepattern\fP documentation. Note that the end of this type of
680 comment is a literal newline sequence in the pattern; escape sequences that
681 happen to represent a newline do not count.
682 .P
683 This option makes it possible to include comments inside complicated patterns.
684 Note, however, that this applies only to data characters. White space characters
685 may never appear within special character sequences in a pattern, for example
686 within the sequence (?( that introduces a conditional subpattern.
687 .sp
688 PCRE_EXTRA
689 .sp
690 This option was invented in order to turn on additional functionality of PCRE
691 that is incompatible with Perl, but it is currently of very little use. When
692 set, any backslash in a pattern that is followed by a letter that has no
693 special meaning causes an error, thus reserving these combinations for future
694 expansion. By default, as in Perl, a backslash followed by a letter with no
695 special meaning is treated as a literal. (Perl can, however, be persuaded to
696 give an error for this, by running it with the -w option.) There are at present
697 no other features controlled by this option. It can also be set by a (?X)
698 option setting within a pattern.
699 .sp
700 PCRE_FIRSTLINE
701 .sp
702 If this option is set, an unanchored pattern is required to match before or at
703 the first newline in the subject string, though the matched text may continue
704 over the newline.
705 .sp
706 PCRE_JAVASCRIPT_COMPAT
707 .sp
708 If this option is set, PCRE's behaviour is changed in some ways so that it is
709 compatible with JavaScript rather than Perl. The changes are as follows:
710 .P
711 (1) A lone closing square bracket in a pattern causes a compile-time error,
712 because this is illegal in JavaScript (by default it is treated as a data
713 character). Thus, the pattern AB]CD becomes illegal when this option is set.
714 .P
715 (2) At run time, a back reference to an unset subpattern group matches an empty
716 string (by default this causes the current matching alternative to fail). A
717 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
718 an "a" in the subject), whereas it fails by default, for Perl compatibility.
719 .P
720 (3) \eU matches an upper case "U" character; by default \eU causes a compile
721 time error (Perl uses \eU to upper case subsequent characters).
722 .P
723 (4) \eu matches a lower case "u" character unless it is followed by four
724 hexadecimal digits, in which case the hexadecimal number defines the code point
725 to match. By default, \eu causes a compile time error (Perl uses it to upper
726 case the following character).
727 .P
728 (5) \ex matches a lower case "x" character unless it is followed by two
729 hexadecimal digits, in which case the hexadecimal number defines the code point
730 to match. By default, as in Perl, a hexadecimal number is always expected after
731 \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
732 binary zero character followed by z).
733 .sp
734 PCRE_MULTILINE
735 .sp
736 By default, PCRE treats the subject string as consisting of a single line of
737 characters (even if it actually contains newlines). The "start of line"
738 metacharacter (^) matches only at the start of the string, while the "end of
739 line" metacharacter ($) matches only at the end of the string, or before a
740 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
741 Perl.
742 .P
743 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
744 match immediately following or immediately before internal newlines in the
745 subject string, respectively, as well as at the very start and end. This is
746 equivalent to Perl's /m option, and it can be changed within a pattern by a
747 (?m) option setting. If there are no newlines in a subject string, or no
748 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
749 .sp
750 PCRE_NEWLINE_CR
751 PCRE_NEWLINE_LF
752 PCRE_NEWLINE_CRLF
753 PCRE_NEWLINE_ANYCRLF
754 PCRE_NEWLINE_ANY
755 .sp
756 These options override the default newline definition that was chosen when PCRE
757 was built. Setting the first or the second specifies that a newline is
758 indicated by a single character (CR or LF, respectively). Setting
759 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
760 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
761 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
762 that any Unicode newline sequence should be recognized.
763 .P
764 In an ASCII/Unicode environment, the Unicode newline sequences are the three
765 just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
766 feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
767 (paragraph separator, U+2029). For the 8-bit library, the last two are
768 recognized only in UTF-8 mode.
769 .P
770 When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
771 CR is 0x0d, the same as ASCII. However, the character code for LF is normally
772 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
773 not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
774 less than 256. For more details, see the
775 .\" HREF
776 \fBpcrebuild\fP
777 .\"
778 documentation.
779 .P
780 The newline setting in the options word uses three bits that are treated
781 as a number, giving eight possibilities. Currently only six are used (default
782 plus the five values above). This means that if you set more than one newline
783 option, the combination may or may not be sensible. For example,
784 PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
785 other combinations may yield unused numbers and cause an error.
786 .P
787 The only time that a line break in a pattern is specially recognized when
788 compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
789 and so are ignored in this mode. Also, an unescaped # outside a character class
790 indicates a comment that lasts until after the next line break sequence. In
791 other circumstances, line break sequences in patterns are treated as literal
792 data.
793 .P
794 The newline option that is set at compile time becomes the default that is used
795 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
796 .sp
797 PCRE_NO_AUTO_CAPTURE
798 .sp
799 If this option is set, it disables the use of numbered capturing parentheses in
800 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
801 were followed by ?: but named parentheses can still be used for capturing (and
802 they acquire numbers in the usual way). There is no equivalent of this option
803 in Perl.
804 .sp
805 NO_START_OPTIMIZE
806 .sp
807 This is an option that acts at matching time; that is, it is really an option
808 for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
809 it is remembered with the compiled pattern and assumed at matching time. For
810 details see the discussion of PCRE_NO_START_OPTIMIZE
811 .\" HTML <a href="#execoptions">
812 .\" </a>
813 below.
814 .\"
815 .sp
816 PCRE_UCP
817 .sp
818 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
819 \ew, and some of the POSIX character classes. By default, only ASCII characters
820 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
821 classify characters. More details are given in the section on
822 .\" HTML <a href="pcre.html#genericchartypes">
823 .\" </a>
824 generic character types
825 .\"
826 in the
827 .\" HREF
828 \fBpcrepattern\fP
829 .\"
830 page. If you set PCRE_UCP, matching one of the items it affects takes much
831 longer. The option is available only if PCRE has been compiled with Unicode
832 property support.
833 .sp
834 PCRE_UNGREEDY
835 .sp
836 This option inverts the "greediness" of the quantifiers so that they are not
837 greedy by default, but become greedy if followed by "?". It is not compatible
838 with Perl. It can also be set by a (?U) option setting within the pattern.
839 .sp
840 PCRE_UTF8
841 .sp
842 This option causes PCRE to regard both the pattern and the subject as strings
843 of UTF-8 characters instead of single-byte strings. However, it is available
844 only when PCRE is built to include UTF support. If not, the use of this option
845 provokes an error. Details of how this option changes the behaviour of PCRE are
846 given in the
847 .\" HREF
848 \fBpcreunicode\fP
849 .\"
850 page.
851 .sp
852 PCRE_NO_UTF8_CHECK
853 .sp
854 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
855 automatically checked. There is a discussion about the
856 .\" HTML <a href="pcreunicode.html#utf8strings">
857 .\" </a>
858 validity of UTF-8 strings
859 .\"
860 in the
861 .\" HREF
862 \fBpcreunicode\fP
863 .\"
864 page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
865 error. If you already know that your pattern is valid, and you want to skip
866 this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
867 When it is set, the effect of passing an invalid UTF-8 string as a pattern is
868 undefined. It may cause your program to crash. Note that this option can also
869 be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress the
870 validity checking of subject strings only. If the same string is being matched
871 many times, the option can be safely set for the second and subsequent
872 matchings to improve performance.
873 .
874 .
875 .SH "COMPILATION ERROR CODES"
876 .rs
877 .sp
878 The following table lists the error codes than may be returned by
879 \fBpcre_compile2()\fP, along with the error messages that may be returned by
880 both compiling functions. Note that error messages are always 8-bit ASCII
881 strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
882 have fallen out of use. To avoid confusion, they have not been re-used.
883 .sp
884 0 no error
885 1 \e at end of pattern
886 2 \ec at end of pattern
887 3 unrecognized character follows \e
888 4 numbers out of order in {} quantifier
889 5 number too big in {} quantifier
890 6 missing terminating ] for character class
891 7 invalid escape sequence in character class
892 8 range out of order in character class
893 9 nothing to repeat
894 10 [this code is not in use]
895 11 internal error: unexpected repeat
896 12 unrecognized character after (? or (?-
897 13 POSIX named classes are supported only within a class
898 14 missing )
899 15 reference to non-existent subpattern
900 16 erroffset passed as NULL
901 17 unknown option bit(s) set
902 18 missing ) after comment
903 19 [this code is not in use]
904 20 regular expression is too large
905 21 failed to get memory
906 22 unmatched parentheses
907 23 internal error: code overflow
908 24 unrecognized character after (?<
909 25 lookbehind assertion is not fixed length
910 26 malformed number or name after (?(
911 27 conditional group contains more than two branches
912 28 assertion expected after (?(
913 29 (?R or (?[+-]digits must be followed by )
914 30 unknown POSIX class name
915 31 POSIX collating elements are not supported
916 32 this version of PCRE is compiled without UTF support
917 33 [this code is not in use]
918 34 character value in \ex{...} sequence is too large
919 35 invalid condition (?(0)
920 36 \eC not allowed in lookbehind assertion
921 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
922 38 number after (?C is > 255
923 39 closing ) for (?C expected
924 40 recursive call could loop indefinitely
925 41 unrecognized character after (?P
926 42 syntax error in subpattern name (missing terminator)
927 43 two named subpatterns have the same name
928 44 invalid UTF-8 string (specifically UTF-8)
929 45 support for \eP, \ep, and \eX has not been compiled
930 46 malformed \eP or \ep sequence
931 47 unknown property name after \eP or \ep
932 48 subpattern name is too long (maximum 32 characters)
933 49 too many named subpatterns (maximum 10000)
934 50 [this code is not in use]
935 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
936 52 internal error: overran compiling workspace
937 53 internal error: previously-checked referenced subpattern
938 not found
939 54 DEFINE group contains more than one branch
940 55 repeating a DEFINE group is not allowed
941 56 inconsistent NEWLINE options
942 57 \eg is not followed by a braced, angle-bracketed, or quoted
943 name/number or by a plain number
944 58 a numbered reference must not be zero
945 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
946 60 (*VERB) not recognized
947 61 number is too big
948 62 subpattern name expected
949 63 digit expected after (?+
950 64 ] is an invalid data character in JavaScript compatibility mode
951 65 different names for subpatterns of the same number are
952 not allowed
953 66 (*MARK) must have an argument
954 67 this version of PCRE is not compiled with Unicode property
955 support
956 68 \ec must be followed by an ASCII character
957 69 \ek is not followed by a braced, angle-bracketed, or quoted name
958 70 internal error: unknown opcode in find_fixedlength()
959 71 \eN is not supported in a class
960 72 too many forward references
961 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
962 74 invalid UTF-16 string (specifically UTF-16)
963 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
964 76 character value in \eu.... sequence is too large
965 77 invalid UTF-32 string (specifically UTF-32)
966 .sp
967 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
968 be used if the limits were changed when PCRE was built.
969 .
970 .
971 .\" HTML <a name="studyingapattern"></a>
972 .SH "STUDYING A PATTERN"
973 .rs
974 .sp
975 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
976 .ti +5n
977 .B const char **\fIerrptr\fP);
978 .PP
979 If a compiled pattern is going to be used several times, it is worth spending
980 more time analyzing it in order to speed up the time taken for matching. The
981 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
982 argument. If studying the pattern produces additional information that will
983 help speed up matching, \fBpcre_study()\fP returns a pointer to a
984 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
985 results of the study.
986 .P
987 The returned value from \fBpcre_study()\fP can be passed directly to
988 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
989 also contains other fields that can be set by the caller before the block is
990 passed; these are described
991 .\" HTML <a href="#extradata">
992 .\" </a>
993 below
994 .\"
995 in the section on matching a pattern.
996 .P
997 If studying the pattern does not produce any useful information,
998 \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
999 calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
1000 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
1001 if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
1002 returns a \fBpcre_extra\fP block even if studying did not find any additional
1003 information. It may still return NULL, however, if an error occurs in
1004 \fBpcre_study()\fP.
1005 .P
1006 The second argument of \fBpcre_study()\fP contains option bits. There are three
1007 further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1008 .sp
1009 PCRE_STUDY_JIT_COMPILE
1010 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1011 PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1012 .sp
1013 If any of these are set, and the just-in-time compiler is available, the
1014 pattern is further compiled into machine code that executes much faster than
1015 the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
1016 compiler is not available, these options are ignored. All undefined bits in the
1017 \fIoptions\fP argument must be zero.
1018 .P
1019 JIT compilation is a heavyweight optimization. It can take some time for
1020 patterns to be analyzed, and for one-off matches and simple patterns the
1021 benefit of faster execution might be offset by a much slower study time.
1022 Not all patterns can be optimized by the JIT compiler. For those that cannot be
1023 handled, matching automatically falls back to the \fBpcre_exec()\fP
1024 interpreter. For more details, see the
1025 .\" HREF
1026 \fBpcrejit\fP
1027 .\"
1028 documentation.
1029 .P
1030 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
1031 studying succeeds (even if no data is returned), the variable it points to is
1032 set to NULL. Otherwise it is set to point to a textual error message. This is a
1033 static string that is part of the library. You must not try to free it. You
1034 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
1035 sure that it has run successfully.
1036 .P
1037 When you are finished with a pattern, you can free the memory used for the
1038 study data by calling \fBpcre_free_study()\fP. This function was added to the
1039 API for release 8.20. For earlier versions, the memory could be freed with
1040 \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1041 where JIT optimization is not used, but it is advisable to change to the new
1042 function when convenient.
1043 .P
1044 This is a typical way in which \fBpcre_study\fP() is used (except that in a
1045 real application there should be tests for errors):
1046 .sp
1047 int rc;
1048 pcre *re;
1049 pcre_extra *sd;
1050 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1051 sd = pcre_study(
1052 re, /* result of pcre_compile() */
1053 0, /* no options */
1054 &error); /* set to NULL or points to a message */
1055 rc = pcre_exec( /* see below for details of pcre_exec() options */
1056 re, sd, "subject", 7, 0, 0, ovector, 30);
1057 ...
1058 pcre_free_study(sd);
1059 pcre_free(re);
1060 .sp
1061 Studying a pattern does two things: first, a lower bound for the length of
1062 subject string that is needed to match the pattern is computed. This does not
1063 mean that there are any strings of that length that match, but it does
1064 guarantee that no shorter strings match. The value is used to avoid wasting
1065 time by trying to match strings that are shorter than the lower bound. You can
1066 find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
1067 .P
1068 Studying a pattern is also useful for non-anchored patterns that do not have a
1069 single fixed starting character. A bitmap of possible starting bytes is
1070 created. This speeds up finding a position in the subject at which to start
1071 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1072 In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1073 .P
1074 These two optimizations apply to both \fBpcre_exec()\fP and
1075 \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1076 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option
1077 when calling \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP, but if this is done,
1078 JIT execution is also disabled. You might want to do this if your pattern
1079 contains callouts or (*MARK) and you want to make use of these facilities in
1080 cases where matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
1081 .\" HTML <a href="#execoptions">
1082 .\" </a>
1083 below.
1084 .\"
1085 .
1086 .
1087 .\" HTML <a name="localesupport"></a>
1088 .SH "LOCALE SUPPORT"
1089 .rs
1090 .sp
1091 PCRE handles caseless matching, and determines whether characters are letters,
1092 digits, or whatever, by reference to a set of tables, indexed by character
1093 value. When running in UTF-8 mode, this applies only to characters
1094 with codes less than 128. By default, higher-valued codes never match escapes
1095 such as \ew or \ed, but they can be tested with \ep if PCRE is built with
1096 Unicode character property support. Alternatively, the PCRE_UCP option can be
1097 set at compile time; this causes \ew and friends to use Unicode property
1098 support instead of built-in tables. The use of locales with Unicode is
1099 discouraged. If you are handling characters with codes greater than 128, you
1100 should either use UTF-8 and Unicode, or use locales, but not try to mix the
1101 two.
1102 .P
1103 PCRE contains an internal set of tables that are used when the final argument
1104 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1105 Normally, the internal tables recognize only ASCII characters. However, when
1106 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1107 default "C" locale of the local system, which may cause them to be different.
1108 .P
1109 The internal tables can always be overridden by tables supplied by the
1110 application that calls PCRE. These may be created in a different locale from
1111 the default. As more and more applications change to using Unicode, the need
1112 for this locale support is expected to die away.
1113 .P
1114 External tables are built by calling the \fBpcre_maketables()\fP function,
1115 which has no arguments, in the relevant locale. The result can then be passed
1116 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
1117 example, to build and use tables that are appropriate for the French locale
1118 (where accented characters with values greater than 128 are treated as letters),
1119 the following code could be used:
1120 .sp
1121 setlocale(LC_CTYPE, "fr_FR");
1122 tables = pcre_maketables();
1123 re = pcre_compile(..., tables);
1124 .sp
1125 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1126 are using Windows, the name for the French locale is "french".
1127 .P
1128 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1129 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1130 that the memory containing the tables remains available for as long as it is
1131 needed.
1132 .P
1133 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1134 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1135 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
1136 pattern, compilation, studying and matching all happen in the same locale, but
1137 different patterns can be compiled in different locales.
1138 .P
1139 It is possible to pass a table pointer or NULL (indicating the use of the
1140 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
1141 this facility could be used to match a pattern in a different locale from the
1142 one in which it was compiled. Passing table pointers at run time is discussed
1143 below in the section on matching a pattern.
1144 .
1145 .
1146 .\" HTML <a name="infoaboutpattern"></a>
1147 .SH "INFORMATION ABOUT A PATTERN"
1148 .rs
1149 .sp
1150 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1151 .ti +5n
1152 .B int \fIwhat\fP, void *\fIwhere\fP);
1153 .PP
1154 The \fBpcre_fullinfo()\fP function returns information about a compiled
1155 pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1156 library at version 8.30, after more than 10 years of obsolescence.
1157 .P
1158 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1159 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1160 the pattern was not studied. The third argument specifies which piece of
1161 information is required, and the fourth argument is a pointer to a variable
1162 to receive the data. The yield of the function is zero for success, or one of
1163 the following negative numbers:
1164 .sp
1165 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1166 the argument \fIwhere\fP was NULL
1167 PCRE_ERROR_BADMAGIC the "magic number" was not found
1168 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1169 endianness
1170 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1171 .sp
1172 The "magic number" is placed at the start of each compiled pattern as an simple
1173 check against passing an arbitrary memory pointer. The endianness error can
1174 occur if a compiled pattern is saved and reloaded on a different host. Here is
1175 a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1176 pattern:
1177 .sp
1178 int rc;
1179 size_t length;
1180 rc = pcre_fullinfo(
1181 re, /* result of pcre_compile() */
1182 sd, /* result of pcre_study(), or NULL */
1183 PCRE_INFO_SIZE, /* what is required */
1184 &length); /* where to put the data */
1185 .sp
1186 The possible values for the third argument are defined in \fBpcre.h\fP, and are
1187 as follows:
1188 .sp
1189 PCRE_INFO_BACKREFMAX
1190 .sp
1191 Return the number of the highest back reference in the pattern. The fourth
1192 argument should point to an \fBint\fP variable. Zero is returned if there are
1193 no back references.
1194 .sp
1195 PCRE_INFO_CAPTURECOUNT
1196 .sp
1197 Return the number of capturing subpatterns in the pattern. The fourth argument
1198 should point to an \fBint\fP variable.
1199 .sp
1200 PCRE_INFO_DEFAULT_TABLES
1201 .sp
1202 Return a pointer to the internal default character tables within PCRE. The
1203 fourth argument should point to an \fBunsigned char *\fP variable. This
1204 information call is provided for internal use by the \fBpcre_study()\fP
1205 function. External callers can cause PCRE to use its internal tables by passing
1206 a NULL table pointer.
1207 .sp
1208 PCRE_INFO_FIRSTBYTE
1209 .sp
1210 Return information about the first data unit of any matched string, for a
1211 non-anchored pattern. (The name of this option refers to the 8-bit library,
1212 where data units are bytes.) The fourth argument should point to an \fBint\fP
1213 variable.
1214 .P
1215 If there is a fixed first value, for example, the letter "c" from a pattern
1216 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1217 value is always less than 256. In the 16-bit library the value can be up to
1218 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1219 .P
1220 If there is no fixed first value, and if either
1221 .sp
1222 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1223 starts with "^", or
1224 .sp
1225 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1226 (if it were set, the pattern would be anchored),
1227 .sp
1228 -1 is returned, indicating that the pattern matches only at the start of a
1229 subject string or after any newline within the string. Otherwise -2 is
1230 returned. For anchored patterns, -2 is returned.
1231 .P
1232 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1233 to return the full 32-bit range of the character, this value is deprecated;
1234 instead the PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER values
1235 should be used.
1236 .sp
1237 PCRE_INFO_FIRSTTABLE
1238 .sp
1239 If the pattern was studied, and this resulted in the construction of a 256-bit
1240 table indicating a fixed set of values for the first data unit in any matching
1241 string, a pointer to the table is returned. Otherwise NULL is returned. The
1242 fourth argument should point to an \fBunsigned char *\fP variable.
1243 .sp
1244 PCRE_INFO_HASCRORLF
1245 .sp
1246 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1247 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1248 explicit match is either a literal CR or LF character, or \er or \en.
1249 .sp
1250 PCRE_INFO_JCHANGED
1251 .sp
1252 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1253 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1254 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1255 .sp
1256 PCRE_INFO_JIT
1257 .sp
1258 Return 1 if the pattern was studied with one of the JIT options, and
1259 just-in-time compiling was successful. The fourth argument should point to an
1260 \fBint\fP variable. A return value of 0 means that JIT support is not available
1261 in this version of PCRE, or that the pattern was not studied with a JIT option,
1262 or that the JIT compiler could not handle this particular pattern. See the
1263 .\" HREF
1264 \fBpcrejit\fP
1265 .\"
1266 documentation for details of what can and cannot be handled.
1267 .sp
1268 PCRE_INFO_JITSIZE
1269 .sp
1270 If the pattern was successfully studied with a JIT option, return the size of
1271 the JIT compiled code, otherwise return zero. The fourth argument should point
1272 to a \fBsize_t\fP variable.
1273 .sp
1274 PCRE_INFO_LASTLITERAL
1275 .sp
1276 Return the value of the rightmost literal data unit that must exist in any
1277 matched string, other than at its start, if such a value has been recorded. The
1278 fourth argument should point to an \fBint\fP variable. If there is no such
1279 value, -1 is returned. For anchored patterns, a last literal value is recorded
1280 only if it follows something of variable length. For example, for the pattern
1281 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1282 is -1.
1283 .P
1284 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1285 to return the full 32-bit range of the character, this value is deprecated;
1286 instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1287 be used.
1288 .sp
1289 PCRE_INFO_MAXLOOKBEHIND
1290 .sp
1291 Return the number of characters (NB not bytes) in the longest lookbehind
1292 assertion in the pattern. Note that the simple assertions \eb and \eB require a
1293 one-character lookbehind. This information is useful when doing multi-segment
1294 matching using the partial matching facilities.
1295 .sp
1296 PCRE_INFO_MINLENGTH
1297 .sp
1298 If the pattern was studied and a minimum length for matching subject strings
1299 was computed, its value is returned. Otherwise the returned value is -1. The
1300 value is a number of characters, which in UTF-8 mode may be different from the
1301 number of bytes. The fourth argument should point to an \fBint\fP variable. A
1302 non-negative value is a lower bound to the length of any matching string. There
1303 may not be any strings of that length that do actually match, but every string
1304 that does match is at least that long.
1305 .sp
1306 PCRE_INFO_NAMECOUNT
1307 PCRE_INFO_NAMEENTRYSIZE
1308 PCRE_INFO_NAMETABLE
1309 .sp
1310 PCRE supports the use of named as well as numbered capturing parentheses. The
1311 names are just an additional way of identifying the parentheses, which still
1312 acquire numbers. Several convenience functions such as
1313 \fBpcre_get_named_substring()\fP are provided for extracting captured
1314 substrings by name. It is also possible to extract the data directly, by first
1315 converting the name to a number in order to access the correct pointers in the
1316 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1317 you need to use the name-to-number map, which is described by these three
1318 values.
1319 .P
1320 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1321 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1322 entry; both of these return an \fBint\fP value. The entry size depends on the
1323 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1324 entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1325 the first two bytes of each entry are the number of the capturing parenthesis,
1326 most significant byte first. In the 16-bit library, the pointer points to
1327 16-bit data units, the first of which contains the parenthesis number.
1328 In the 32-bit library, the pointer points to 32-bit data units, the first of
1329 which contains the parenthesis number. The rest
1330 of the entry is the corresponding name, zero terminated.
1331 .P
1332 The names are in alphabetical order. Duplicate names may appear if (?| is used
1333 to create multiple groups with the same number, as described in the
1334 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1335 .\" </a>
1336 section on duplicate subpattern numbers
1337 .\"
1338 in the
1339 .\" HREF
1340 \fBpcrepattern\fP
1341 .\"
1342 page. Duplicate names for subpatterns with different numbers are permitted only
1343 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1344 table in the order in which they were found in the pattern. In the absence of
1345 (?| this is the order of increasing number; when (?| is used this is not
1346 necessarily the case because later subpatterns may have lower numbers.
1347 .P
1348 As a simple example of the name/number table, consider the following pattern
1349 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1350 space - including newlines - is ignored):
1351 .sp
1352 .\" JOIN
1353 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1354 (?<month>\ed\ed) - (?<day>\ed\ed) )
1355 .sp
1356 There are four named subpatterns, so the table has four entries, and each entry
1357 in the table is eight bytes long. The table is as follows, with non-printing
1358 bytes shows in hexadecimal, and undefined bytes shown as ??:
1359 .sp
1360 00 01 d a t e 00 ??
1361 00 05 d a y 00 ?? ??
1362 00 04 m o n t h 00
1363 00 02 y e a r 00 ??
1364 .sp
1365 When writing code to extract data from named subpatterns using the
1366 name-to-number map, remember that the length of the entries is likely to be
1367 different for each compiled pattern.
1368 .sp
1369 PCRE_INFO_OKPARTIAL
1370 .sp
1371 Return 1 if the pattern can be used for partial matching with
1372 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1373 \fBint\fP variable. From release 8.00, this always returns 1, because the
1374 restrictions that previously applied to partial matching have been lifted. The
1375 .\" HREF
1376 \fBpcrepartial\fP
1377 .\"
1378 documentation gives details of partial matching.
1379 .sp
1380 PCRE_INFO_OPTIONS
1381 .sp
1382 Return a copy of the options with which the pattern was compiled. The fourth
1383 argument should point to an \fBunsigned long int\fP variable. These option bits
1384 are those specified in the call to \fBpcre_compile()\fP, modified by any
1385 top-level option settings at the start of the pattern itself. In other words,
1386 they are the options that will be in force when matching starts. For example,
1387 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1388 result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1389 .P
1390 A pattern is automatically anchored by PCRE if all of its top-level
1391 alternatives begin with one of the following:
1392 .sp
1393 ^ unless PCRE_MULTILINE is set
1394 \eA always
1395 \eG always
1396 .\" JOIN
1397 .* if PCRE_DOTALL is set and there are no back
1398 references to the subpattern in which .* appears
1399 .sp
1400 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1401 \fBpcre_fullinfo()\fP.
1402 .sp
1403 PCRE_INFO_SIZE
1404 .sp
1405 Return the size of the compiled pattern in bytes (for both libraries). The
1406 fourth argument should point to a \fBsize_t\fP variable. This value does not
1407 include the size of the \fBpcre\fP structure that is returned by
1408 \fBpcre_compile()\fP. The value that is passed as the argument to
1409 \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1410 place the compiled data is the value returned by this option plus the size of
1411 the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1412 does not alter the value returned by this option.
1413 .sp
1414 PCRE_INFO_STUDYSIZE
1415 .sp
1416 Return the size in bytes of the data block pointed to by the \fIstudy_data\fP
1417 field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP is NULL, or there is no
1418 study data, zero is returned. The fourth argument should point to a
1419 \fBsize_t\fP variable. The \fIstudy_data\fP field is set by \fBpcre_study()\fP
1420 to record information that will speed up matching (see the section entitled
1421 .\" HTML <a href="#studyingapattern">
1422 .\" </a>
1423 "Studying a pattern"
1424 .\"
1425 above). The format of the \fIstudy_data\fP block is private, but its length
1426 is made available via this option so that it can be saved and restored (see the
1427 .\" HREF
1428 \fBpcreprecompile\fP
1429 .\"
1430 documentation for details).
1431 .sp
1432 PCRE_INFO_FIRSTCHARACTERFLAGS
1433 .sp
1434 Return information about the first data unit of any matched string, for a
1435 non-anchored pattern. The fourth argument should point to an \fBint\fP
1436 variable.
1437 .P
1438 If there is a fixed first value, for example, the letter "c" from a pattern
1439 such as (cat|cow|coyote), 1 is returned, and the character value can be
1440 retrieved using PCRE_INFO_FIRSTCHARACTER.
1441 .P
1442 If there is no fixed first value, and if either
1443 .sp
1444 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1445 starts with "^", or
1446 .sp
1447 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1448 (if it were set, the pattern would be anchored),
1449 .sp
1450 2 is returned, indicating that the pattern matches only at the start of a
1451 subject string or after any newline within the string. Otherwise 0 is
1452 returned. For anchored patterns, 0 is returned.
1453 .sp
1454 PCRE_INFO_FIRSTCHARACTER
1455 .sp
1456 Return the fixed first character value, if PCRE_INFO_FIRSTCHARACTERFLAGS
1457 returned 1; otherwise returns 0. The fourth argument should point to an
1458 \fBuint_t\fP variable.
1459 .P
1460 In the 8-bit library, the value is always less than 256. In the 16-bit library
1461 the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1462 can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1463 .P
1464 If there is no fixed first value, and if either
1465 .sp
1466 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1467 starts with "^", or
1468 .sp
1469 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1470 (if it were set, the pattern would be anchored),
1471 .sp
1472 -1 is returned, indicating that the pattern matches only at the start of a
1473 subject string or after any newline within the string. Otherwise -2 is
1474 returned. For anchored patterns, -2 is returned.
1475 .sp
1476 PCRE_INFO_REQUIREDCHARFLAGS
1477 .sp
1478 Returns 1 if there is a rightmost literal data unit that must exist in any
1479 matched string, other than at its start. The fourth argument should point to
1480 an \fBint\fP variable. If there is no such value, 0 is returned. If returning
1481 1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1482 .P
1483 For anchored patterns, a last literal value is recorded only if it follows
1484 something of variable length. For example, for the pattern /^a\ed+z\ed+/ the
1485 returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1486 /^a\edz\ed/ the returned value is 0.
1487 .sp
1488 PCRE_INFO_REQUIREDCHAR
1489 .sp
1490 Return the value of the rightmost literal data unit that must exist in any
1491 matched string, other than at its start, if such a value has been recorded. The
1492 fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1493 value, 0 is returned.
1494 .
1495 .
1496 .SH "REFERENCE COUNTS"
1497 .rs
1498 .sp
1499 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1500 .PP
1501 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1502 data block that contains a compiled pattern. It is provided for the benefit of
1503 applications that operate in an object-oriented manner, where different parts
1504 of the application may be using the same compiled pattern, but you want to free
1505 the block when they are all done.
1506 .P
1507 When a pattern is compiled, the reference count field is initialized to zero.
1508 It is changed only by calling this function, whose action is to add the
1509 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1510 function is the new value. However, the value of the count is constrained to
1511 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1512 it is forced to the appropriate limit value.
1513 .P
1514 Except when it is zero, the reference count is not correctly preserved if a
1515 pattern is compiled on one host and then transferred to a host whose byte-order
1516 is different. (This seems a highly unlikely scenario.)
1517 .
1518 .
1519 .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1520 .rs
1521 .sp
1522 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1523 .ti +5n
1524 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1525 .ti +5n
1526 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1527 .P
1528 The function \fBpcre_exec()\fP is called to match a subject string against a
1529 compiled pattern, which is passed in the \fIcode\fP argument. If the
1530 pattern was studied, the result of the study should be passed in the
1531 \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1532 and \fIextra\fP arguments as many times as you like, in order to match
1533 different subject strings with the same pattern.
1534 .P
1535 This function is the main matching facility of the library, and it operates in
1536 a Perl-like manner. For specialist use there is also an alternative matching
1537 function, which is described
1538 .\" HTML <a href="#dfamatch">
1539 .\" </a>
1540 below
1541 .\"
1542 in the section about the \fBpcre_dfa_exec()\fP function.
1543 .P
1544 In most applications, the pattern will have been compiled (and optionally
1545 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1546 possible to save compiled patterns and study data, and then use them later
1547 in different processes, possibly even on different hosts. For a discussion
1548 about this, see the
1549 .\" HREF
1550 \fBpcreprecompile\fP
1551 .\"
1552 documentation.
1553 .P
1554 Here is an example of a simple call to \fBpcre_exec()\fP:
1555 .sp
1556 int rc;
1557 int ovector[30];
1558 rc = pcre_exec(
1559 re, /* result of pcre_compile() */
1560 NULL, /* we didn't study the pattern */
1561 "some string", /* the subject string */
1562 11, /* the length of the subject string */
1563 0, /* start at offset 0 in the subject */
1564 0, /* default options */
1565 ovector, /* vector of integers for substring information */
1566 30); /* number of elements (NOT size in bytes) */
1567 .
1568 .
1569 .\" HTML <a name="extradata"></a>
1570 .SS "Extra data for \fBpcre_exec()\fR"
1571 .rs
1572 .sp
1573 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1574 data block. The \fBpcre_study()\fP function returns such a block (when it
1575 doesn't return NULL), but you can also create one for yourself, and pass
1576 additional information in it. The \fBpcre_extra\fP block contains the following
1577 fields (not necessarily in this order):
1578 .sp
1579 unsigned long int \fIflags\fP;
1580 void *\fIstudy_data\fP;
1581 void *\fIexecutable_jit\fP;
1582 unsigned long int \fImatch_limit\fP;
1583 unsigned long int \fImatch_limit_recursion\fP;
1584 void *\fIcallout_data\fP;
1585 const unsigned char *\fItables\fP;
1586 unsigned char **\fImark\fP;
1587 .sp
1588 In the 16-bit version of this structure, the \fImark\fP field has type
1589 "PCRE_UCHAR16 **".
1590 .sp
1591 In the 32-bit version of this structure, the \fImark\fP field has type
1592 "PCRE_UCHAR32 **".
1593 .P
1594 The \fIflags\fP field is used to specify which of the other fields are set. The
1595 flag bits are:
1596 .sp
1597 PCRE_EXTRA_CALLOUT_DATA
1598 PCRE_EXTRA_EXECUTABLE_JIT
1599 PCRE_EXTRA_MARK
1600 PCRE_EXTRA_MATCH_LIMIT
1601 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1602 PCRE_EXTRA_STUDY_DATA
1603 PCRE_EXTRA_TABLES
1604 .sp
1605 Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1606 the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1607 returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1608 should not set these yourself, but you may add to the block by setting other
1609 fields and their corresponding flag bits.
1610 .P
1611 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1612 vast amount of resources when running patterns that are not going to match,
1613 but which have a very large number of possibilities in their search trees. The
1614 classic example is a pattern that uses nested unlimited repeats.
1615 .P
1616 Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1617 calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1618 imposed on the number of times this function is called during a match, which
1619 has the effect of limiting the amount of backtracking that can take place. For
1620 patterns that are not anchored, the count restarts from zero for each position
1621 in the subject string.
1622 .P
1623 When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1624 with a JIT option, the way that the matching is executed is entirely different.
1625 However, there is still the possibility of runaway matching that goes on for a
1626 very long time, and so the \fImatch_limit\fP value is also used in this case
1627 (but in a different way) to limit how long the matching can continue.
1628 .P
1629 The default value for the limit can be set when PCRE is built; the default
1630 default is 10 million, which handles all but the most extreme cases. You can
1631 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1632 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1633 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1634 PCRE_ERROR_MATCHLIMIT.
1635 .P
1636 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1637 instead of limiting the total number of times that \fBmatch()\fP is called, it
1638 limits the depth of recursion. The recursion depth is a smaller number than the
1639 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1640 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1641 .P
1642 Limiting the recursion depth limits the amount of machine stack that can be
1643 used, or, when PCRE has been compiled to use memory on the heap instead of the
1644 stack, the amount of heap memory that can be used. This limit is not relevant,
1645 and is ignored, when matching is done using JIT compiled code.
1646 .P
1647 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1648 built; the default default is the same value as the default for
1649 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1650 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1651 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1652 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1653 .P
1654 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1655 and is described in the
1656 .\" HREF
1657 \fBpcrecallout\fP
1658 .\"
1659 documentation.
1660 .P
1661 The \fItables\fP field is used to pass a character tables pointer to
1662 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1663 pattern. A non-NULL value is stored with the compiled pattern only if custom
1664 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1665 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1666 internal tables to be used. This facility is helpful when re-using patterns
1667 that have been saved after compiling with an external set of tables, because
1668 the external tables might be at a different address when \fBpcre_exec()\fP is
1669 called. See the
1670 .\" HREF
1671 \fBpcreprecompile\fP
1672 .\"
1673 documentation for a discussion of saving compiled patterns for later use.
1674 .P
1675 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1676 be set to point to a suitable variable. If the pattern contains any
1677 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1678 a name to pass back, a pointer to the name string (zero terminated) is placed
1679 in the variable pointed to by the \fImark\fP field. The names are within the
1680 compiled pattern; if you wish to retain such a name you must copy it before
1681 freeing the memory of a compiled pattern. If there is no name to pass back, the
1682 variable pointed to by the \fImark\fP field is set to NULL. For details of the
1683 backtracking control verbs, see the section entitled
1684 .\" HTML <a href="pcrepattern#backtrackcontrol">
1685 .\" </a>
1686 "Backtracking control"
1687 .\"
1688 in the
1689 .\" HREF
1690 \fBpcrepattern\fP
1691 .\"
1692 documentation.
1693 .
1694 .
1695 .\" HTML <a name="execoptions"></a>
1696 .SS "Option bits for \fBpcre_exec()\fP"
1697 .rs
1698 .sp
1699 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1700 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1701 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1702 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1703 PCRE_PARTIAL_SOFT.
1704 .P
1705 If the pattern was successfully studied with one of the just-in-time (JIT)
1706 compile options, the only supported options for JIT execution are
1707 PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1708 PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1709 unsupported option is used, JIT execution is disabled and the normal
1710 interpretive code in \fBpcre_exec()\fP is run.
1711 .sp
1712 PCRE_ANCHORED
1713 .sp
1714 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1715 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1716 to be anchored by virtue of its contents, it cannot be made unachored at
1717 matching time.
1718 .sp
1719 PCRE_BSR_ANYCRLF
1720 PCRE_BSR_UNICODE
1721 .sp
1722 These options (which are mutually exclusive) control what the \eR escape
1723 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1724 match any Unicode newline sequence. These options override the choice that was
1725 made or defaulted when the pattern was compiled.
1726 .sp
1727 PCRE_NEWLINE_CR
1728 PCRE_NEWLINE_LF
1729 PCRE_NEWLINE_CRLF
1730 PCRE_NEWLINE_ANYCRLF
1731 PCRE_NEWLINE_ANY
1732 .sp
1733 These options override the newline definition that was chosen or defaulted when
1734 the pattern was compiled. For details, see the description of
1735 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1736 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1737 the way the match position is advanced after a match failure for an unanchored
1738 pattern.
1739 .P
1740 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1741 match attempt for an unanchored pattern fails when the current position is at a
1742 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1743 characters, the match position is advanced by two characters instead of one, in
1744 other words, to after the CRLF.
1745 .P
1746 The above rule is a compromise that makes the most common cases work as
1747 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1748 set), it does not match the string "\er\enA" because, after failing at the
1749 start, it skips both the CR and the LF before retrying. However, the pattern
1750 [\er\en]A does match that string, because it contains an explicit CR or LF
1751 reference, and so advances only by one character after the first failure.
1752 .P
1753 An explicit match for CR of LF is either a literal appearance of one of those
1754 characters, or one of the \er or \en escape sequences. Implicit matches such as
1755 [^X] do not count, nor does \es (which includes CR and LF in the characters
1756 that it matches).
1757 .P
1758 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1759 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1760 .sp
1761 PCRE_NOTBOL
1762 .sp
1763 This option specifies that first character of the subject string is not the
1764 beginning of a line, so the circumflex metacharacter should not match before
1765 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1766 never to match. This option affects only the behaviour of the circumflex
1767 metacharacter. It does not affect \eA.
1768 .sp
1769 PCRE_NOTEOL
1770 .sp
1771 This option specifies that the end of the subject string is not the end of a
1772 line, so the dollar metacharacter should not match it nor (except in multiline
1773 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1774 compile time) causes dollar never to match. This option affects only the
1775 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1776 .sp
1777 PCRE_NOTEMPTY
1778 .sp
1779 An empty string is not considered to be a valid match if this option is set. If
1780 there are alternatives in the pattern, they are tried. If all the alternatives
1781 match the empty string, the entire match fails. For example, if the pattern
1782 .sp
1783 a?b?
1784 .sp
1785 is applied to a string not beginning with "a" or "b", it matches an empty
1786 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1787 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1788 .sp
1789 PCRE_NOTEMPTY_ATSTART
1790 .sp
1791 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1792 the start of the subject is permitted. If the pattern is anchored, such a match
1793 can occur only if the pattern contains \eK.
1794 .P
1795 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1796 does make a special case of a pattern match of the empty string within its
1797 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1798 emulate Perl's behaviour after matching a null string by first trying the match
1799 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1800 if that fails, by advancing the starting offset (see below) and trying an
1801 ordinary match again. There is some code that demonstrates how to do this in
1802 the
1803 .\" HREF
1804 \fBpcredemo\fP
1805 .\"
1806 sample program. In the most general case, you have to check to see if the
1807 newline convention recognizes CRLF as a newline, and if so, and the current
1808 character is CR followed by LF, advance the starting offset by two characters
1809 instead of one.
1810 .sp
1811 PCRE_NO_START_OPTIMIZE
1812 .sp
1813 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1814 a match, in order to speed up the process. For example, if it is known that an
1815 unanchored match must start with a specific character, it searches the subject
1816 for that character, and fails immediately if it cannot find it, without
1817 actually running the main matching function. This means that a special item
1818 such as (*COMMIT) at the start of a pattern is not considered until after a
1819 suitable starting point for the match has been found. When callouts or (*MARK)
1820 items are in use, these "start-up" optimizations can cause them to be skipped
1821 if the pattern is never actually used. The start-up optimizations are in effect
1822 a pre-scan of the subject that takes place before the pattern is run.
1823 .P
1824 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1825 causing performance to suffer, but ensuring that in cases where the result is
1826 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1827 are considered at every possible starting position in the subject string. If
1828 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1829 time. The use of PCRE_NO_START_OPTIMIZE disables JIT execution; when it is set,
1830 matching is always done using interpretively.
1831 .P
1832 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1833 Consider the pattern
1834 .sp
1835 (*COMMIT)ABC
1836 .sp
1837 When this is compiled, PCRE records the fact that a match must start with the
1838 character "A". Suppose the subject string is "DEFABC". The start-up
1839 optimization scans along the subject, finds "A" and runs the first match
1840 attempt from there. The (*COMMIT) item means that the pattern must match the
1841 current starting position, which in this case, it does. However, if the same
1842 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1843 subject string does not happen. The first match attempt is run starting from
1844 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1845 the overall result is "no match". If the pattern is studied, more start-up
1846 optimizations may be used. For example, a minimum length for the subject may be
1847 recorded. Consider the pattern
1848 .sp
1849 (*MARK:A)(X|Y)
1850 .sp
1851 The minimum length for a match is one character. If the subject is "ABC", there
1852 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1853 If the pattern is studied, the final attempt does not take place, because PCRE
1854 knows that the subject is too short, and so the (*MARK) is never encountered.
1855 In this case, studying the pattern does not affect the overall match result,
1856 which is still "no match", but it does affect the auxiliary information that is
1857 returned.
1858 .sp
1859 PCRE_NO_UTF8_CHECK
1860 .sp
1861 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1862 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1863 The entire string is checked before any other processing takes place. The value
1864 of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1865 UTF-8 character. There is a discussion about the
1866 .\" HTML <a href="pcreunicode.html#utf8strings">
1867 .\" </a>
1868 validity of UTF-8 strings
1869 .\"
1870 in the
1871 .\" HREF
1872 \fBpcreunicode\fP
1873 .\"
1874 page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1875 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1876 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1877 cases, information about the precise nature of the error may also be returned
1878 (see the descriptions of these errors in the section entitled \fIError return
1879 values from\fP \fBpcre_exec()\fP
1880 .\" HTML <a href="#errorlist">
1881 .\" </a>
1882 below).
1883 .\"
1884 If \fIstartoffset\fP contains a value that does not point to the start of a
1885 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1886 returned.
1887 .P
1888 If you already know that your subject is valid, and you want to skip these
1889 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1890 calling \fBpcre_exec()\fP. You might want to do this for the second and
1891 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1892 all the matches in a single subject string. However, you should be sure that
1893 the value of \fIstartoffset\fP points to the start of a character (or the end
1894 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1895 invalid string as a subject or an invalid value of \fIstartoffset\fP is
1896 undefined. Your program may crash.
1897 .sp
1898 PCRE_PARTIAL_HARD
1899 PCRE_PARTIAL_SOFT
1900 .sp
1901 These options turn on the partial matching feature. For backwards
1902 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1903 occurs if the end of the subject string is reached successfully, but there are
1904 not enough subject characters to complete the match. If this happens when
1905 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1906 testing any remaining alternatives. Only if no complete match can be found is
1907 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1908 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1909 but only if no complete match can be found.
1910 .P
1911 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1912 partial match is found, \fBpcre_exec()\fP immediately returns
1913 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1914 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1915 important that an alternative complete match.
1916 .P
1917 In both cases, the portion of the string that was inspected when the partial
1918 match was found is set as the first matching string. There is a more detailed
1919 discussion of partial and multi-segment matching, with examples, in the
1920 .\" HREF
1921 \fBpcrepartial\fP
1922 .\"
1923 documentation.
1924 .
1925 .
1926 .SS "The string to be matched by \fBpcre_exec()\fP"
1927 .rs
1928 .sp
1929 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1930 \fIsubject\fP, a length in bytes in \fIlength\fP, and a starting byte offset
1931 in \fIstartoffset\fP. If this is negative or greater than the length of the
1932 subject, \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting
1933 offset is zero, the search for a match starts at the beginning of the subject,
1934 and this is by far the most common case. In UTF-8 mode, the byte offset must
1935 point to the start of a UTF-8 character (or the end of the subject). Unlike the
1936 pattern string, the subject may contain binary zero bytes.
1937 .P
1938 A non-zero starting offset is useful when searching for another match in the
1939 same subject by calling \fBpcre_exec()\fP again after a previous success.
1940 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1941 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1942 lookbehind. For example, consider the pattern
1943 .sp
1944 \eBiss\eB
1945 .sp
1946 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1947 the current position in the subject is not a word boundary.) When applied to
1948 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1949 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1950 subject, namely "issipi", it does not match, because \eB is always false at the
1951 start of the subject, which is deemed to be a word boundary. However, if
1952 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1953 set to 4, it finds the second occurrence of "iss" because it is able to look
1954 behind the starting point to discover that it is preceded by a letter.
1955 .P
1956 Finding all the matches in a subject is tricky when the pattern can match an
1957 empty string. It is possible to emulate Perl's /g behaviour by first trying the
1958 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1959 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1960 and trying an ordinary match again. There is some code that demonstrates how to
1961 do this in the
1962 .\" HREF
1963 \fBpcredemo\fP
1964 .\"
1965 sample program. In the most general case, you have to check to see if the
1966 newline convention recognizes CRLF as a newline, and if so, and the current
1967 character is CR followed by LF, advance the starting offset by two characters
1968 instead of one.
1969 .P
1970 If a non-zero starting offset is passed when the pattern is anchored, one
1971 attempt to match at the given offset is made. This can only succeed if the
1972 pattern does not require the match to be at the start of the subject.
1973 .
1974 .
1975 .SS "How \fBpcre_exec()\fP returns captured substrings"
1976 .rs
1977 .sp
1978 In general, a pattern matches a certain portion of the subject, and in
1979 addition, further substrings from the subject may be picked out by parts of the
1980 pattern. Following the usage in Jeffrey Friedl's book, this is called
1981 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1982 a fragment of a pattern that picks out a substring. PCRE supports several other
1983 kinds of parenthesized subpattern that do not cause substrings to be captured.
1984 .P
1985 Captured substrings are returned to the caller via a vector of integers whose
1986 address is passed in \fIovector\fP. The number of elements in the vector is
1987 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
1988 argument is NOT the size of \fIovector\fP in bytes.
1989 .P
1990 The first two-thirds of the vector is used to pass back captured substrings,
1991 each substring using a pair of integers. The remaining third of the vector is
1992 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1993 and is not available for passing back information. The number passed in
1994 \fIovecsize\fP should always be a multiple of three. If it is not, it is
1995 rounded down.
1996 .P
1997 When a match is successful, information about captured substrings is returned
1998 in pairs of integers, starting at the beginning of \fIovector\fP, and
1999 continuing up to two-thirds of its length at the most. The first element of
2000 each pair is set to the byte offset of the first character in a substring, and
2001 the second is set to the byte offset of the first character after the end of a
2002 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
2003 mode. They are not character counts.
2004 .P
2005 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
2006 portion of the subject string matched by the entire pattern. The next pair is
2007 used for the first capturing subpattern, and so on. The value returned by
2008 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
2009 For example, if two substrings have been captured, the returned value is 3. If
2010 there are no capturing subpatterns, the return value from a successful match is
2011 1, indicating that just the first pair of offsets has been set.
2012 .P
2013 If a capturing subpattern is matched repeatedly, it is the last portion of the
2014 string that it matched that is returned.
2015 .P
2016 If the vector is too small to hold all the captured substring offsets, it is
2017 used as far as possible (up to two-thirds of its length), and the function
2018 returns a value of zero. If neither the actual string matched nor any captured
2019 substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2020 passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2021 back references and the \fIovector\fP is not big enough to remember the related
2022 substrings, PCRE has to get additional memory for use during matching. Thus it
2023 is usually advisable to supply an \fIovector\fP of reasonable size.
2024 .P
2025 There are some cases where zero is returned (indicating vector overflow) when
2026 in fact the vector is exactly the right size for the final match. For example,
2027 consider the pattern
2028 .sp
2029 (a)(?:(b)c|bd)
2030 .sp
2031 If a vector of 6 elements (allowing for only 1 captured substring) is given
2032 with subject string "abd", \fBpcre_exec()\fP will try to set the second
2033 captured string, thereby recording a vector overflow, before failing to match
2034 "c" and backing up to try the second alternative. The zero return, however,
2035 does correctly indicate that the maximum number of slots (namely 2) have been
2036 filled. In similar cases where there is temporary overflow, but the final
2037 number of used slots is actually less than the maximum, a non-zero value is
2038 returned.
2039 .P
2040 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
2041 subpatterns there are in a compiled pattern. The smallest size for
2042 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
2043 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
2044 .P
2045 It is possible for capturing subpattern number \fIn+1\fP to match some part of
2046 the subject when subpattern \fIn\fP has not been used at all. For example, if
2047 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2048 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2049 happens, both values in the offset pairs corresponding to unused subpatterns
2050 are set to -1.
2051 .P
2052 Offset values that correspond to unused subpatterns at the end of the
2053 expression are also set to -1. For example, if the string "abc" is matched
2054 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2055 return from the function is 2, because the highest used capturing subpattern
2056 number is 1, and the offsets for for the second and third capturing subpatterns
2057 (assuming the vector is large enough, of course) are set to -1.
2058 .P
2059 \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
2060 correspond to capturing parentheses in the pattern are never changed. That is,
2061 if a pattern contains \fIn\fP capturing parentheses, no more than
2062 \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
2063 elements (in the first two-thirds) retain whatever values they previously had.
2064 .P
2065 Some convenience functions are provided for extracting the captured substrings
2066 as separate strings. These are described below.
2067 .
2068 .
2069 .\" HTML <a name="errorlist"></a>
2070 .SS "Error return values from \fBpcre_exec()\fP"
2071 .rs
2072 .sp
2073 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
2074 defined in the header file:
2075 .sp
2076 PCRE_ERROR_NOMATCH (-1)
2077 .sp
2078 The subject string did not match the pattern.
2079 .sp
2080 PCRE_ERROR_NULL (-2)
2081 .sp
2082 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
2083 NULL and \fIovecsize\fP was not zero.
2084 .sp
2085 PCRE_ERROR_BADOPTION (-3)
2086 .sp
2087 An unrecognized bit was set in the \fIoptions\fP argument.
2088 .sp
2089 PCRE_ERROR_BADMAGIC (-4)
2090 .sp
2091 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2092 the case when it is passed a junk pointer and to detect when a pattern that was
2093 compiled in an environment of one endianness is run in an environment with the
2094 other endianness. This is the error that PCRE gives when the magic number is
2095 not present.
2096 .sp
2097 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2098 .sp
2099 While running the pattern match, an unknown item was encountered in the
2100 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2101 of the compiled pattern.
2102 .sp
2103 PCRE_ERROR_NOMEMORY (-6)
2104 .sp
2105 If a pattern contains back references, but the \fIovector\fP that is passed to
2106 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2107 gets a block of memory at the start of matching to use for this purpose. If the
2108 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2109 automatically freed at the end of matching.
2110 .P
2111 This error is also given if \fBpcre_stack_malloc()\fP fails in
2112 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2113 \fB--disable-stack-for-recursion\fP.
2114 .sp
2115 PCRE_ERROR_NOSUBSTRING (-7)
2116 .sp
2117 This error is used by the \fBpcre_copy_substring()\fP,
2118 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2119 below). It is never returned by \fBpcre_exec()\fP.
2120 .sp
2121 PCRE_ERROR_MATCHLIMIT (-8)
2122 .sp
2123 The backtracking limit, as specified by the \fImatch_limit\fP field in a
2124 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2125 above.
2126 .sp
2127 PCRE_ERROR_CALLOUT (-9)
2128 .sp
2129 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2130 use by callout functions that want to yield a distinctive error code. See the
2131 .\" HREF
2132 \fBpcrecallout\fP
2133 .\"
2134 documentation for details.
2135 .sp
2136 PCRE_ERROR_BADUTF8 (-10)
2137 .sp
2138 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2139 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2140 (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2141 UTF-8 character is placed in the first element, and a reason code is placed in
2142 the second element. The reason codes are listed in the
2143 .\" HTML <a href="#badutf8reasons">
2144 .\" </a>
2145 following section.
2146 .\"
2147 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2148 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2149 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2150 .sp
2151 PCRE_ERROR_BADUTF8_OFFSET (-11)
2152 .sp
2153 The UTF-8 byte sequence that was passed as a subject was checked and found to
2154 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2155 \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2156 end of the subject.
2157 .sp
2158 PCRE_ERROR_PARTIAL (-12)
2159 .sp
2160 The subject string did not match, but it did match partially. See the
2161 .\" HREF
2162 \fBpcrepartial\fP
2163 .\"
2164 documentation for details of partial matching.
2165 .sp
2166 PCRE_ERROR_BADPARTIAL (-13)
2167 .sp
2168 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2169 option was used with a compiled pattern containing items that were not
2170 supported for partial matching. From release 8.00 onwards, there are no
2171 restrictions on partial matching.
2172 .sp
2173 PCRE_ERROR_INTERNAL (-14)
2174 .sp
2175 An unexpected internal error has occurred. This error could be caused by a bug
2176 in PCRE or by overwriting of the compiled pattern.
2177 .sp
2178 PCRE_ERROR_BADCOUNT (-15)
2179 .sp
2180 This error is given if the value of the \fIovecsize\fP argument is negative.
2181 .sp
2182 PCRE_ERROR_RECURSIONLIMIT (-21)
2183 .sp
2184 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2185 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2186 description above.
2187 .sp
2188 PCRE_ERROR_BADNEWLINE (-23)
2189 .sp
2190 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2191 .sp
2192 PCRE_ERROR_BADOFFSET (-24)
2193 .sp
2194 The value of \fIstartoffset\fP was negative or greater than the length of the
2195 subject, that is, the value in \fIlength\fP.
2196 .sp
2197 PCRE_ERROR_SHORTUTF8 (-25)
2198 .sp
2199 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2200 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2201 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2202 fact sufficient to detect this case, but this special error code for
2203 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2204 retained for backwards compatibility.
2205 .sp
2206 PCRE_ERROR_RECURSELOOP (-26)
2207 .sp
2208 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2209 the pattern. Specifically, it means that either the whole pattern or a
2210 subpattern has been called recursively for the second time at the same position
2211 in the subject string. Some simple patterns that might do this are detected and
2212 faulted at compile time, but more complicated cases, in particular mutual
2213 recursions between two different subpatterns, cannot be detected until run
2214 time.
2215 .sp
2216 PCRE_ERROR_JIT_STACKLIMIT (-27)
2217 .sp
2218 This error is returned when a pattern that was successfully studied using a
2219 JIT compile option is being matched, but the memory available for the
2220 just-in-time processing stack is not large enough. See the
2221 .\" HREF
2222 \fBpcrejit\fP
2223 .\"
2224 documentation for more details.
2225 .sp
2226 PCRE_ERROR_BADMODE (-28)
2227 .sp
2228 This error is given if a pattern that was compiled by the 8-bit library is
2229 passed to a 16-bit or 32-bit library function, or vice versa.
2230 .sp
2231 PCRE_ERROR_BADENDIANNESS (-29)
2232 .sp
2233 This error is given if a pattern that was compiled and saved is reloaded on a
2234 host with different endianness. The utility function
2235 \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2236 so that it runs on the new host.
2237 .sp
2238 PCRE_ERROR_BADLENGTH (-32)
2239 .sp
2240 This error is given if \fBpcre_exec()\fP is called with a negative value for
2241 the \fIlength\fP argument.
2242 .P
2243 Error numbers -16 to -20, -22, 30, and -31 are not used by \fBpcre_exec()\fP.
2244 .
2245 .
2246 .\" HTML <a name="badutf8reasons"></a>
2247 .SS "Reason codes for invalid UTF-8 strings"
2248 .rs
2249 .sp
2250 This section applies only to the 8-bit library. The corresponding information
2251 for the 16-bit library is given in the
2252 .\" HREF
2253 \fBpcre16\fP
2254 .\"
2255 page. The corresponding information for the 32-bit library is given in the
2256 .\" HREF
2257 \fBpcre32\fP
2258 .\"
2259 page.
2260 .P
2261 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2262 PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2263 least 2, the offset of the start of the invalid UTF-8 character is placed in
2264 the first output vector element (\fIovector[0]\fP) and a reason code is placed
2265 in the second element (\fIovector[1]\fP). The reason codes are given names in
2266 the \fBpcre.h\fP header file:
2267 .sp
2268 PCRE_UTF8_ERR1
2269 PCRE_UTF8_ERR2
2270 PCRE_UTF8_ERR3
2271 PCRE_UTF8_ERR4
2272 PCRE_UTF8_ERR5
2273 .sp
2274 The string ends with a truncated UTF-8 character; the code specifies how many
2275 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2276 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2277 allows for up to 6 bytes, and this is checked first; hence the possibility of
2278 4 or 5 missing bytes.
2279 .sp
2280 PCRE_UTF8_ERR6
2281 PCRE_UTF8_ERR7
2282 PCRE_UTF8_ERR8
2283 PCRE_UTF8_ERR9
2284 PCRE_UTF8_ERR10
2285 .sp
2286 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2287 character do not have the binary value 0b10 (that is, either the most
2288 significant bit is 0, or the next bit is 1).
2289 .sp
2290 PCRE_UTF8_ERR11
2291 PCRE_UTF8_ERR12
2292 .sp
2293 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2294 these code points are excluded by RFC 3629.
2295 .sp
2296 PCRE_UTF8_ERR13
2297 .sp
2298 A 4-byte character has a value greater than 0x10fff; these code points are
2299 excluded by RFC 3629.
2300 .sp
2301 PCRE_UTF8_ERR14
2302 .sp
2303 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2304 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2305 from UTF-8.
2306 .sp
2307 PCRE_UTF8_ERR15
2308 PCRE_UTF8_ERR16
2309 PCRE_UTF8_ERR17
2310 PCRE_UTF8_ERR18
2311 PCRE_UTF8_ERR19
2312 .sp
2313 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2314 value that can be represented by fewer bytes, which is invalid. For example,
2315 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2316 one byte.
2317 .sp
2318 PCRE_UTF8_ERR20
2319 .sp
2320 The two most significant bits of the first byte of a character have the binary
2321 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2322 byte can only validly occur as the second or subsequent byte of a multi-byte
2323 character.
2324 .sp
2325 PCRE_UTF8_ERR21
2326 .sp
2327 The first byte of a character has the value 0xfe or 0xff. These values can
2328 never occur in a valid UTF-8 string.
2329 .sp
2330 PCRE_UTF8_ERR2
2331 .sp
2332 Non-character. These are the last two characters in each plane (0xfffe, 0xffff,
2333 0x1fffe, 0x1ffff .. 0x10fffe, 0x10ffff), and the characters 0xfdd0..0xfdef.
2334 .
2335 .
2336 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2337 .rs
2338 .sp
2339 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2340 .ti +5n
2341 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
2342 .ti +5n
2343 .B int \fIbuffersize\fP);
2344 .PP
2345 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2346 .ti +5n
2347 .B int \fIstringcount\fP, int \fIstringnumber\fP,
2348 .ti +5n
2349 .B const char **\fIstringptr\fP);
2350 .PP
2351 .B int pcre_get_substring_list(const char *\fIsubject\fP,
2352 .ti +5n
2353 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
2354 .PP
2355 Captured substrings can be accessed directly by using the offsets returned by
2356 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2357 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2358 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2359 as new, separate, zero-terminated strings. These functions identify substrings
2360 by number. The next section describes functions for extracting named
2361 substrings.
2362 .P
2363 A substring that contains a binary zero is correctly extracted and has a
2364 further zero added on the end, but the result is not, of course, a C string.
2365 However, you can process such a string by referring to the length that is
2366 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2367 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2368 for handling strings containing binary zeros, because the end of the final
2369 string is not independently indicated.
2370 .P
2371 The first three arguments are the same for all three of these functions:
2372 \fIsubject\fP is the subject string that has just been successfully matched,
2373 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2374 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2375 captured by the match, including the substring that matched the entire regular
2376 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2377 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2378 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2379 number of elements in the vector divided by three.
2380 .P
2381 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2382 extract a single substring, whose number is given as \fIstringnumber\fP. A
2383 value of zero extracts the substring that matched the entire pattern, whereas
2384 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2385 the string is placed in \fIbuffer\fP, whose length is given by
2386 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2387 obtained via \fBpcre_malloc\fP, and its address is returned via
2388 \fIstringptr\fP. The yield of the function is the length of the string, not
2389 including the terminating zero, or one of these error codes:
2390 .sp
2391 PCRE_ERROR_NOMEMORY (-6)
2392 .sp
2393 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2394 memory failed for \fBpcre_get_substring()\fP.
2395 .sp
2396 PCRE_ERROR_NOSUBSTRING (-7)
2397 .sp
2398 There is no substring whose number is \fIstringnumber\fP.
2399 .P
2400 The \fBpcre_get_substring_list()\fP function extracts all available substrings
2401 and builds a list of pointers to them. All this is done in a single block of
2402 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2403 is returned via \fIlistptr\fP, which is also the start of the list of string
2404 pointers. The end of the list is marked by a NULL pointer. The yield of the
2405 function is zero if all went well, or the error code
2406 .sp
2407 PCRE_ERROR_NOMEMORY (-6)
2408 .sp
2409 if the attempt to get the memory block failed.
2410 .P
2411 When any of these functions encounter a substring that is unset, which can
2412 happen when capturing subpattern number \fIn+1\fP matches some part of the
2413 subject, but subpattern \fIn\fP has not been used at all, they return an empty
2414 string. This can be distinguished from a genuine zero-length substring by
2415 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2416 substrings.
2417 .P
2418 The two convenience functions \fBpcre_free_substring()\fP and
2419 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2420 a previous call of \fBpcre_get_substring()\fP or
2421 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2422 the function pointed to by \fBpcre_free\fP, which of course could be called
2423 directly from a C program. However, PCRE is used in some situations where it is
2424 linked via a special interface to another programming language that cannot use
2425 \fBpcre_free\fP directly; it is for these cases that the functions are
2426 provided.
2427 .
2428 .
2429 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2430 .rs
2431 .sp
2432 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2433 .ti +5n
2434 .B const char *\fIname\fP);
2435 .PP
2436 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2437 .ti +5n
2438 .B const char *\fIsubject\fP, int *\fIovector\fP,
2439 .ti +5n
2440 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2441 .ti +5n
2442 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
2443 .PP
2444 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2445 .ti +5n
2446 .B const char *\fIsubject\fP, int *\fIovector\fP,
2447 .ti +5n
2448 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2449 .ti +5n
2450 .B const char **\fIstringptr\fP);
2451 .PP
2452 To extract a substring by name, you first have to find associated number.
2453 For example, for this pattern
2454 .sp
2455 (a+)b(?<xxx>\ed+)...
2456 .sp
2457 the number of the subpattern called "xxx" is 2. If the name is known to be
2458 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2459 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2460 pattern, and the second is the name. The yield of the function is the
2461 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2462 that name.
2463 .P
2464 Given the number, you can extract the substring directly, or use one of the
2465 functions described in the previous section. For convenience, there are also
2466 two functions that do the whole job.
2467 .P
2468 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2469 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2470 functions that extract by number. As these are described in the previous
2471 section, they are not re-described here. There are just two differences:
2472 .P
2473 First, instead of a substring number, a substring name is given. Second, there
2474 is an extra argument, given at the start, which is a pointer to the compiled
2475 pattern. This is needed in order to gain access to the name-to-number
2476 translation table.
2477 .P
2478 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2479 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2480 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2481 the behaviour may not be what you want (see the next section).
2482 .P
2483 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2484 subpatterns with the same number, as described in the
2485 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2486 .\" </a>
2487 section on duplicate subpattern numbers
2488 .\"
2489 in the
2490 .\" HREF
2491 \fBpcrepattern\fP
2492 .\"
2493 page, you cannot use names to distinguish the different subpatterns, because
2494 names are not included in the compiled code. The matching process uses only
2495 numbers. For this reason, the use of different names for subpatterns of the
2496 same number causes an error at compile time.
2497 .
2498 .
2499 .SH "DUPLICATE SUBPATTERN NAMES"
2500 .rs
2501 .sp
2502 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2503 .ti +5n
2504 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
2505 .PP
2506 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2507 are not required to be unique. (Duplicate names are always allowed for
2508 subpatterns with the same number, created by using the (?| feature. Indeed, if
2509 such subpatterns are named, they are required to use the same names.)
2510 .P
2511 Normally, patterns with duplicate names are such that in any one match, only
2512 one of the named subpatterns participates. An example is shown in the
2513 .\" HREF
2514 \fBpcrepattern\fP
2515 .\"
2516 documentation.
2517 .P
2518 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2519 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2520 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2521 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2522 returns one of the numbers that are associated with the name, but it is not
2523 defined which it is.
2524 .P
2525 If you want to get full details of all captured substrings for a given name,
2526 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2527 argument is the compiled pattern, and the second is the name. The third and
2528 fourth are pointers to variables which are updated by the function. After it
2529 has run, they point to the first and last entries in the name-to-number table
2530 for the given name. The function itself returns the length of each entry, or
2531 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2532 described above in the section entitled \fIInformation about a pattern\fP
2533 .\" HTML <a href="#infoaboutpattern">
2534 .\" </a>
2535 above.
2536 .\"
2537 Given all the relevant entries for the name, you can extract each of their
2538 numbers, and hence the captured data, if any.
2539 .
2540 .
2541 .SH "FINDING ALL POSSIBLE MATCHES"
2542 .rs
2543 .sp
2544 The traditional matching function uses a similar algorithm to Perl, which stops
2545 when it finds the first match, starting at a given point in the subject. If you
2546 want to find all possible matches, or the longest possible match, consider
2547 using the alternative matching function (see below) instead. If you cannot use
2548 the alternative function, but still need to find all possible matches, you
2549 can kludge it up by making use of the callout facility, which is described in
2550 the
2551 .\" HREF
2552 \fBpcrecallout\fP
2553 .\"
2554 documentation.
2555 .P
2556 What you have to do is to insert a callout right at the end of the pattern.
2557 When your callout function is called, extract and save the current matched
2558 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2559 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2560 will yield PCRE_ERROR_NOMATCH.
2561 .
2562 .
2563 .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2564 .rs
2565 .sp
2566 Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2567 stack, which in certain environments can be rather limited in size. Some users
2568 find it helpful to have an estimate of the amount of stack that is used by
2569 \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2570 .\" HREF
2571 \fBpcrestack\fP
2572 .\"
2573 documentation. The estimate that is output by \fBpcretest\fP when called with
2574 the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2575 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2576 .P
2577 Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2578 the negative error code PCRE_ERROR_NULL, but with this special combination of
2579 arguments, it returns instead a negative number whose absolute value is the
2580 approximate stack frame size in bytes. (A negative number is used so that it is
2581 clear that no match has happened.) The value is approximate because in some
2582 cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2583 additional variables on the stack.
2584 .P
2585 If PCRE has been compiled to use the heap instead of the stack for recursion,
2586 the value returned is the size of each block that is obtained from the heap.
2587 .
2588 .
2589 .\" HTML <a name="dfamatch"></a>
2590 .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2591 .rs
2592 .sp
2593 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2594 .ti +5n
2595 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2596 .ti +5n
2597 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2598 .ti +5n
2599 .B int *\fIworkspace\fP, int \fIwscount\fP);
2600 .P
2601 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2602 a compiled pattern, using a matching algorithm that scans the subject string
2603 just once, and does not backtrack. This has different characteristics to the
2604 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2605 patterns are not supported. Nevertheless, there are times when this kind of
2606 matching can be useful. For a discussion of the two matching algorithms, and a
2607 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2608 .\" HREF
2609 \fBpcrematching\fP
2610 .\"
2611 documentation.
2612 .P
2613 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2614 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2615 different way, and this is described below. The other common arguments are used
2616 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2617 here.
2618 .P
2619 The two additional arguments provide workspace for the function. The workspace
2620 vector should contain at least 20 elements. It is used for keeping track of
2621 multiple paths through the pattern tree. More workspace will be needed for
2622 patterns and subjects where there are a lot of potential matches.
2623 .P
2624 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2625 .sp
2626 int rc;
2627 int ovector[10];
2628 int wspace[20];
2629 rc = pcre_dfa_exec(
2630 re, /* result of pcre_compile() */
2631 NULL, /* we didn't study the pattern */
2632 "some string", /* the subject string */
2633 11, /* the length of the subject string */
2634 0, /* start at offset 0 in the subject */
2635 0, /* default options */
2636 ovector, /* vector of integers for substring information */
2637 10, /* number of elements (NOT size in bytes) */
2638 wspace, /* working space vector */
2639 20); /* number of elements (NOT size in bytes) */
2640 .
2641 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2642 .rs
2643 .sp
2644 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2645 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2646 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2647 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2648 PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2649 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2650 so their description is not repeated here.
2651 .sp
2652 PCRE_PARTIAL_HARD
2653 PCRE_PARTIAL_SOFT
2654 .sp
2655 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2656 details are slightly different. When PCRE_PARTIAL_HARD is set for
2657 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2658 is reached and there is still at least one matching possibility that requires
2659 additional characters. This happens even if some complete matches have also
2660 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2661 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2662 there have been no complete matches, but there is still at least one matching
2663 possibility. The portion of the string that was inspected when the longest
2664 partial match was found is set as the first matching string in both cases.
2665 There is a more detailed discussion of partial and multi-segment matching, with
2666 examples, in the
2667 .\" HREF
2668 \fBpcrepartial\fP
2669 .\"
2670 documentation.
2671 .sp
2672 PCRE_DFA_SHORTEST
2673 .sp
2674 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2675 soon as it has found one match. Because of the way the alternative algorithm
2676 works, this is necessarily the shortest possible match at the first possible
2677 matching point in the subject string.
2678 .sp
2679 PCRE_DFA_RESTART
2680 .sp
2681 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2682 again, with additional subject characters, and have it continue with the same
2683 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2684 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2685 before because data about the match so far is left in them after a partial
2686 match. There is more discussion of this facility in the
2687 .\" HREF
2688 \fBpcrepartial\fP
2689 .\"
2690 documentation.
2691 .
2692 .
2693 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2694 .rs
2695 .sp
2696 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2697 substring in the subject. Note, however, that all the matches from one run of
2698 the function start at the same point in the subject. The shorter matches are
2699 all initial substrings of the longer matches. For example, if the pattern
2700 .sp
2701 <.*>
2702 .sp
2703 is matched against the string
2704 .sp
2705 This is <something> <something else> <something further> no more
2706 .sp
2707 the three matched strings are
2708 .sp
2709 <something>
2710 <something> <something else>
2711 <something> <something else> <something further>
2712 .sp
2713 On success, the yield of the function is a number greater than zero, which is
2714 the number of matched substrings. The substrings themselves are returned in
2715 \fIovector\fP. Each string uses two elements; the first is the offset to the
2716 start, and the second is the offset to the end. In fact, all the strings have
2717 the same start offset. (Space could have been saved by giving this only once,
2718 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2719 returns data, even though the meaning of the strings is different.)
2720 .P
2721 The strings are returned in reverse order of length; that is, the longest
2722 matching string is given first. If there were too many matches to fit into
2723 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2724 the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2725 the entire \fIovector\fP for returning matched strings.
2726 .
2727 .
2728 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2729 .rs
2730 .sp
2731 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2732 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2733 described
2734 .\" HTML <a href="#errorlist">
2735 .\" </a>
2736 above.
2737 .\"
2738 There are in addition the following errors that are specific to
2739 \fBpcre_dfa_exec()\fP:
2740 .sp
2741 PCRE_ERROR_DFA_UITEM (-16)
2742 .sp
2743 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2744 that it does not support, for instance, the use of \eC or a back reference.
2745 .sp
2746 PCRE_ERROR_DFA_UCOND (-17)
2747 .sp
2748 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2749 uses a back reference for the condition, or a test for recursion in a specific
2750 group. These are not supported.
2751 .sp
2752 PCRE_ERROR_DFA_UMLIMIT (-18)
2753 .sp
2754 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2755 block that contains a setting of the \fImatch_limit\fP or
2756 \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2757 meaningless for DFA matching).
2758 .sp
2759 PCRE_ERROR_DFA_WSSIZE (-19)
2760 .sp
2761 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2762 \fIworkspace\fP vector.
2763 .sp
2764 PCRE_ERROR_DFA_RECURSE (-20)
2765 .sp
2766 When a recursive subpattern is processed, the matching function calls itself
2767 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2768 error is given if the output vector is not large enough. This should be
2769 extremely rare, as a vector of size 1000 is used.
2770 .sp
2771 PCRE_ERROR_DFA_BADRESTART (-30)
2772 .sp
2773 When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2774 some plausibility checks are made on the contents of the workspace, which
2775 should contain data about the previous partial match. If any of these checks
2776 fail, this error is given.
2777 .
2778 .
2779 .SH "SEE ALSO"
2780 .rs
2781 .sp
2782 \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2783 \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2784 \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2785 \fBpcrestack\fP(3).
2786 .
2787 .
2788 .SH AUTHOR
2789 .rs
2790 .sp
2791 .nf
2792 Philip Hazel
2793 University Computing Service
2794 Cambridge CB2 3QH, England.
2795 .fi
2796 .
2797 .
2798 .SH REVISION
2799 .rs
2800 .sp
2801 .nf
2802 Last updated: 29 October 2012
2803 Copyright (c) 1997-2012 University of Cambridge.
2804 .fi

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5