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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1363 - (show annotations)
Tue Oct 1 16:54:40 2013 UTC (5 years, 11 months ago) by ph10
File size: 126313 byte(s)
Refactored auto-possessification code.
1 .TH PCREAPI 3 "01 October 2013" "PCRE 8.34"
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 .nf
12 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
13 .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
14 .B " const unsigned char *\fItableptr\fP);"
15 .sp
16 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
17 .B " int *\fIerrorcodeptr\fP,"
18 .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
19 .B " const unsigned char *\fItableptr\fP);"
20 .sp
21 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
22 .B " const char **\fIerrptr\fP);"
23 .sp
24 .B void pcre_free_study(pcre_extra *\fIextra\fP);
25 .sp
26 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
27 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
28 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);"
29 .sp
30 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
31 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
32 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
33 .B " int *\fIworkspace\fP, int \fIwscount\fP);"
34 .fi
35 .
36 .
37 .SH "PCRE NATIVE API STRING EXTRACTION FUNCTIONS"
38 .rs
39 .sp
40 .nf
41 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
42 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
43 .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
44 .B " char *\fIbuffer\fP, int \fIbuffersize\fP);"
45 .sp
46 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
47 .B " int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,"
48 .B " int \fIbuffersize\fP);"
49 .sp
50 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
51 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
52 .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
53 .B " const char **\fIstringptr\fP);"
54 .sp
55 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
56 .B " const char *\fIname\fP);"
57 .sp
58 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
59 .B " const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);"
60 .sp
61 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
62 .B " int \fIstringcount\fP, int \fIstringnumber\fP,"
63 .B " const char **\fIstringptr\fP);"
64 .sp
65 .B int pcre_get_substring_list(const char *\fIsubject\fP,
66 .B " int *\fIovector\fP, int \fIstringcount\fP, const char ***\fIlistptr\fP);"
67 .sp
68 .B void pcre_free_substring(const char *\fIstringptr\fP);
69 .sp
70 .B void pcre_free_substring_list(const char **\fIstringptr\fP);
71 .fi
72 .
73 .
74 .SH "PCRE NATIVE API AUXILIARY FUNCTIONS"
75 .rs
76 .sp
77 .nf
78 .B int pcre_jit_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
79 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
80 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
81 .B " pcre_jit_stack *\fIjstack\fP);"
82 .sp
83 .B pcre_jit_stack *pcre_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
84 .sp
85 .B void pcre_jit_stack_free(pcre_jit_stack *\fIstack\fP);
86 .sp
87 .B void pcre_assign_jit_stack(pcre_extra *\fIextra\fP,
88 .B " pcre_jit_callback \fIcallback\fP, void *\fIdata\fP);"
89 .sp
90 .B const unsigned char *pcre_maketables(void);
91 .sp
92 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
93 .B " int \fIwhat\fP, void *\fIwhere\fP);"
94 .sp
95 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
96 .sp
97 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
98 .sp
99 .B const char *pcre_version(void);
100 .sp
101 .B int pcre_pattern_to_host_byte_order(pcre *\fIcode\fP,
102 .B " pcre_extra *\fIextra\fP, const unsigned char *\fItables\fP);"
103 .fi
104 .
105 .
106 .SH "PCRE NATIVE API INDIRECTED FUNCTIONS"
107 .rs
108 .sp
109 .nf
110 .B void *(*pcre_malloc)(size_t);
111 .sp
112 .B void (*pcre_free)(void *);
113 .sp
114 .B void *(*pcre_stack_malloc)(size_t);
115 .sp
116 .B void (*pcre_stack_free)(void *);
117 .sp
118 .B int (*pcre_callout)(pcre_callout_block *);
119 .fi
120 .
121 .
122 .SH "PCRE 8-BIT, 16-BIT, AND 32-BIT LIBRARIES"
123 .rs
124 .sp
125 As well as support for 8-bit character strings, PCRE also supports 16-bit
126 strings (from release 8.30) and 32-bit strings (from release 8.32), by means of
127 two additional libraries. They can be built as well as, or instead of, the
128 8-bit library. To avoid too much complication, this document describes the
129 8-bit versions of the functions, with only occasional references to the 16-bit
130 and 32-bit libraries.
131 .P
132 The 16-bit and 32-bit functions operate in the same way as their 8-bit
133 counterparts; they just use different data types for their arguments and
134 results, and their names start with \fBpcre16_\fP or \fBpcre32_\fP instead of
135 \fBpcre_\fP. For every option that has UTF8 in its name (for example,
136 PCRE_UTF8), there are corresponding 16-bit and 32-bit names with UTF8 replaced
137 by UTF16 or UTF32, respectively. This facility is in fact just cosmetic; the
138 16-bit and 32-bit option names define the same bit values.
139 .P
140 References to bytes and UTF-8 in this document should be read as references to
141 16-bit data units and UTF-16 when using the 16-bit library, or 32-bit data
142 units and UTF-32 when using the 32-bit library, unless specified otherwise.
143 More details of the specific differences for the 16-bit and 32-bit libraries
144 are given in the
145 .\" HREF
146 \fBpcre16\fP
147 .\"
148 and
149 .\" HREF
150 \fBpcre32\fP
151 .\"
152 pages.
153 .
154 .
155 .SH "PCRE API OVERVIEW"
156 .rs
157 .sp
158 PCRE has its own native API, which is described in this document. There are
159 also some wrapper functions (for the 8-bit library only) that correspond to the
160 POSIX regular expression API, but they do not give access to all the
161 functionality. They are described in the
162 .\" HREF
163 \fBpcreposix\fP
164 .\"
165 documentation. Both of these APIs define a set of C function calls. A C++
166 wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
167 documented in the
168 .\" HREF
169 \fBpcrecpp\fP
170 .\"
171 page.
172 .P
173 The native API C function prototypes are defined in the header file
174 \fBpcre.h\fP, and on Unix-like systems the (8-bit) library itself is called
175 \fBlibpcre\fP. It can normally be accessed by adding \fB-lpcre\fP to the
176 command for linking an application that uses PCRE. The header file defines the
177 macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
178 for the library. Applications can use these to include support for different
179 releases of PCRE.
180 .P
181 In a Windows environment, if you want to statically link an application program
182 against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
183 including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
184 \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
185 \fB__declspec(dllimport)\fP, with unwanted results.
186 .P
187 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
188 and \fBpcre_exec()\fP are used for compiling and matching regular expressions
189 in a Perl-compatible manner. A sample program that demonstrates the simplest
190 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
191 source distribution. A listing of this program is given in the
192 .\" HREF
193 \fBpcredemo\fP
194 .\"
195 documentation, and the
196 .\" HREF
197 \fBpcresample\fP
198 .\"
199 documentation describes how to compile and run it.
200 .P
201 Just-in-time compiler support is an optional feature of PCRE that can be built
202 in appropriate hardware environments. It greatly speeds up the matching
203 performance of many patterns. Simple programs can easily request that it be
204 used if available, by setting an option that is ignored when it is not
205 relevant. More complicated programs might need to make use of the functions
206 \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
207 \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
208 .P
209 From release 8.32 there is also a direct interface for JIT execution, which
210 gives improved performance. The JIT-specific functions are discussed in the
211 .\" HREF
212 \fBpcrejit\fP
213 .\"
214 documentation.
215 .P
216 A second matching function, \fBpcre_dfa_exec()\fP, which is not
217 Perl-compatible, is also provided. This uses a different algorithm for the
218 matching. The alternative algorithm finds all possible matches (at a given
219 point in the subject), and scans the subject just once (unless there are
220 lookbehind assertions). However, this algorithm does not return captured
221 substrings. A description of the two matching algorithms and their advantages
222 and disadvantages is given in the
223 .\" HREF
224 \fBpcrematching\fP
225 .\"
226 documentation.
227 .P
228 In addition to the main compiling and matching functions, there are convenience
229 functions for extracting captured substrings from a subject string that is
230 matched by \fBpcre_exec()\fP. They are:
231 .sp
232 \fBpcre_copy_substring()\fP
233 \fBpcre_copy_named_substring()\fP
234 \fBpcre_get_substring()\fP
235 \fBpcre_get_named_substring()\fP
236 \fBpcre_get_substring_list()\fP
237 \fBpcre_get_stringnumber()\fP
238 \fBpcre_get_stringtable_entries()\fP
239 .sp
240 \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
241 provided, to free the memory used for extracted strings.
242 .P
243 The function \fBpcre_maketables()\fP is used to build a set of character tables
244 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
245 or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
246 specialist use. Most commonly, no special tables are passed, in which case
247 internal tables that are generated when PCRE is built are used.
248 .P
249 The function \fBpcre_fullinfo()\fP is used to find out information about a
250 compiled pattern. The function \fBpcre_version()\fP returns a pointer to a
251 string containing the version of PCRE and its date of release.
252 .P
253 The function \fBpcre_refcount()\fP maintains a reference count in a data block
254 containing a compiled pattern. This is provided for the benefit of
255 object-oriented applications.
256 .P
257 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
258 the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
259 respectively. PCRE calls the memory management functions via these variables,
260 so a calling program can replace them if it wishes to intercept the calls. This
261 should be done before calling any PCRE functions.
262 .P
263 The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
264 indirections to memory management functions. These special functions are used
265 only when PCRE is compiled to use the heap for remembering data, instead of
266 recursive function calls, when running the \fBpcre_exec()\fP function. See the
267 .\" HREF
268 \fBpcrebuild\fP
269 .\"
270 documentation for details of how to do this. It is a non-standard way of
271 building PCRE, for use in environments that have limited stacks. Because of the
272 greater use of memory management, it runs more slowly. Separate functions are
273 provided so that special-purpose external code can be used for this case. When
274 used, these functions are always called in a stack-like manner (last obtained,
275 first freed), and always for memory blocks of the same size. There is a
276 discussion about PCRE's stack usage in the
277 .\" HREF
278 \fBpcrestack\fP
279 .\"
280 documentation.
281 .P
282 The global variable \fBpcre_callout\fP initially contains NULL. It can be set
283 by the caller to a "callout" function, which PCRE will then call at specified
284 points during a matching operation. Details are given in the
285 .\" HREF
286 \fBpcrecallout\fP
287 .\"
288 documentation.
289 .
290 .
291 .\" HTML <a name="newlines"></a>
292 .SH NEWLINES
293 .rs
294 .sp
295 PCRE supports five different conventions for indicating line breaks in
296 strings: a single CR (carriage return) character, a single LF (linefeed)
297 character, the two-character sequence CRLF, any of the three preceding, or any
298 Unicode newline sequence. The Unicode newline sequences are the three just
299 mentioned, plus the single characters VT (vertical tab, U+000B), FF (form feed,
300 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
301 (paragraph separator, U+2029).
302 .P
303 Each of the first three conventions is used by at least one operating system as
304 its standard newline sequence. When PCRE is built, a default can be specified.
305 The default default is LF, which is the Unix standard. When PCRE is run, the
306 default can be overridden, either when a pattern is compiled, or when it is
307 matched.
308 .P
309 At compile time, the newline convention can be specified by the \fIoptions\fP
310 argument of \fBpcre_compile()\fP, or it can be specified by special text at the
311 start of the pattern itself; this overrides any other settings. See the
312 .\" HREF
313 \fBpcrepattern\fP
314 .\"
315 page for details of the special character sequences.
316 .P
317 In the PCRE documentation the word "newline" is used to mean "the character or
318 pair of characters that indicate a line break". The choice of newline
319 convention affects the handling of the dot, circumflex, and dollar
320 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
321 recognized line ending sequence, the match position advancement for a
322 non-anchored pattern. There is more detail about this in the
323 .\" HTML <a href="#execoptions">
324 .\" </a>
325 section on \fBpcre_exec()\fP options
326 .\"
327 below.
328 .P
329 The choice of newline convention does not affect the interpretation of
330 the \en or \er escape sequences, nor does it affect what \eR matches, which is
331 controlled in a similar way, but by separate options.
332 .
333 .
334 .SH MULTITHREADING
335 .rs
336 .sp
337 The PCRE functions can be used in multi-threading applications, with the
338 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
339 \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
340 callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
341 .P
342 The compiled form of a regular expression is not altered during matching, so
343 the same compiled pattern can safely be used by several threads at once.
344 .P
345 If the just-in-time optimization feature is being used, it needs separate
346 memory stack areas for each thread. See the
347 .\" HREF
348 \fBpcrejit\fP
349 .\"
350 documentation for more details.
351 .
352 .
353 .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
354 .rs
355 .sp
356 The compiled form of a regular expression can be saved and re-used at a later
357 time, possibly by a different program, and even on a host other than the one on
358 which it was compiled. Details are given in the
359 .\" HREF
360 \fBpcreprecompile\fP
361 .\"
362 documentation, which includes a description of the
363 \fBpcre_pattern_to_host_byte_order()\fP function. However, compiling a regular
364 expression with one version of PCRE for use with a different version is not
365 guaranteed to work and may cause crashes.
366 .
367 .
368 .SH "CHECKING BUILD-TIME OPTIONS"
369 .rs
370 .sp
371 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
372 .PP
373 The function \fBpcre_config()\fP makes it possible for a PCRE client to
374 discover which optional features have been compiled into the PCRE library. The
375 .\" HREF
376 \fBpcrebuild\fP
377 .\"
378 documentation has more details about these optional features.
379 .P
380 The first argument for \fBpcre_config()\fP is an integer, specifying which
381 information is required; the second argument is a pointer to a variable into
382 which the information is placed. The returned value is zero on success, or the
383 negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
384 not recognized. The following information is available:
385 .sp
386 PCRE_CONFIG_UTF8
387 .sp
388 The output is an integer that is set to one if UTF-8 support is available;
389 otherwise it is set to zero. This value should normally be given to the 8-bit
390 version of this function, \fBpcre_config()\fP. If it is given to the 16-bit
391 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
392 .sp
393 PCRE_CONFIG_UTF16
394 .sp
395 The output is an integer that is set to one if UTF-16 support is available;
396 otherwise it is set to zero. This value should normally be given to the 16-bit
397 version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
398 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
399 .sp
400 PCRE_CONFIG_UTF32
401 .sp
402 The output is an integer that is set to one if UTF-32 support is available;
403 otherwise it is set to zero. This value should normally be given to the 32-bit
404 version of this function, \fBpcre32_config()\fP. If it is given to the 8-bit
405 or 16-bit version of this function, the result is PCRE_ERROR_BADOPTION.
406 .sp
407 PCRE_CONFIG_UNICODE_PROPERTIES
408 .sp
409 The output is an integer that is set to one if support for Unicode character
410 properties is available; otherwise it is set to zero.
411 .sp
412 PCRE_CONFIG_JIT
413 .sp
414 The output is an integer that is set to one if support for just-in-time
415 compiling is available; otherwise it is set to zero.
416 .sp
417 PCRE_CONFIG_JITTARGET
418 .sp
419 The output is a pointer to a zero-terminated "const char *" string. If JIT
420 support is available, the string contains the name of the architecture for
421 which the JIT compiler is configured, for example "x86 32bit (little endian +
422 unaligned)". If JIT support is not available, the result is NULL.
423 .sp
424 PCRE_CONFIG_NEWLINE
425 .sp
426 The output is an integer whose value specifies the default character sequence
427 that is recognized as meaning "newline". The values that are supported in
428 ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for
429 ANYCRLF, and -1 for ANY. In EBCDIC environments, CR, ANYCRLF, and ANY yield the
430 same values. However, the value for LF is normally 21, though some EBCDIC
431 environments use 37. The corresponding values for CRLF are 3349 and 3365. The
432 default should normally correspond to the standard sequence for your operating
433 system.
434 .sp
435 PCRE_CONFIG_BSR
436 .sp
437 The output is an integer whose value indicates what character sequences the \eR
438 escape sequence matches by default. A value of 0 means that \eR matches any
439 Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
440 or CRLF. The default can be overridden when a pattern is compiled or matched.
441 .sp
442 PCRE_CONFIG_LINK_SIZE
443 .sp
444 The output is an integer that contains the number of bytes used for internal
445 linkage in compiled regular expressions. For the 8-bit library, the value can
446 be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
447 a number of bytes. For the 32-bit library, the value is either 2 or 4 and is
448 still a number of bytes. The default value of 2 is sufficient for all but the
449 most massive patterns, since it allows the compiled pattern to be up to 64K in
450 size. Larger values allow larger regular expressions to be compiled, at the
451 expense of slower matching.
452 .sp
453 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
454 .sp
455 The output is an integer that contains the threshold above which the POSIX
456 interface uses \fBmalloc()\fP for output vectors. Further details are given in
457 the
458 .\" HREF
459 \fBpcreposix\fP
460 .\"
461 documentation.
462 .sp
463 PCRE_CONFIG_MATCH_LIMIT
464 .sp
465 The output is a long integer that gives the default limit for the number of
466 internal matching function calls in a \fBpcre_exec()\fP execution. Further
467 details are given with \fBpcre_exec()\fP below.
468 .sp
469 PCRE_CONFIG_MATCH_LIMIT_RECURSION
470 .sp
471 The output is a long integer that gives the default limit for the depth of
472 recursion when calling the internal matching function in a \fBpcre_exec()\fP
473 execution. Further details are given with \fBpcre_exec()\fP below.
474 .sp
475 PCRE_CONFIG_STACKRECURSE
476 .sp
477 The output is an integer that is set to one if internal recursion when running
478 \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
479 to remember their state. This is the usual way that PCRE is compiled. The
480 output is zero if PCRE was compiled to use blocks of data on the heap instead
481 of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
482 \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
483 avoiding the use of the stack.
484 .
485 .
486 .SH "COMPILING A PATTERN"
487 .rs
488 .sp
489 .nf
490 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
491 .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
492 .B " const unsigned char *\fItableptr\fP);"
493 .sp
494 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
495 .B " int *\fIerrorcodeptr\fP,"
496 .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
497 .B " const unsigned char *\fItableptr\fP);"
498 .fi
499 .P
500 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
501 called to compile a pattern into an internal form. The only difference between
502 the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
503 \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
504 too much repetition, we refer just to \fBpcre_compile()\fP below, but the
505 information applies equally to \fBpcre_compile2()\fP.
506 .P
507 The pattern is a C string terminated by a binary zero, and is passed in the
508 \fIpattern\fP argument. A pointer to a single block of memory that is obtained
509 via \fBpcre_malloc\fP is returned. This contains the compiled code and related
510 data. The \fBpcre\fP type is defined for the returned block; this is a typedef
511 for a structure whose contents are not externally defined. It is up to the
512 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
513 .P
514 Although the compiled code of a PCRE regex is relocatable, that is, it does not
515 depend on memory location, the complete \fBpcre\fP data block is not
516 fully relocatable, because it may contain a copy of the \fItableptr\fP
517 argument, which is an address (see below).
518 .P
519 The \fIoptions\fP argument contains various bit settings that affect the
520 compilation. It should be zero if no options are required. The available
521 options are described below. Some of them (in particular, those that are
522 compatible with Perl, but some others as well) can also be set and unset from
523 within the pattern (see the detailed description in the
524 .\" HREF
525 \fBpcrepattern\fP
526 .\"
527 documentation). For those options that can be different in different parts of
528 the pattern, the contents of the \fIoptions\fP argument specifies their
529 settings at the start of compilation and execution. The PCRE_ANCHORED,
530 PCRE_BSR_\fIxxx\fP, PCRE_NEWLINE_\fIxxx\fP, PCRE_NO_UTF8_CHECK, and
531 PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
532 compile time.
533 .P
534 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
535 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
536 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
537 error message. This is a static string that is part of the library. You must
538 not try to free it. Normally, the offset from the start of the pattern to the
539 data unit that was being processed when the error was discovered is placed in
540 the variable pointed to by \fIerroffset\fP, which must not be NULL (if it is,
541 an immediate error is given). However, for an invalid UTF-8 or UTF-16 string,
542 the offset is that of the first data unit of the failing character.
543 .P
544 Some errors are not detected until the whole pattern has been scanned; in these
545 cases, the offset passed back is the length of the pattern. Note that the
546 offset is in data units, not characters, even in a UTF mode. It may sometimes
547 point into the middle of a UTF-8 or UTF-16 character.
548 .P
549 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
550 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
551 returned via this argument in the event of an error. This is in addition to the
552 textual error message. Error codes and messages are listed below.
553 .P
554 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
555 character tables that are built when PCRE is compiled, using the default C
556 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
557 call to \fBpcre_maketables()\fP. This value is stored with the compiled
558 pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
559 passed to it. For more discussion, see the section on locale support below.
560 .P
561 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
562 .sp
563 pcre *re;
564 const char *error;
565 int erroffset;
566 re = pcre_compile(
567 "^A.*Z", /* the pattern */
568 0, /* default options */
569 &error, /* for error message */
570 &erroffset, /* for error offset */
571 NULL); /* use default character tables */
572 .sp
573 The following names for option bits are defined in the \fBpcre.h\fP header
574 file:
575 .sp
576 PCRE_ANCHORED
577 .sp
578 If this bit is set, the pattern is forced to be "anchored", that is, it is
579 constrained to match only at the first matching point in the string that is
580 being searched (the "subject string"). This effect can also be achieved by
581 appropriate constructs in the pattern itself, which is the only way to do it in
582 Perl.
583 .sp
584 PCRE_AUTO_CALLOUT
585 .sp
586 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
587 all with number 255, before each pattern item. For discussion of the callout
588 facility, see the
589 .\" HREF
590 \fBpcrecallout\fP
591 .\"
592 documentation.
593 .sp
594 PCRE_BSR_ANYCRLF
595 PCRE_BSR_UNICODE
596 .sp
597 These options (which are mutually exclusive) control what the \eR escape
598 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
599 match any Unicode newline sequence. The default is specified when PCRE is
600 built. It can be overridden from within the pattern, or by setting an option
601 when a compiled pattern is matched.
602 .sp
603 PCRE_CASELESS
604 .sp
605 If this bit is set, letters in the pattern match both upper and lower case
606 letters. It is equivalent to Perl's /i option, and it can be changed within a
607 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
608 concept of case for characters whose values are less than 128, so caseless
609 matching is always possible. For characters with higher values, the concept of
610 case is supported if PCRE is compiled with Unicode property support, but not
611 otherwise. If you want to use caseless matching for characters 128 and above,
612 you must ensure that PCRE is compiled with Unicode property support as well as
613 with UTF-8 support.
614 .sp
615 PCRE_DOLLAR_ENDONLY
616 .sp
617 If this bit is set, a dollar metacharacter in the pattern matches only at the
618 end of the subject string. Without this option, a dollar also matches
619 immediately before a newline at the end of the string (but not before any other
620 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
621 There is no equivalent to this option in Perl, and no way to set it within a
622 pattern.
623 .sp
624 PCRE_DOTALL
625 .sp
626 If this bit is set, a dot metacharacter in the pattern matches a character of
627 any value, including one that indicates a newline. However, it only ever
628 matches one character, even if newlines are coded as CRLF. Without this option,
629 a dot does not match when the current position is at a newline. This option is
630 equivalent to Perl's /s option, and it can be changed within a pattern by a
631 (?s) option setting. A negative class such as [^a] always matches newline
632 characters, independent of the setting of this option.
633 .sp
634 PCRE_DUPNAMES
635 .sp
636 If this bit is set, names used to identify capturing subpatterns need not be
637 unique. This can be helpful for certain types of pattern when it is known that
638 only one instance of the named subpattern can ever be matched. There are more
639 details of named subpatterns below; see also the
640 .\" HREF
641 \fBpcrepattern\fP
642 .\"
643 documentation.
644 .sp
645 PCRE_EXTENDED
646 .sp
647 If this bit is set, white space data characters in the pattern are totally
648 ignored except when escaped or inside a character class. White space does not
649 include the VT character (code 11). In addition, characters between an
650 unescaped # outside a character class and the next newline, inclusive, are also
651 ignored. This is equivalent to Perl's /x option, and it can be changed within a
652 pattern by a (?x) option setting.
653 .P
654 Which characters are interpreted as newlines is controlled by the options
655 passed to \fBpcre_compile()\fP or by a special sequence at the start of the
656 pattern, as described in the section entitled
657 .\" HTML <a href="pcrepattern.html#newlines">
658 .\" </a>
659 "Newline conventions"
660 .\"
661 in the \fBpcrepattern\fP documentation. Note that the end of this type of
662 comment is a literal newline sequence in the pattern; escape sequences that
663 happen to represent a newline do not count.
664 .P
665 This option makes it possible to include comments inside complicated patterns.
666 Note, however, that this applies only to data characters. White space characters
667 may never appear within special character sequences in a pattern, for example
668 within the sequence (?( that introduces a conditional subpattern.
669 .sp
670 PCRE_EXTRA
671 .sp
672 This option was invented in order to turn on additional functionality of PCRE
673 that is incompatible with Perl, but it is currently of very little use. When
674 set, any backslash in a pattern that is followed by a letter that has no
675 special meaning causes an error, thus reserving these combinations for future
676 expansion. By default, as in Perl, a backslash followed by a letter with no
677 special meaning is treated as a literal. (Perl can, however, be persuaded to
678 give an error for this, by running it with the -w option.) There are at present
679 no other features controlled by this option. It can also be set by a (?X)
680 option setting within a pattern.
681 .sp
682 PCRE_FIRSTLINE
683 .sp
684 If this option is set, an unanchored pattern is required to match before or at
685 the first newline in the subject string, though the matched text may continue
686 over the newline.
687 .sp
688 PCRE_JAVASCRIPT_COMPAT
689 .sp
690 If this option is set, PCRE's behaviour is changed in some ways so that it is
691 compatible with JavaScript rather than Perl. The changes are as follows:
692 .P
693 (1) A lone closing square bracket in a pattern causes a compile-time error,
694 because this is illegal in JavaScript (by default it is treated as a data
695 character). Thus, the pattern AB]CD becomes illegal when this option is set.
696 .P
697 (2) At run time, a back reference to an unset subpattern group matches an empty
698 string (by default this causes the current matching alternative to fail). A
699 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
700 an "a" in the subject), whereas it fails by default, for Perl compatibility.
701 .P
702 (3) \eU matches an upper case "U" character; by default \eU causes a compile
703 time error (Perl uses \eU to upper case subsequent characters).
704 .P
705 (4) \eu matches a lower case "u" character unless it is followed by four
706 hexadecimal digits, in which case the hexadecimal number defines the code point
707 to match. By default, \eu causes a compile time error (Perl uses it to upper
708 case the following character).
709 .P
710 (5) \ex matches a lower case "x" character unless it is followed by two
711 hexadecimal digits, in which case the hexadecimal number defines the code point
712 to match. By default, as in Perl, a hexadecimal number is always expected after
713 \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
714 binary zero character followed by z).
715 .sp
716 PCRE_MULTILINE
717 .sp
718 By default, for the purposes of matching "start of line" and "end of line",
719 PCRE treats the subject string as consisting of a single line of characters,
720 even if it actually contains newlines. The "start of line" metacharacter (^)
721 matches only at the start of the string, and the "end of line" metacharacter
722 ($) matches only at the end of the string, or before a terminating newline
723 (except when PCRE_DOLLAR_ENDONLY is set). Note, however, that unless
724 PCRE_DOTALL is set, the "any character" metacharacter (.) does not match at a
725 newline. This behaviour (for ^, $, and dot) is the same as Perl.
726 .P
727 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
728 match immediately following or immediately before internal newlines in the
729 subject string, respectively, as well as at the very start and end. This is
730 equivalent to Perl's /m option, and it can be changed within a pattern by a
731 (?m) option setting. If there are no newlines in a subject string, or no
732 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
733 .sp
734 PCRE_NEVER_UTF
735 .sp
736 This option locks out interpretation of the pattern as UTF-8 (or UTF-16 or
737 UTF-32 in the 16-bit and 32-bit libraries). In particular, it prevents the
738 creator of the pattern from switching to UTF interpretation by starting the
739 pattern with (*UTF). This may be useful in applications that process patterns
740 from external sources. The combination of PCRE_UTF8 and PCRE_NEVER_UTF also
741 causes an error.
742 .sp
743 PCRE_NEWLINE_CR
744 PCRE_NEWLINE_LF
745 PCRE_NEWLINE_CRLF
746 PCRE_NEWLINE_ANYCRLF
747 PCRE_NEWLINE_ANY
748 .sp
749 These options override the default newline definition that was chosen when PCRE
750 was built. Setting the first or the second specifies that a newline is
751 indicated by a single character (CR or LF, respectively). Setting
752 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
753 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
754 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
755 that any Unicode newline sequence should be recognized.
756 .P
757 In an ASCII/Unicode environment, the Unicode newline sequences are the three
758 just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
759 feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
760 (paragraph separator, U+2029). For the 8-bit library, the last two are
761 recognized only in UTF-8 mode.
762 .P
763 When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
764 CR is 0x0d, the same as ASCII. However, the character code for LF is normally
765 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
766 not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
767 less than 256. For more details, see the
768 .\" HREF
769 \fBpcrebuild\fP
770 .\"
771 documentation.
772 .P
773 The newline setting in the options word uses three bits that are treated
774 as a number, giving eight possibilities. Currently only six are used (default
775 plus the five values above). This means that if you set more than one newline
776 option, the combination may or may not be sensible. For example,
777 PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
778 other combinations may yield unused numbers and cause an error.
779 .P
780 The only time that a line break in a pattern is specially recognized when
781 compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
782 and so are ignored in this mode. Also, an unescaped # outside a character class
783 indicates a comment that lasts until after the next line break sequence. In
784 other circumstances, line break sequences in patterns are treated as literal
785 data.
786 .P
787 The newline option that is set at compile time becomes the default that is used
788 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
789 .sp
790 PCRE_NO_AUTO_CAPTURE
791 .sp
792 If this option is set, it disables the use of numbered capturing parentheses in
793 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
794 were followed by ?: but named parentheses can still be used for capturing (and
795 they acquire numbers in the usual way). There is no equivalent of this option
796 in Perl.
797 .sp
798 PCRE_NO_AUTO_POSSESSIFY
799 .sp
800 If this option is set, it disables "auto-possessification". This is an
801 optimization that, for example, turns a+b into a++b in order to avoid
802 backtracks into a+ that can never be successful.
803 .sp
804 PCRE_NO_START_OPTIMIZE
805 .sp
806 This is an option that acts at matching time; that is, it is really an option
807 for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
808 it is remembered with the compiled pattern and assumed at matching time. This
809 is necessary if you want to use JIT execution, because the JIT compiler needs
810 to know whether or not this option is set. For details see the discussion of
811 PCRE_NO_START_OPTIMIZE
812 .\" HTML <a href="#execoptions">
813 .\" </a>
814 below.
815 .\"
816 .sp
817 PCRE_UCP
818 .sp
819 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
820 \ew, and some of the POSIX character classes. By default, only ASCII characters
821 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
822 classify characters. More details are given in the section on
823 .\" HTML <a href="pcre.html#genericchartypes">
824 .\" </a>
825 generic character types
826 .\"
827 in the
828 .\" HREF
829 \fBpcrepattern\fP
830 .\"
831 page. If you set PCRE_UCP, matching one of the items it affects takes much
832 longer. The option is available only if PCRE has been compiled with Unicode
833 property support.
834 .sp
835 PCRE_UNGREEDY
836 .sp
837 This option inverts the "greediness" of the quantifiers so that they are not
838 greedy by default, but become greedy if followed by "?". It is not compatible
839 with Perl. It can also be set by a (?U) option setting within the pattern.
840 .sp
841 PCRE_UTF8
842 .sp
843 This option causes PCRE to regard both the pattern and the subject as strings
844 of UTF-8 characters instead of single-byte strings. However, it is available
845 only when PCRE is built to include UTF support. If not, the use of this option
846 provokes an error. Details of how this option changes the behaviour of PCRE are
847 given in the
848 .\" HREF
849 \fBpcreunicode\fP
850 .\"
851 page.
852 .sp
853 PCRE_NO_UTF8_CHECK
854 .sp
855 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
856 automatically checked. There is a discussion about the
857 .\" HTML <a href="pcreunicode.html#utf8strings">
858 .\" </a>
859 validity of UTF-8 strings
860 .\"
861 in the
862 .\" HREF
863 \fBpcreunicode\fP
864 .\"
865 page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
866 error. If you already know that your pattern is valid, and you want to skip
867 this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
868 When it is set, the effect of passing an invalid UTF-8 string as a pattern is
869 undefined. It may cause your program to crash or loop. Note that this option
870 can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
871 the validity checking of subject strings only. If the same string is being
872 matched many times, the option can be safely set for the second and subsequent
873 matchings to improve performance.
874 .
875 .
876 .SH "COMPILATION ERROR CODES"
877 .rs
878 .sp
879 The following table lists the error codes than may be returned by
880 \fBpcre_compile2()\fP, along with the error messages that may be returned by
881 both compiling functions. Note that error messages are always 8-bit ASCII
882 strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
883 have fallen out of use. To avoid confusion, they have not been re-used.
884 .sp
885 0 no error
886 1 \e at end of pattern
887 2 \ec at end of pattern
888 3 unrecognized character follows \e
889 4 numbers out of order in {} quantifier
890 5 number too big in {} quantifier
891 6 missing terminating ] for character class
892 7 invalid escape sequence in character class
893 8 range out of order in character class
894 9 nothing to repeat
895 10 [this code is not in use]
896 11 internal error: unexpected repeat
897 12 unrecognized character after (? or (?-
898 13 POSIX named classes are supported only within a class
899 14 missing )
900 15 reference to non-existent subpattern
901 16 erroffset passed as NULL
902 17 unknown option bit(s) set
903 18 missing ) after comment
904 19 [this code is not in use]
905 20 regular expression is too large
906 21 failed to get memory
907 22 unmatched parentheses
908 23 internal error: code overflow
909 24 unrecognized character after (?<
910 25 lookbehind assertion is not fixed length
911 26 malformed number or name after (?(
912 27 conditional group contains more than two branches
913 28 assertion expected after (?(
914 29 (?R or (?[+-]digits must be followed by )
915 30 unknown POSIX class name
916 31 POSIX collating elements are not supported
917 32 this version of PCRE is compiled without UTF support
918 33 [this code is not in use]
919 34 character value in \ex{...} sequence is too large
920 35 invalid condition (?(0)
921 36 \eC not allowed in lookbehind assertion
922 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
923 38 number after (?C is > 255
924 39 closing ) for (?C expected
925 40 recursive call could loop indefinitely
926 41 unrecognized character after (?P
927 42 syntax error in subpattern name (missing terminator)
928 43 two named subpatterns have the same name
929 44 invalid UTF-8 string (specifically UTF-8)
930 45 support for \eP, \ep, and \eX has not been compiled
931 46 malformed \eP or \ep sequence
932 47 unknown property name after \eP or \ep
933 48 subpattern name is too long (maximum 32 characters)
934 49 too many named subpatterns (maximum 10000)
935 50 [this code is not in use]
936 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
937 52 internal error: overran compiling workspace
938 53 internal error: previously-checked referenced subpattern
939 not found
940 54 DEFINE group contains more than one branch
941 55 repeating a DEFINE group is not allowed
942 56 inconsistent NEWLINE options
943 57 \eg is not followed by a braced, angle-bracketed, or quoted
944 name/number or by a plain number
945 58 a numbered reference must not be zero
946 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
947 60 (*VERB) not recognized or malformed
948 61 number is too big
949 62 subpattern name expected
950 63 digit expected after (?+
951 64 ] is an invalid data character in JavaScript compatibility mode
952 65 different names for subpatterns of the same number are
953 not allowed
954 66 (*MARK) must have an argument
955 67 this version of PCRE is not compiled with Unicode property
956 support
957 68 \ec must be followed by an ASCII character
958 69 \ek is not followed by a braced, angle-bracketed, or quoted name
959 70 internal error: unknown opcode in find_fixedlength()
960 71 \eN is not supported in a class
961 72 too many forward references
962 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
963 74 invalid UTF-16 string (specifically UTF-16)
964 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
965 76 character value in \eu.... sequence is too large
966 77 invalid UTF-32 string (specifically UTF-32)
967 .sp
968 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
969 be used if the limits were changed when PCRE was built.
970 .
971 .
972 .\" HTML <a name="studyingapattern"></a>
973 .SH "STUDYING A PATTERN"
974 .rs
975 .sp
976 .nf
977 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
978 .B " const char **\fIerrptr\fP);"
979 .fi
980 .PP
981 If a compiled pattern is going to be used several times, it is worth spending
982 more time analyzing it in order to speed up the time taken for matching. The
983 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
984 argument. If studying the pattern produces additional information that will
985 help speed up matching, \fBpcre_study()\fP returns a pointer to a
986 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
987 results of the study.
988 .P
989 The returned value from \fBpcre_study()\fP can be passed directly to
990 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
991 also contains other fields that can be set by the caller before the block is
992 passed; these are described
993 .\" HTML <a href="#extradata">
994 .\" </a>
995 below
996 .\"
997 in the section on matching a pattern.
998 .P
999 If studying the pattern does not produce any useful information,
1000 \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
1001 calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
1002 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
1003 if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
1004 returns a \fBpcre_extra\fP block even if studying did not find any additional
1005 information. It may still return NULL, however, if an error occurs in
1006 \fBpcre_study()\fP.
1007 .P
1008 The second argument of \fBpcre_study()\fP contains option bits. There are three
1009 further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1010 .sp
1011 PCRE_STUDY_JIT_COMPILE
1012 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1013 PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1014 .sp
1015 If any of these are set, and the just-in-time compiler is available, the
1016 pattern is further compiled into machine code that executes much faster than
1017 the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
1018 compiler is not available, these options are ignored. All undefined bits in the
1019 \fIoptions\fP argument must be zero.
1020 .P
1021 JIT compilation is a heavyweight optimization. It can take some time for
1022 patterns to be analyzed, and for one-off matches and simple patterns the
1023 benefit of faster execution might be offset by a much slower study time.
1024 Not all patterns can be optimized by the JIT compiler. For those that cannot be
1025 handled, matching automatically falls back to the \fBpcre_exec()\fP
1026 interpreter. For more details, see the
1027 .\" HREF
1028 \fBpcrejit\fP
1029 .\"
1030 documentation.
1031 .P
1032 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
1033 studying succeeds (even if no data is returned), the variable it points to is
1034 set to NULL. Otherwise it is set to point to a textual error message. This is a
1035 static string that is part of the library. You must not try to free it. You
1036 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
1037 sure that it has run successfully.
1038 .P
1039 When you are finished with a pattern, you can free the memory used for the
1040 study data by calling \fBpcre_free_study()\fP. This function was added to the
1041 API for release 8.20. For earlier versions, the memory could be freed with
1042 \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1043 where JIT optimization is not used, but it is advisable to change to the new
1044 function when convenient.
1045 .P
1046 This is a typical way in which \fBpcre_study\fP() is used (except that in a
1047 real application there should be tests for errors):
1048 .sp
1049 int rc;
1050 pcre *re;
1051 pcre_extra *sd;
1052 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1053 sd = pcre_study(
1054 re, /* result of pcre_compile() */
1055 0, /* no options */
1056 &error); /* set to NULL or points to a message */
1057 rc = pcre_exec( /* see below for details of pcre_exec() options */
1058 re, sd, "subject", 7, 0, 0, ovector, 30);
1059 ...
1060 pcre_free_study(sd);
1061 pcre_free(re);
1062 .sp
1063 Studying a pattern does two things: first, a lower bound for the length of
1064 subject string that is needed to match the pattern is computed. This does not
1065 mean that there are any strings of that length that match, but it does
1066 guarantee that no shorter strings match. The value is used to avoid wasting
1067 time by trying to match strings that are shorter than the lower bound. You can
1068 find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
1069 .P
1070 Studying a pattern is also useful for non-anchored patterns that do not have a
1071 single fixed starting character. A bitmap of possible starting bytes is
1072 created. This speeds up finding a position in the subject at which to start
1073 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1074 In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1075 .P
1076 These two optimizations apply to both \fBpcre_exec()\fP and
1077 \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1078 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option.
1079 You might want to do this if your pattern contains callouts or (*MARK) and you
1080 want to make use of these facilities in cases where matching fails.
1081 .P
1082 PCRE_NO_START_OPTIMIZE can be specified at either compile time or execution
1083 time. However, if PCRE_NO_START_OPTIMIZE is passed to \fBpcre_exec()\fP, (that
1084 is, after any JIT compilation has happened) JIT execution is disabled. For JIT
1085 execution to work with PCRE_NO_START_OPTIMIZE, the option must be set at
1086 compile time.
1087 .P
1088 There is a longer discussion of PCRE_NO_START_OPTIMIZE
1089 .\" HTML <a href="#execoptions">
1090 .\" </a>
1091 below.
1092 .\"
1093 .
1094 .
1095 .\" HTML <a name="localesupport"></a>
1096 .SH "LOCALE SUPPORT"
1097 .rs
1098 .sp
1099 PCRE handles caseless matching, and determines whether characters are letters,
1100 digits, or whatever, by reference to a set of tables, indexed by character
1101 value. When running in UTF-8 mode, this applies only to characters
1102 with codes less than 128. By default, higher-valued codes never match escapes
1103 such as \ew or \ed, but they can be tested with \ep if PCRE is built with
1104 Unicode character property support. Alternatively, the PCRE_UCP option can be
1105 set at compile time; this causes \ew and friends to use Unicode property
1106 support instead of built-in tables. The use of locales with Unicode is
1107 discouraged. If you are handling characters with codes greater than 128, you
1108 should either use UTF-8 and Unicode, or use locales, but not try to mix the
1109 two.
1110 .P
1111 PCRE contains an internal set of tables that are used when the final argument
1112 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1113 Normally, the internal tables recognize only ASCII characters. However, when
1114 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1115 default "C" locale of the local system, which may cause them to be different.
1116 .P
1117 The internal tables can always be overridden by tables supplied by the
1118 application that calls PCRE. These may be created in a different locale from
1119 the default. As more and more applications change to using Unicode, the need
1120 for this locale support is expected to die away.
1121 .P
1122 External tables are built by calling the \fBpcre_maketables()\fP function,
1123 which has no arguments, in the relevant locale. The result can then be passed
1124 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
1125 example, to build and use tables that are appropriate for the French locale
1126 (where accented characters with values greater than 128 are treated as letters),
1127 the following code could be used:
1128 .sp
1129 setlocale(LC_CTYPE, "fr_FR");
1130 tables = pcre_maketables();
1131 re = pcre_compile(..., tables);
1132 .sp
1133 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1134 are using Windows, the name for the French locale is "french".
1135 .P
1136 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1137 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1138 that the memory containing the tables remains available for as long as it is
1139 needed.
1140 .P
1141 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1142 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1143 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
1144 pattern, compilation, studying and matching all happen in the same locale, but
1145 different patterns can be compiled in different locales.
1146 .P
1147 It is possible to pass a table pointer or NULL (indicating the use of the
1148 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
1149 this facility could be used to match a pattern in a different locale from the
1150 one in which it was compiled. Passing table pointers at run time is discussed
1151 below in the section on matching a pattern.
1152 .
1153 .
1154 .\" HTML <a name="infoaboutpattern"></a>
1155 .SH "INFORMATION ABOUT A PATTERN"
1156 .rs
1157 .sp
1158 .nf
1159 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1160 .B " int \fIwhat\fP, void *\fIwhere\fP);"
1161 .fi
1162 .PP
1163 The \fBpcre_fullinfo()\fP function returns information about a compiled
1164 pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1165 library at version 8.30, after more than 10 years of obsolescence.
1166 .P
1167 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1168 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1169 the pattern was not studied. The third argument specifies which piece of
1170 information is required, and the fourth argument is a pointer to a variable
1171 to receive the data. The yield of the function is zero for success, or one of
1172 the following negative numbers:
1173 .sp
1174 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1175 the argument \fIwhere\fP was NULL
1176 PCRE_ERROR_BADMAGIC the "magic number" was not found
1177 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1178 endianness
1179 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1180 PCRE_ERROR_UNSET the requested field is not set
1181 .sp
1182 The "magic number" is placed at the start of each compiled pattern as an simple
1183 check against passing an arbitrary memory pointer. The endianness error can
1184 occur if a compiled pattern is saved and reloaded on a different host. Here is
1185 a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1186 pattern:
1187 .sp
1188 int rc;
1189 size_t length;
1190 rc = pcre_fullinfo(
1191 re, /* result of pcre_compile() */
1192 sd, /* result of pcre_study(), or NULL */
1193 PCRE_INFO_SIZE, /* what is required */
1194 &length); /* where to put the data */
1195 .sp
1196 The possible values for the third argument are defined in \fBpcre.h\fP, and are
1197 as follows:
1198 .sp
1199 PCRE_INFO_BACKREFMAX
1200 .sp
1201 Return the number of the highest back reference in the pattern. The fourth
1202 argument should point to an \fBint\fP variable. Zero is returned if there are
1203 no back references.
1204 .sp
1205 PCRE_INFO_CAPTURECOUNT
1206 .sp
1207 Return the number of capturing subpatterns in the pattern. The fourth argument
1208 should point to an \fBint\fP variable.
1209 .sp
1210 PCRE_INFO_DEFAULT_TABLES
1211 .sp
1212 Return a pointer to the internal default character tables within PCRE. The
1213 fourth argument should point to an \fBunsigned char *\fP variable. This
1214 information call is provided for internal use by the \fBpcre_study()\fP
1215 function. External callers can cause PCRE to use its internal tables by passing
1216 a NULL table pointer.
1217 .sp
1218 PCRE_INFO_FIRSTBYTE
1219 .sp
1220 Return information about the first data unit of any matched string, for a
1221 non-anchored pattern. (The name of this option refers to the 8-bit library,
1222 where data units are bytes.) The fourth argument should point to an \fBint\fP
1223 variable.
1224 .P
1225 If there is a fixed first value, for example, the letter "c" from a pattern
1226 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1227 value is always less than 256. In the 16-bit library the value can be up to
1228 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1229 .P
1230 If there is no fixed first value, and if either
1231 .sp
1232 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1233 starts with "^", or
1234 .sp
1235 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1236 (if it were set, the pattern would be anchored),
1237 .sp
1238 -1 is returned, indicating that the pattern matches only at the start of a
1239 subject string or after any newline within the string. Otherwise -2 is
1240 returned. For anchored patterns, -2 is returned.
1241 .P
1242 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1243 to return the full 32-bit range of the character, this value is deprecated;
1244 instead the PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER values
1245 should be used.
1246 .sp
1247 PCRE_INFO_FIRSTTABLE
1248 .sp
1249 If the pattern was studied, and this resulted in the construction of a 256-bit
1250 table indicating a fixed set of values for the first data unit in any matching
1251 string, a pointer to the table is returned. Otherwise NULL is returned. The
1252 fourth argument should point to an \fBunsigned char *\fP variable.
1253 .sp
1254 PCRE_INFO_HASCRORLF
1255 .sp
1256 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1257 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1258 explicit match is either a literal CR or LF character, or \er or \en.
1259 .sp
1260 PCRE_INFO_JCHANGED
1261 .sp
1262 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1263 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1264 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1265 .sp
1266 PCRE_INFO_JIT
1267 .sp
1268 Return 1 if the pattern was studied with one of the JIT options, and
1269 just-in-time compiling was successful. The fourth argument should point to an
1270 \fBint\fP variable. A return value of 0 means that JIT support is not available
1271 in this version of PCRE, or that the pattern was not studied with a JIT option,
1272 or that the JIT compiler could not handle this particular pattern. See the
1273 .\" HREF
1274 \fBpcrejit\fP
1275 .\"
1276 documentation for details of what can and cannot be handled.
1277 .sp
1278 PCRE_INFO_JITSIZE
1279 .sp
1280 If the pattern was successfully studied with a JIT option, return the size of
1281 the JIT compiled code, otherwise return zero. The fourth argument should point
1282 to a \fBsize_t\fP variable.
1283 .sp
1284 PCRE_INFO_LASTLITERAL
1285 .sp
1286 Return the value of the rightmost literal data unit that must exist in any
1287 matched string, other than at its start, if such a value has been recorded. The
1288 fourth argument should point to an \fBint\fP variable. If there is no such
1289 value, -1 is returned. For anchored patterns, a last literal value is recorded
1290 only if it follows something of variable length. For example, for the pattern
1291 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1292 is -1.
1293 .P
1294 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1295 to return the full 32-bit range of characters, this value is deprecated;
1296 instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1297 be used.
1298 .sp
1299 PCRE_INFO_MATCH_EMPTY
1300 .sp
1301 Return 1 if the pattern can match an empty string, otherwise 0. The fourth
1302 argument should point to an \fBint\fP variable.
1303 .sp
1304 PCRE_INFO_MATCHLIMIT
1305 .sp
1306 If the pattern set a match limit by including an item of the form
1307 (*LIMIT_MATCH=nnnn) at the start, the value is returned. The fourth argument
1308 should point to an unsigned 32-bit integer. If no such value has been set, the
1309 call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1310 .sp
1311 PCRE_INFO_MAXLOOKBEHIND
1312 .sp
1313 Return the number of characters (NB not data units) in the longest lookbehind
1314 assertion in the pattern. This information is useful when doing multi-segment
1315 matching using the partial matching facilities. Note that the simple assertions
1316 \eb and \eB require a one-character lookbehind. \eA also registers a
1317 one-character lookbehind, though it does not actually inspect the previous
1318 character. This is to ensure that at least one character from the old segment
1319 is retained when a new segment is processed. Otherwise, if there are no
1320 lookbehinds in the pattern, \eA might match incorrectly at the start of a new
1321 segment.
1322 .sp
1323 PCRE_INFO_MINLENGTH
1324 .sp
1325 If the pattern was studied and a minimum length for matching subject strings
1326 was computed, its value is returned. Otherwise the returned value is -1. The
1327 value is a number of characters, which in UTF mode may be different from the
1328 number of data units. The fourth argument should point to an \fBint\fP
1329 variable. A non-negative value is a lower bound to the length of any matching
1330 string. There may not be any strings of that length that do actually match, but
1331 every string that does match is at least that long.
1332 .sp
1333 PCRE_INFO_NAMECOUNT
1334 PCRE_INFO_NAMEENTRYSIZE
1335 PCRE_INFO_NAMETABLE
1336 .sp
1337 PCRE supports the use of named as well as numbered capturing parentheses. The
1338 names are just an additional way of identifying the parentheses, which still
1339 acquire numbers. Several convenience functions such as
1340 \fBpcre_get_named_substring()\fP are provided for extracting captured
1341 substrings by name. It is also possible to extract the data directly, by first
1342 converting the name to a number in order to access the correct pointers in the
1343 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1344 you need to use the name-to-number map, which is described by these three
1345 values.
1346 .P
1347 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1348 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1349 entry; both of these return an \fBint\fP value. The entry size depends on the
1350 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1351 entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1352 the first two bytes of each entry are the number of the capturing parenthesis,
1353 most significant byte first. In the 16-bit library, the pointer points to
1354 16-bit data units, the first of which contains the parenthesis number. In the
1355 32-bit library, the pointer points to 32-bit data units, the first of which
1356 contains the parenthesis number. The rest of the entry is the corresponding
1357 name, zero terminated.
1358 .P
1359 The names are in alphabetical order. If (?| is used to create multiple groups
1360 with the same number, as described in the
1361 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1362 .\" </a>
1363 section on duplicate subpattern numbers
1364 .\"
1365 in the
1366 .\" HREF
1367 \fBpcrepattern\fP
1368 .\"
1369 page, the groups may be given the same name, but there is only one entry in the
1370 table. Different names for groups of the same number are not permitted.
1371 Duplicate names for subpatterns with different numbers are permitted,
1372 but only if PCRE_DUPNAMES is set. They appear in the table in the order in
1373 which they were found in the pattern. In the absence of (?| this is the order
1374 of increasing number; when (?| is used this is not necessarily the case because
1375 later subpatterns may have lower numbers.
1376 .P
1377 As a simple example of the name/number table, consider the following pattern
1378 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1379 space - including newlines - is ignored):
1380 .sp
1381 .\" JOIN
1382 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1383 (?<month>\ed\ed) - (?<day>\ed\ed) )
1384 .sp
1385 There are four named subpatterns, so the table has four entries, and each entry
1386 in the table is eight bytes long. The table is as follows, with non-printing
1387 bytes shows in hexadecimal, and undefined bytes shown as ??:
1388 .sp
1389 00 01 d a t e 00 ??
1390 00 05 d a y 00 ?? ??
1391 00 04 m o n t h 00
1392 00 02 y e a r 00 ??
1393 .sp
1394 When writing code to extract data from named subpatterns using the
1395 name-to-number map, remember that the length of the entries is likely to be
1396 different for each compiled pattern.
1397 .sp
1398 PCRE_INFO_OKPARTIAL
1399 .sp
1400 Return 1 if the pattern can be used for partial matching with
1401 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1402 \fBint\fP variable. From release 8.00, this always returns 1, because the
1403 restrictions that previously applied to partial matching have been lifted. The
1404 .\" HREF
1405 \fBpcrepartial\fP
1406 .\"
1407 documentation gives details of partial matching.
1408 .sp
1409 PCRE_INFO_OPTIONS
1410 .sp
1411 Return a copy of the options with which the pattern was compiled. The fourth
1412 argument should point to an \fBunsigned long int\fP variable. These option bits
1413 are those specified in the call to \fBpcre_compile()\fP, modified by any
1414 top-level option settings at the start of the pattern itself. In other words,
1415 they are the options that will be in force when matching starts. For example,
1416 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1417 result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1418 .P
1419 A pattern is automatically anchored by PCRE if all of its top-level
1420 alternatives begin with one of the following:
1421 .sp
1422 ^ unless PCRE_MULTILINE is set
1423 \eA always
1424 \eG always
1425 .\" JOIN
1426 .* if PCRE_DOTALL is set and there are no back
1427 references to the subpattern in which .* appears
1428 .sp
1429 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1430 \fBpcre_fullinfo()\fP.
1431 .sp
1432 PCRE_INFO_RECURSIONLIMIT
1433 .sp
1434 If the pattern set a recursion limit by including an item of the form
1435 (*LIMIT_RECURSION=nnnn) at the start, the value is returned. The fourth
1436 argument should point to an unsigned 32-bit integer. If no such value has been
1437 set, the call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1438 .sp
1439 PCRE_INFO_SIZE
1440 .sp
1441 Return the size of the compiled pattern in bytes (for all three libraries). The
1442 fourth argument should point to a \fBsize_t\fP variable. This value does not
1443 include the size of the \fBpcre\fP structure that is returned by
1444 \fBpcre_compile()\fP. The value that is passed as the argument to
1445 \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1446 place the compiled data is the value returned by this option plus the size of
1447 the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1448 does not alter the value returned by this option.
1449 .sp
1450 PCRE_INFO_STUDYSIZE
1451 .sp
1452 Return the size in bytes (for all three libraries) of the data block pointed to
1453 by the \fIstudy_data\fP field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP
1454 is NULL, or there is no study data, zero is returned. The fourth argument
1455 should point to a \fBsize_t\fP variable. The \fIstudy_data\fP field is set by
1456 \fBpcre_study()\fP to record information that will speed up matching (see the
1457 section entitled
1458 .\" HTML <a href="#studyingapattern">
1459 .\" </a>
1460 "Studying a pattern"
1461 .\"
1462 above). The format of the \fIstudy_data\fP block is private, but its length
1463 is made available via this option so that it can be saved and restored (see the
1464 .\" HREF
1465 \fBpcreprecompile\fP
1466 .\"
1467 documentation for details).
1468 .sp
1469 PCRE_INFO_FIRSTCHARACTERFLAGS
1470 .sp
1471 Return information about the first data unit of any matched string, for a
1472 non-anchored pattern. The fourth argument should point to an \fBint\fP
1473 variable.
1474 .P
1475 If there is a fixed first value, for example, the letter "c" from a pattern
1476 such as (cat|cow|coyote), 1 is returned, and the character value can be
1477 retrieved using PCRE_INFO_FIRSTCHARACTER.
1478 .P
1479 If there is no fixed first value, and if either
1480 .sp
1481 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1482 starts with "^", or
1483 .sp
1484 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1485 (if it were set, the pattern would be anchored),
1486 .sp
1487 2 is returned, indicating that the pattern matches only at the start of a
1488 subject string or after any newline within the string. Otherwise 0 is
1489 returned. For anchored patterns, 0 is returned.
1490 .sp
1491 PCRE_INFO_FIRSTCHARACTER
1492 .sp
1493 Return the fixed first character value in the situation where
1494 PCRE_INFO_FIRSTCHARACTERFLAGS returns 1; otherwise return 0. The fourth
1495 argument should point to an \fBuint_t\fP variable.
1496 .P
1497 In the 8-bit library, the value is always less than 256. In the 16-bit library
1498 the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1499 can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1500 .sp
1501 PCRE_INFO_REQUIREDCHARFLAGS
1502 .sp
1503 Returns 1 if there is a rightmost literal data unit that must exist in any
1504 matched string, other than at its start. The fourth argument should point to
1505 an \fBint\fP variable. If there is no such value, 0 is returned. If returning
1506 1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1507 .P
1508 For anchored patterns, a last literal value is recorded only if it follows
1509 something of variable length. For example, for the pattern /^a\ed+z\ed+/ the
1510 returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1511 /^a\edz\ed/ the returned value is 0.
1512 .sp
1513 PCRE_INFO_REQUIREDCHAR
1514 .sp
1515 Return the value of the rightmost literal data unit that must exist in any
1516 matched string, other than at its start, if such a value has been recorded. The
1517 fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1518 value, 0 is returned.
1519 .
1520 .
1521 .SH "REFERENCE COUNTS"
1522 .rs
1523 .sp
1524 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1525 .PP
1526 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1527 data block that contains a compiled pattern. It is provided for the benefit of
1528 applications that operate in an object-oriented manner, where different parts
1529 of the application may be using the same compiled pattern, but you want to free
1530 the block when they are all done.
1531 .P
1532 When a pattern is compiled, the reference count field is initialized to zero.
1533 It is changed only by calling this function, whose action is to add the
1534 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1535 function is the new value. However, the value of the count is constrained to
1536 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1537 it is forced to the appropriate limit value.
1538 .P
1539 Except when it is zero, the reference count is not correctly preserved if a
1540 pattern is compiled on one host and then transferred to a host whose byte-order
1541 is different. (This seems a highly unlikely scenario.)
1542 .
1543 .
1544 .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1545 .rs
1546 .sp
1547 .nf
1548 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1549 .B " const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1550 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);"
1551 .fi
1552 .P
1553 The function \fBpcre_exec()\fP is called to match a subject string against a
1554 compiled pattern, which is passed in the \fIcode\fP argument. If the
1555 pattern was studied, the result of the study should be passed in the
1556 \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1557 and \fIextra\fP arguments as many times as you like, in order to match
1558 different subject strings with the same pattern.
1559 .P
1560 This function is the main matching facility of the library, and it operates in
1561 a Perl-like manner. For specialist use there is also an alternative matching
1562 function, which is described
1563 .\" HTML <a href="#dfamatch">
1564 .\" </a>
1565 below
1566 .\"
1567 in the section about the \fBpcre_dfa_exec()\fP function.
1568 .P
1569 In most applications, the pattern will have been compiled (and optionally
1570 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1571 possible to save compiled patterns and study data, and then use them later
1572 in different processes, possibly even on different hosts. For a discussion
1573 about this, see the
1574 .\" HREF
1575 \fBpcreprecompile\fP
1576 .\"
1577 documentation.
1578 .P
1579 Here is an example of a simple call to \fBpcre_exec()\fP:
1580 .sp
1581 int rc;
1582 int ovector[30];
1583 rc = pcre_exec(
1584 re, /* result of pcre_compile() */
1585 NULL, /* we didn't study the pattern */
1586 "some string", /* the subject string */
1587 11, /* the length of the subject string */
1588 0, /* start at offset 0 in the subject */
1589 0, /* default options */
1590 ovector, /* vector of integers for substring information */
1591 30); /* number of elements (NOT size in bytes) */
1592 .
1593 .
1594 .\" HTML <a name="extradata"></a>
1595 .SS "Extra data for \fBpcre_exec()\fR"
1596 .rs
1597 .sp
1598 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1599 data block. The \fBpcre_study()\fP function returns such a block (when it
1600 doesn't return NULL), but you can also create one for yourself, and pass
1601 additional information in it. The \fBpcre_extra\fP block contains the following
1602 fields (not necessarily in this order):
1603 .sp
1604 unsigned long int \fIflags\fP;
1605 void *\fIstudy_data\fP;
1606 void *\fIexecutable_jit\fP;
1607 unsigned long int \fImatch_limit\fP;
1608 unsigned long int \fImatch_limit_recursion\fP;
1609 void *\fIcallout_data\fP;
1610 const unsigned char *\fItables\fP;
1611 unsigned char **\fImark\fP;
1612 .sp
1613 In the 16-bit version of this structure, the \fImark\fP field has type
1614 "PCRE_UCHAR16 **".
1615 .sp
1616 In the 32-bit version of this structure, the \fImark\fP field has type
1617 "PCRE_UCHAR32 **".
1618 .P
1619 The \fIflags\fP field is used to specify which of the other fields are set. The
1620 flag bits are:
1621 .sp
1622 PCRE_EXTRA_CALLOUT_DATA
1623 PCRE_EXTRA_EXECUTABLE_JIT
1624 PCRE_EXTRA_MARK
1625 PCRE_EXTRA_MATCH_LIMIT
1626 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1627 PCRE_EXTRA_STUDY_DATA
1628 PCRE_EXTRA_TABLES
1629 .sp
1630 Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1631 the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1632 returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1633 should not set these yourself, but you may add to the block by setting other
1634 fields and their corresponding flag bits.
1635 .P
1636 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1637 vast amount of resources when running patterns that are not going to match,
1638 but which have a very large number of possibilities in their search trees. The
1639 classic example is a pattern that uses nested unlimited repeats.
1640 .P
1641 Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1642 calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1643 imposed on the number of times this function is called during a match, which
1644 has the effect of limiting the amount of backtracking that can take place. For
1645 patterns that are not anchored, the count restarts from zero for each position
1646 in the subject string.
1647 .P
1648 When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1649 with a JIT option, the way that the matching is executed is entirely different.
1650 However, there is still the possibility of runaway matching that goes on for a
1651 very long time, and so the \fImatch_limit\fP value is also used in this case
1652 (but in a different way) to limit how long the matching can continue.
1653 .P
1654 The default value for the limit can be set when PCRE is built; the default
1655 default is 10 million, which handles all but the most extreme cases. You can
1656 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1657 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1658 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1659 PCRE_ERROR_MATCHLIMIT.
1660 .P
1661 A value for the match limit may also be supplied by an item at the start of a
1662 pattern of the form
1663 .sp
1664 (*LIMIT_MATCH=d)
1665 .sp
1666 where d is a decimal number. However, such a setting is ignored unless d is
1667 less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1668 is set, less than the default.
1669 .P
1670 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1671 instead of limiting the total number of times that \fBmatch()\fP is called, it
1672 limits the depth of recursion. The recursion depth is a smaller number than the
1673 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1674 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1675 .P
1676 Limiting the recursion depth limits the amount of machine stack that can be
1677 used, or, when PCRE has been compiled to use memory on the heap instead of the
1678 stack, the amount of heap memory that can be used. This limit is not relevant,
1679 and is ignored, when matching is done using JIT compiled code.
1680 .P
1681 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1682 built; the default default is the same value as the default for
1683 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1684 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1685 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1686 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1687 .P
1688 A value for the recursion limit may also be supplied by an item at the start of
1689 a pattern of the form
1690 .sp
1691 (*LIMIT_RECURSION=d)
1692 .sp
1693 where d is a decimal number. However, such a setting is ignored unless d is
1694 less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1695 is set, less than the default.
1696 .P
1697 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1698 and is described in the
1699 .\" HREF
1700 \fBpcrecallout\fP
1701 .\"
1702 documentation.
1703 .P
1704 The \fItables\fP field is used to pass a character tables pointer to
1705 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1706 pattern. A non-NULL value is stored with the compiled pattern only if custom
1707 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1708 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1709 internal tables to be used. This facility is helpful when re-using patterns
1710 that have been saved after compiling with an external set of tables, because
1711 the external tables might be at a different address when \fBpcre_exec()\fP is
1712 called. See the
1713 .\" HREF
1714 \fBpcreprecompile\fP
1715 .\"
1716 documentation for a discussion of saving compiled patterns for later use.
1717 .P
1718 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1719 be set to point to a suitable variable. If the pattern contains any
1720 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1721 a name to pass back, a pointer to the name string (zero terminated) is placed
1722 in the variable pointed to by the \fImark\fP field. The names are within the
1723 compiled pattern; if you wish to retain such a name you must copy it before
1724 freeing the memory of a compiled pattern. If there is no name to pass back, the
1725 variable pointed to by the \fImark\fP field is set to NULL. For details of the
1726 backtracking control verbs, see the section entitled
1727 .\" HTML <a href="pcrepattern#backtrackcontrol">
1728 .\" </a>
1729 "Backtracking control"
1730 .\"
1731 in the
1732 .\" HREF
1733 \fBpcrepattern\fP
1734 .\"
1735 documentation.
1736 .
1737 .
1738 .\" HTML <a name="execoptions"></a>
1739 .SS "Option bits for \fBpcre_exec()\fP"
1740 .rs
1741 .sp
1742 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1743 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1744 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1745 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1746 PCRE_PARTIAL_SOFT.
1747 .P
1748 If the pattern was successfully studied with one of the just-in-time (JIT)
1749 compile options, the only supported options for JIT execution are
1750 PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1751 PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1752 unsupported option is used, JIT execution is disabled and the normal
1753 interpretive code in \fBpcre_exec()\fP is run.
1754 .sp
1755 PCRE_ANCHORED
1756 .sp
1757 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1758 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1759 to be anchored by virtue of its contents, it cannot be made unachored at
1760 matching time.
1761 .sp
1762 PCRE_BSR_ANYCRLF
1763 PCRE_BSR_UNICODE
1764 .sp
1765 These options (which are mutually exclusive) control what the \eR escape
1766 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1767 match any Unicode newline sequence. These options override the choice that was
1768 made or defaulted when the pattern was compiled.
1769 .sp
1770 PCRE_NEWLINE_CR
1771 PCRE_NEWLINE_LF
1772 PCRE_NEWLINE_CRLF
1773 PCRE_NEWLINE_ANYCRLF
1774 PCRE_NEWLINE_ANY
1775 .sp
1776 These options override the newline definition that was chosen or defaulted when
1777 the pattern was compiled. For details, see the description of
1778 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1779 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1780 the way the match position is advanced after a match failure for an unanchored
1781 pattern.
1782 .P
1783 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1784 match attempt for an unanchored pattern fails when the current position is at a
1785 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1786 characters, the match position is advanced by two characters instead of one, in
1787 other words, to after the CRLF.
1788 .P
1789 The above rule is a compromise that makes the most common cases work as
1790 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1791 set), it does not match the string "\er\enA" because, after failing at the
1792 start, it skips both the CR and the LF before retrying. However, the pattern
1793 [\er\en]A does match that string, because it contains an explicit CR or LF
1794 reference, and so advances only by one character after the first failure.
1795 .P
1796 An explicit match for CR of LF is either a literal appearance of one of those
1797 characters, or one of the \er or \en escape sequences. Implicit matches such as
1798 [^X] do not count, nor does \es (which includes CR and LF in the characters
1799 that it matches).
1800 .P
1801 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1802 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1803 .sp
1804 PCRE_NOTBOL
1805 .sp
1806 This option specifies that first character of the subject string is not the
1807 beginning of a line, so the circumflex metacharacter should not match before
1808 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1809 never to match. This option affects only the behaviour of the circumflex
1810 metacharacter. It does not affect \eA.
1811 .sp
1812 PCRE_NOTEOL
1813 .sp
1814 This option specifies that the end of the subject string is not the end of a
1815 line, so the dollar metacharacter should not match it nor (except in multiline
1816 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1817 compile time) causes dollar never to match. This option affects only the
1818 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1819 .sp
1820 PCRE_NOTEMPTY
1821 .sp
1822 An empty string is not considered to be a valid match if this option is set. If
1823 there are alternatives in the pattern, they are tried. If all the alternatives
1824 match the empty string, the entire match fails. For example, if the pattern
1825 .sp
1826 a?b?
1827 .sp
1828 is applied to a string not beginning with "a" or "b", it matches an empty
1829 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1830 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1831 .sp
1832 PCRE_NOTEMPTY_ATSTART
1833 .sp
1834 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1835 the start of the subject is permitted. If the pattern is anchored, such a match
1836 can occur only if the pattern contains \eK.
1837 .P
1838 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1839 does make a special case of a pattern match of the empty string within its
1840 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1841 emulate Perl's behaviour after matching a null string by first trying the match
1842 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1843 if that fails, by advancing the starting offset (see below) and trying an
1844 ordinary match again. There is some code that demonstrates how to do this in
1845 the
1846 .\" HREF
1847 \fBpcredemo\fP
1848 .\"
1849 sample program. In the most general case, you have to check to see if the
1850 newline convention recognizes CRLF as a newline, and if so, and the current
1851 character is CR followed by LF, advance the starting offset by two characters
1852 instead of one.
1853 .sp
1854 PCRE_NO_START_OPTIMIZE
1855 .sp
1856 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1857 a match, in order to speed up the process. For example, if it is known that an
1858 unanchored match must start with a specific character, it searches the subject
1859 for that character, and fails immediately if it cannot find it, without
1860 actually running the main matching function. This means that a special item
1861 such as (*COMMIT) at the start of a pattern is not considered until after a
1862 suitable starting point for the match has been found. Also, when callouts or
1863 (*MARK) items are in use, these "start-up" optimizations can cause them to be
1864 skipped if the pattern is never actually used. The start-up optimizations are
1865 in effect a pre-scan of the subject that takes place before the pattern is run.
1866 .P
1867 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1868 causing performance to suffer, but ensuring that in cases where the result is
1869 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1870 are considered at every possible starting position in the subject string. If
1871 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1872 time. The use of PCRE_NO_START_OPTIMIZE at matching time (that is, passing it
1873 to \fBpcre_exec()\fP) disables JIT execution; in this situation, matching is
1874 always done using interpretively.
1875 .P
1876 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1877 Consider the pattern
1878 .sp
1879 (*COMMIT)ABC
1880 .sp
1881 When this is compiled, PCRE records the fact that a match must start with the
1882 character "A". Suppose the subject string is "DEFABC". The start-up
1883 optimization scans along the subject, finds "A" and runs the first match
1884 attempt from there. The (*COMMIT) item means that the pattern must match the
1885 current starting position, which in this case, it does. However, if the same
1886 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1887 subject string does not happen. The first match attempt is run starting from
1888 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1889 the overall result is "no match". If the pattern is studied, more start-up
1890 optimizations may be used. For example, a minimum length for the subject may be
1891 recorded. Consider the pattern
1892 .sp
1893 (*MARK:A)(X|Y)
1894 .sp
1895 The minimum length for a match is one character. If the subject is "ABC", there
1896 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1897 If the pattern is studied, the final attempt does not take place, because PCRE
1898 knows that the subject is too short, and so the (*MARK) is never encountered.
1899 In this case, studying the pattern does not affect the overall match result,
1900 which is still "no match", but it does affect the auxiliary information that is
1901 returned.
1902 .sp
1903 PCRE_NO_UTF8_CHECK
1904 .sp
1905 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1906 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1907 The entire string is checked before any other processing takes place. The value
1908 of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1909 UTF-8 character. There is a discussion about the
1910 .\" HTML <a href="pcreunicode.html#utf8strings">
1911 .\" </a>
1912 validity of UTF-8 strings
1913 .\"
1914 in the
1915 .\" HREF
1916 \fBpcreunicode\fP
1917 .\"
1918 page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1919 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1920 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1921 cases, information about the precise nature of the error may also be returned
1922 (see the descriptions of these errors in the section entitled \fIError return
1923 values from\fP \fBpcre_exec()\fP
1924 .\" HTML <a href="#errorlist">
1925 .\" </a>
1926 below).
1927 .\"
1928 If \fIstartoffset\fP contains a value that does not point to the start of a
1929 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1930 returned.
1931 .P
1932 If you already know that your subject is valid, and you want to skip these
1933 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1934 calling \fBpcre_exec()\fP. You might want to do this for the second and
1935 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1936 all the matches in a single subject string. However, you should be sure that
1937 the value of \fIstartoffset\fP points to the start of a character (or the end
1938 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1939 invalid string as a subject or an invalid value of \fIstartoffset\fP is
1940 undefined. Your program may crash or loop.
1941 .sp
1942 PCRE_PARTIAL_HARD
1943 PCRE_PARTIAL_SOFT
1944 .sp
1945 These options turn on the partial matching feature. For backwards
1946 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1947 occurs if the end of the subject string is reached successfully, but there are
1948 not enough subject characters to complete the match. If this happens when
1949 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1950 testing any remaining alternatives. Only if no complete match can be found is
1951 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1952 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1953 but only if no complete match can be found.
1954 .P
1955 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1956 partial match is found, \fBpcre_exec()\fP immediately returns
1957 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1958 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1959 important that an alternative complete match.
1960 .P
1961 In both cases, the portion of the string that was inspected when the partial
1962 match was found is set as the first matching string. There is a more detailed
1963 discussion of partial and multi-segment matching, with examples, in the
1964 .\" HREF
1965 \fBpcrepartial\fP
1966 .\"
1967 documentation.
1968 .
1969 .
1970 .SS "The string to be matched by \fBpcre_exec()\fP"
1971 .rs
1972 .sp
1973 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1974 \fIsubject\fP, a length in \fIlength\fP, and a starting offset in
1975 \fIstartoffset\fP. The units for \fIlength\fP and \fIstartoffset\fP are bytes
1976 for the 8-bit library, 16-bit data items for the 16-bit library, and 32-bit
1977 data items for the 32-bit library.
1978 .P
1979 If \fIstartoffset\fP is negative or greater than the length of the subject,
1980 \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting offset is
1981 zero, the search for a match starts at the beginning of the subject, and this
1982 is by far the most common case. In UTF-8 or UTF-16 mode, the offset must point
1983 to the start of a character, or the end of the subject (in UTF-32 mode, one
1984 data unit equals one character, so all offsets are valid). Unlike the pattern
1985 string, the subject may contain binary zeroes.
1986 .P
1987 A non-zero starting offset is useful when searching for another match in the
1988 same subject by calling \fBpcre_exec()\fP again after a previous success.
1989 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1990 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1991 lookbehind. For example, consider the pattern
1992 .sp
1993 \eBiss\eB
1994 .sp
1995 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1996 the current position in the subject is not a word boundary.) When applied to
1997 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1998 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1999 subject, namely "issipi", it does not match, because \eB is always false at the
2000 start of the subject, which is deemed to be a word boundary. However, if
2001 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
2002 set to 4, it finds the second occurrence of "iss" because it is able to look
2003 behind the starting point to discover that it is preceded by a letter.
2004 .P
2005 Finding all the matches in a subject is tricky when the pattern can match an
2006 empty string. It is possible to emulate Perl's /g behaviour by first trying the
2007 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
2008 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
2009 and trying an ordinary match again. There is some code that demonstrates how to
2010 do this in the
2011 .\" HREF
2012 \fBpcredemo\fP
2013 .\"
2014 sample program. In the most general case, you have to check to see if the
2015 newline convention recognizes CRLF as a newline, and if so, and the current
2016 character is CR followed by LF, advance the starting offset by two characters
2017 instead of one.
2018 .P
2019 If a non-zero starting offset is passed when the pattern is anchored, one
2020 attempt to match at the given offset is made. This can only succeed if the
2021 pattern does not require the match to be at the start of the subject.
2022 .
2023 .
2024 .SS "How \fBpcre_exec()\fP returns captured substrings"
2025 .rs
2026 .sp
2027 In general, a pattern matches a certain portion of the subject, and in
2028 addition, further substrings from the subject may be picked out by parts of the
2029 pattern. Following the usage in Jeffrey Friedl's book, this is called
2030 "capturing" in what follows, and the phrase "capturing subpattern" is used for
2031 a fragment of a pattern that picks out a substring. PCRE supports several other
2032 kinds of parenthesized subpattern that do not cause substrings to be captured.
2033 .P
2034 Captured substrings are returned to the caller via a vector of integers whose
2035 address is passed in \fIovector\fP. The number of elements in the vector is
2036 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
2037 argument is NOT the size of \fIovector\fP in bytes.
2038 .P
2039 The first two-thirds of the vector is used to pass back captured substrings,
2040 each substring using a pair of integers. The remaining third of the vector is
2041 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
2042 and is not available for passing back information. The number passed in
2043 \fIovecsize\fP should always be a multiple of three. If it is not, it is
2044 rounded down.
2045 .P
2046 When a match is successful, information about captured substrings is returned
2047 in pairs of integers, starting at the beginning of \fIovector\fP, and
2048 continuing up to two-thirds of its length at the most. The first element of
2049 each pair is set to the offset of the first character in a substring, and the
2050 second is set to the offset of the first character after the end of a
2051 substring. These values are always data unit offsets, even in UTF mode. They
2052 are byte offsets in the 8-bit library, 16-bit data item offsets in the 16-bit
2053 library, and 32-bit data item offsets in the 32-bit library. \fBNote\fP: they
2054 are not character counts.
2055 .P
2056 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
2057 portion of the subject string matched by the entire pattern. The next pair is
2058 used for the first capturing subpattern, and so on. The value returned by
2059 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
2060 For example, if two substrings have been captured, the returned value is 3. If
2061 there are no capturing subpatterns, the return value from a successful match is
2062 1, indicating that just the first pair of offsets has been set.
2063 .P
2064 If a capturing subpattern is matched repeatedly, it is the last portion of the
2065 string that it matched that is returned.
2066 .P
2067 If the vector is too small to hold all the captured substring offsets, it is
2068 used as far as possible (up to two-thirds of its length), and the function
2069 returns a value of zero. If neither the actual string matched nor any captured
2070 substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2071 passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2072 back references and the \fIovector\fP is not big enough to remember the related
2073 substrings, PCRE has to get additional memory for use during matching. Thus it
2074 is usually advisable to supply an \fIovector\fP of reasonable size.
2075 .P
2076 There are some cases where zero is returned (indicating vector overflow) when
2077 in fact the vector is exactly the right size for the final match. For example,
2078 consider the pattern
2079 .sp
2080 (a)(?:(b)c|bd)
2081 .sp
2082 If a vector of 6 elements (allowing for only 1 captured substring) is given
2083 with subject string "abd", \fBpcre_exec()\fP will try to set the second
2084 captured string, thereby recording a vector overflow, before failing to match
2085 "c" and backing up to try the second alternative. The zero return, however,
2086 does correctly indicate that the maximum number of slots (namely 2) have been
2087 filled. In similar cases where there is temporary overflow, but the final
2088 number of used slots is actually less than the maximum, a non-zero value is
2089 returned.
2090 .P
2091 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
2092 subpatterns there are in a compiled pattern. The smallest size for
2093 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
2094 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
2095 .P
2096 It is possible for capturing subpattern number \fIn+1\fP to match some part of
2097 the subject when subpattern \fIn\fP has not been used at all. For example, if
2098 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2099 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2100 happens, both values in the offset pairs corresponding to unused subpatterns
2101 are set to -1.
2102 .P
2103 Offset values that correspond to unused subpatterns at the end of the
2104 expression are also set to -1. For example, if the string "abc" is matched
2105 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2106 return from the function is 2, because the highest used capturing subpattern
2107 number is 1, and the offsets for for the second and third capturing subpatterns
2108 (assuming the vector is large enough, of course) are set to -1.
2109 .P
2110 \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
2111 correspond to capturing parentheses in the pattern are never changed. That is,
2112 if a pattern contains \fIn\fP capturing parentheses, no more than
2113 \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
2114 elements (in the first two-thirds) retain whatever values they previously had.
2115 .P
2116 Some convenience functions are provided for extracting the captured substrings
2117 as separate strings. These are described below.
2118 .
2119 .
2120 .\" HTML <a name="errorlist"></a>
2121 .SS "Error return values from \fBpcre_exec()\fP"
2122 .rs
2123 .sp
2124 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
2125 defined in the header file:
2126 .sp
2127 PCRE_ERROR_NOMATCH (-1)
2128 .sp
2129 The subject string did not match the pattern.
2130 .sp
2131 PCRE_ERROR_NULL (-2)
2132 .sp
2133 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
2134 NULL and \fIovecsize\fP was not zero.
2135 .sp
2136 PCRE_ERROR_BADOPTION (-3)
2137 .sp
2138 An unrecognized bit was set in the \fIoptions\fP argument.
2139 .sp
2140 PCRE_ERROR_BADMAGIC (-4)
2141 .sp
2142 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2143 the case when it is passed a junk pointer and to detect when a pattern that was
2144 compiled in an environment of one endianness is run in an environment with the
2145 other endianness. This is the error that PCRE gives when the magic number is
2146 not present.
2147 .sp
2148 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2149 .sp
2150 While running the pattern match, an unknown item was encountered in the
2151 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2152 of the compiled pattern.
2153 .sp
2154 PCRE_ERROR_NOMEMORY (-6)
2155 .sp
2156 If a pattern contains back references, but the \fIovector\fP that is passed to
2157 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2158 gets a block of memory at the start of matching to use for this purpose. If the
2159 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2160 automatically freed at the end of matching.
2161 .P
2162 This error is also given if \fBpcre_stack_malloc()\fP fails in
2163 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2164 \fB--disable-stack-for-recursion\fP.
2165 .sp
2166 PCRE_ERROR_NOSUBSTRING (-7)
2167 .sp
2168 This error is used by the \fBpcre_copy_substring()\fP,
2169 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2170 below). It is never returned by \fBpcre_exec()\fP.
2171 .sp
2172 PCRE_ERROR_MATCHLIMIT (-8)
2173 .sp
2174 The backtracking limit, as specified by the \fImatch_limit\fP field in a
2175 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2176 above.
2177 .sp
2178 PCRE_ERROR_CALLOUT (-9)
2179 .sp
2180 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2181 use by callout functions that want to yield a distinctive error code. See the
2182 .\" HREF
2183 \fBpcrecallout\fP
2184 .\"
2185 documentation for details.
2186 .sp
2187 PCRE_ERROR_BADUTF8 (-10)
2188 .sp
2189 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2190 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2191 (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2192 UTF-8 character is placed in the first element, and a reason code is placed in
2193 the second element. The reason codes are listed in the
2194 .\" HTML <a href="#badutf8reasons">
2195 .\" </a>
2196 following section.
2197 .\"
2198 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2199 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2200 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2201 .sp
2202 PCRE_ERROR_BADUTF8_OFFSET (-11)
2203 .sp
2204 The UTF-8 byte sequence that was passed as a subject was checked and found to
2205 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2206 \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2207 end of the subject.
2208 .sp
2209 PCRE_ERROR_PARTIAL (-12)
2210 .sp
2211 The subject string did not match, but it did match partially. See the
2212 .\" HREF
2213 \fBpcrepartial\fP
2214 .\"
2215 documentation for details of partial matching.
2216 .sp
2217 PCRE_ERROR_BADPARTIAL (-13)
2218 .sp
2219 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2220 option was used with a compiled pattern containing items that were not
2221 supported for partial matching. From release 8.00 onwards, there are no
2222 restrictions on partial matching.
2223 .sp
2224 PCRE_ERROR_INTERNAL (-14)
2225 .sp
2226 An unexpected internal error has occurred. This error could be caused by a bug
2227 in PCRE or by overwriting of the compiled pattern.
2228 .sp
2229 PCRE_ERROR_BADCOUNT (-15)
2230 .sp
2231 This error is given if the value of the \fIovecsize\fP argument is negative.
2232 .sp
2233 PCRE_ERROR_RECURSIONLIMIT (-21)
2234 .sp
2235 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2236 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2237 description above.
2238 .sp
2239 PCRE_ERROR_BADNEWLINE (-23)
2240 .sp
2241 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2242 .sp
2243 PCRE_ERROR_BADOFFSET (-24)
2244 .sp
2245 The value of \fIstartoffset\fP was negative or greater than the length of the
2246 subject, that is, the value in \fIlength\fP.
2247 .sp
2248 PCRE_ERROR_SHORTUTF8 (-25)
2249 .sp
2250 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2251 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2252 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2253 fact sufficient to detect this case, but this special error code for
2254 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2255 retained for backwards compatibility.
2256 .sp
2257 PCRE_ERROR_RECURSELOOP (-26)
2258 .sp
2259 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2260 the pattern. Specifically, it means that either the whole pattern or a
2261 subpattern has been called recursively for the second time at the same position
2262 in the subject string. Some simple patterns that might do this are detected and
2263 faulted at compile time, but more complicated cases, in particular mutual
2264 recursions between two different subpatterns, cannot be detected until run
2265 time.
2266 .sp
2267 PCRE_ERROR_JIT_STACKLIMIT (-27)
2268 .sp
2269 This error is returned when a pattern that was successfully studied using a
2270 JIT compile option is being matched, but the memory available for the
2271 just-in-time processing stack is not large enough. See the
2272 .\" HREF
2273 \fBpcrejit\fP
2274 .\"
2275 documentation for more details.
2276 .sp
2277 PCRE_ERROR_BADMODE (-28)
2278 .sp
2279 This error is given if a pattern that was compiled by the 8-bit library is
2280 passed to a 16-bit or 32-bit library function, or vice versa.
2281 .sp
2282 PCRE_ERROR_BADENDIANNESS (-29)
2283 .sp
2284 This error is given if a pattern that was compiled and saved is reloaded on a
2285 host with different endianness. The utility function
2286 \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2287 so that it runs on the new host.
2288 .sp
2289 PCRE_ERROR_JIT_BADOPTION
2290 .sp
2291 This error is returned when a pattern that was successfully studied using a JIT
2292 compile option is being matched, but the matching mode (partial or complete
2293 match) does not correspond to any JIT compilation mode. When the JIT fast path
2294 function is used, this error may be also given for invalid options. See the
2295 .\" HREF
2296 \fBpcrejit\fP
2297 .\"
2298 documentation for more details.
2299 .sp
2300 PCRE_ERROR_BADLENGTH (-32)
2301 .sp
2302 This error is given if \fBpcre_exec()\fP is called with a negative value for
2303 the \fIlength\fP argument.
2304 .P
2305 Error numbers -16 to -20, -22, and 30 are not used by \fBpcre_exec()\fP.
2306 .
2307 .
2308 .\" HTML <a name="badutf8reasons"></a>
2309 .SS "Reason codes for invalid UTF-8 strings"
2310 .rs
2311 .sp
2312 This section applies only to the 8-bit library. The corresponding information
2313 for the 16-bit and 32-bit libraries is given in the
2314 .\" HREF
2315 \fBpcre16\fP
2316 .\"
2317 and
2318 .\" HREF
2319 \fBpcre32\fP
2320 .\"
2321 pages.
2322 .P
2323 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2324 PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2325 least 2, the offset of the start of the invalid UTF-8 character is placed in
2326 the first output vector element (\fIovector[0]\fP) and a reason code is placed
2327 in the second element (\fIovector[1]\fP). The reason codes are given names in
2328 the \fBpcre.h\fP header file:
2329 .sp
2330 PCRE_UTF8_ERR1
2331 PCRE_UTF8_ERR2
2332 PCRE_UTF8_ERR3
2333 PCRE_UTF8_ERR4
2334 PCRE_UTF8_ERR5
2335 .sp
2336 The string ends with a truncated UTF-8 character; the code specifies how many
2337 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2338 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2339 allows for up to 6 bytes, and this is checked first; hence the possibility of
2340 4 or 5 missing bytes.
2341 .sp
2342 PCRE_UTF8_ERR6
2343 PCRE_UTF8_ERR7
2344 PCRE_UTF8_ERR8
2345 PCRE_UTF8_ERR9
2346 PCRE_UTF8_ERR10
2347 .sp
2348 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2349 character do not have the binary value 0b10 (that is, either the most
2350 significant bit is 0, or the next bit is 1).
2351 .sp
2352 PCRE_UTF8_ERR11
2353 PCRE_UTF8_ERR12
2354 .sp
2355 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2356 these code points are excluded by RFC 3629.
2357 .sp
2358 PCRE_UTF8_ERR13
2359 .sp
2360 A 4-byte character has a value greater than 0x10fff; these code points are
2361 excluded by RFC 3629.
2362 .sp
2363 PCRE_UTF8_ERR14
2364 .sp
2365 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2366 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2367 from UTF-8.
2368 .sp
2369 PCRE_UTF8_ERR15
2370 PCRE_UTF8_ERR16
2371 PCRE_UTF8_ERR17
2372 PCRE_UTF8_ERR18
2373 PCRE_UTF8_ERR19
2374 .sp
2375 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2376 value that can be represented by fewer bytes, which is invalid. For example,
2377 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2378 one byte.
2379 .sp
2380 PCRE_UTF8_ERR20
2381 .sp
2382 The two most significant bits of the first byte of a character have the binary
2383 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2384 byte can only validly occur as the second or subsequent byte of a multi-byte
2385 character.
2386 .sp
2387 PCRE_UTF8_ERR21
2388 .sp
2389 The first byte of a character has the value 0xfe or 0xff. These values can
2390 never occur in a valid UTF-8 string.
2391 .sp
2392 PCRE_UTF8_ERR22
2393 .sp
2394 This error code was formerly used when the presence of a so-called
2395 "non-character" caused an error. Unicode corrigendum #9 makes it clear that
2396 such characters should not cause a string to be rejected, and so this code is
2397 no longer in use and is never returned.
2398 .
2399 .
2400 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2401 .rs
2402 .sp
2403 .nf
2404 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2405 .B " int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,"
2406 .B " int \fIbuffersize\fP);"
2407 .sp
2408 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2409 .B " int \fIstringcount\fP, int \fIstringnumber\fP,"
2410 .B " const char **\fIstringptr\fP);"
2411 .sp
2412 .B int pcre_get_substring_list(const char *\fIsubject\fP,
2413 .B " int *\fIovector\fP, int \fIstringcount\fP, const char ***\fIlistptr\fP);"
2414 .fi
2415 .PP
2416 Captured substrings can be accessed directly by using the offsets returned by
2417 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2418 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2419 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2420 as new, separate, zero-terminated strings. These functions identify substrings
2421 by number. The next section describes functions for extracting named
2422 substrings.
2423 .P
2424 A substring that contains a binary zero is correctly extracted and has a
2425 further zero added on the end, but the result is not, of course, a C string.
2426 However, you can process such a string by referring to the length that is
2427 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2428 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2429 for handling strings containing binary zeros, because the end of the final
2430 string is not independently indicated.
2431 .P
2432 The first three arguments are the same for all three of these functions:
2433 \fIsubject\fP is the subject string that has just been successfully matched,
2434 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2435 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2436 captured by the match, including the substring that matched the entire regular
2437 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2438 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2439 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2440 number of elements in the vector divided by three.
2441 .P
2442 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2443 extract a single substring, whose number is given as \fIstringnumber\fP. A
2444 value of zero extracts the substring that matched the entire pattern, whereas
2445 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2446 the string is placed in \fIbuffer\fP, whose length is given by
2447 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2448 obtained via \fBpcre_malloc\fP, and its address is returned via
2449 \fIstringptr\fP. The yield of the function is the length of the string, not
2450 including the terminating zero, or one of these error codes:
2451 .sp
2452 PCRE_ERROR_NOMEMORY (-6)
2453 .sp
2454 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2455 memory failed for \fBpcre_get_substring()\fP.
2456 .sp
2457 PCRE_ERROR_NOSUBSTRING (-7)
2458 .sp
2459 There is no substring whose number is \fIstringnumber\fP.
2460 .P
2461 The \fBpcre_get_substring_list()\fP function extracts all available substrings
2462 and builds a list of pointers to them. All this is done in a single block of
2463 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2464 is returned via \fIlistptr\fP, which is also the start of the list of string
2465 pointers. The end of the list is marked by a NULL pointer. The yield of the
2466 function is zero if all went well, or the error code
2467 .sp
2468 PCRE_ERROR_NOMEMORY (-6)
2469 .sp
2470 if the attempt to get the memory block failed.
2471 .P
2472 When any of these functions encounter a substring that is unset, which can
2473 happen when capturing subpattern number \fIn+1\fP matches some part of the
2474 subject, but subpattern \fIn\fP has not been used at all, they return an empty
2475 string. This can be distinguished from a genuine zero-length substring by
2476 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2477 substrings.
2478 .P
2479 The two convenience functions \fBpcre_free_substring()\fP and
2480 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2481 a previous call of \fBpcre_get_substring()\fP or
2482 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2483 the function pointed to by \fBpcre_free\fP, which of course could be called
2484 directly from a C program. However, PCRE is used in some situations where it is
2485 linked via a special interface to another programming language that cannot use
2486 \fBpcre_free\fP directly; it is for these cases that the functions are
2487 provided.
2488 .
2489 .
2490 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2491 .rs
2492 .sp
2493 .nf
2494 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2495 .B " const char *\fIname\fP);"
2496 .sp
2497 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2498 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
2499 .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
2500 .B " char *\fIbuffer\fP, int \fIbuffersize\fP);"
2501 .sp
2502 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2503 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
2504 .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
2505 .B " const char **\fIstringptr\fP);"
2506 .fi
2507 .PP
2508 To extract a substring by name, you first have to find associated number.
2509 For example, for this pattern
2510 .sp
2511 (a+)b(?<xxx>\ed+)...
2512 .sp
2513 the number of the subpattern called "xxx" is 2. If the name is known to be
2514 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2515 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2516 pattern, and the second is the name. The yield of the function is the
2517 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2518 that name.
2519 .P
2520 Given the number, you can extract the substring directly, or use one of the
2521 functions described in the previous section. For convenience, there are also
2522 two functions that do the whole job.
2523 .P
2524 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2525 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2526 functions that extract by number. As these are described in the previous
2527 section, they are not re-described here. There are just two differences:
2528 .P
2529 First, instead of a substring number, a substring name is given. Second, there
2530 is an extra argument, given at the start, which is a pointer to the compiled
2531 pattern. This is needed in order to gain access to the name-to-number
2532 translation table.
2533 .P
2534 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2535 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2536 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2537 the behaviour may not be what you want (see the next section).
2538 .P
2539 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2540 subpatterns with the same number, as described in the
2541 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2542 .\" </a>
2543 section on duplicate subpattern numbers
2544 .\"
2545 in the
2546 .\" HREF
2547 \fBpcrepattern\fP
2548 .\"
2549 page, you cannot use names to distinguish the different subpatterns, because
2550 names are not included in the compiled code. The matching process uses only
2551 numbers. For this reason, the use of different names for subpatterns of the
2552 same number causes an error at compile time.
2553 .
2554 .
2555 .SH "DUPLICATE SUBPATTERN NAMES"
2556 .rs
2557 .sp
2558 .nf
2559 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2560 .B " const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);"
2561 .fi
2562 .PP
2563 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2564 are not required to be unique. (Duplicate names are always allowed for
2565 subpatterns with the same number, created by using the (?| feature. Indeed, if
2566 such subpatterns are named, they are required to use the same names.)
2567 .P
2568 Normally, patterns with duplicate names are such that in any one match, only
2569 one of the named subpatterns participates. An example is shown in the
2570 .\" HREF
2571 \fBpcrepattern\fP
2572 .\"
2573 documentation.
2574 .P
2575 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2576 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2577 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2578 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2579 returns one of the numbers that are associated with the name, but it is not
2580 defined which it is.
2581 .P
2582 If you want to get full details of all captured substrings for a given name,
2583 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2584 argument is the compiled pattern, and the second is the name. The third and
2585 fourth are pointers to variables which are updated by the function. After it
2586 has run, they point to the first and last entries in the name-to-number table
2587 for the given name. The function itself returns the length of each entry, or
2588 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2589 described above in the section entitled \fIInformation about a pattern\fP
2590 .\" HTML <a href="#infoaboutpattern">
2591 .\" </a>
2592 above.
2593 .\"
2594 Given all the relevant entries for the name, you can extract each of their
2595 numbers, and hence the captured data, if any.
2596 .
2597 .
2598 .SH "FINDING ALL POSSIBLE MATCHES"
2599 .rs
2600 .sp
2601 The traditional matching function uses a similar algorithm to Perl, which stops
2602 when it finds the first match, starting at a given point in the subject. If you
2603 want to find all possible matches, or the longest possible match, consider
2604 using the alternative matching function (see below) instead. If you cannot use
2605 the alternative function, but still need to find all possible matches, you
2606 can kludge it up by making use of the callout facility, which is described in
2607 the
2608 .\" HREF
2609 \fBpcrecallout\fP
2610 .\"
2611 documentation.
2612 .P
2613 What you have to do is to insert a callout right at the end of the pattern.
2614 When your callout function is called, extract and save the current matched
2615 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2616 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2617 will yield PCRE_ERROR_NOMATCH.
2618 .
2619 .
2620 .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2621 .rs
2622 .sp
2623 Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2624 stack, which in certain environments can be rather limited in size. Some users
2625 find it helpful to have an estimate of the amount of stack that is used by
2626 \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2627 .\" HREF
2628 \fBpcrestack\fP
2629 .\"
2630 documentation. The estimate that is output by \fBpcretest\fP when called with
2631 the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2632 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2633 .P
2634 Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2635 the negative error code PCRE_ERROR_NULL, but with this special combination of
2636 arguments, it returns instead a negative number whose absolute value is the
2637 approximate stack frame size in bytes. (A negative number is used so that it is
2638 clear that no match has happened.) The value is approximate because in some
2639 cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2640 additional variables on the stack.
2641 .P
2642 If PCRE has been compiled to use the heap instead of the stack for recursion,
2643 the value returned is the size of each block that is obtained from the heap.
2644 .
2645 .
2646 .\" HTML <a name="dfamatch"></a>
2647 .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2648 .rs
2649 .sp
2650 .nf
2651 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2652 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
2653 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
2654 .B " int *\fIworkspace\fP, int \fIwscount\fP);"
2655 .fi
2656 .P
2657 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2658 a compiled pattern, using a matching algorithm that scans the subject string
2659 just once, and does not backtrack. This has different characteristics to the
2660 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2661 patterns are not supported. Nevertheless, there are times when this kind of
2662 matching can be useful. For a discussion of the two matching algorithms, and a
2663 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2664 .\" HREF
2665 \fBpcrematching\fP
2666 .\"
2667 documentation.
2668 .P
2669 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2670 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2671 different way, and this is described below. The other common arguments are used
2672 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2673 here.
2674 .P
2675 The two additional arguments provide workspace for the function. The workspace
2676 vector should contain at least 20 elements. It is used for keeping track of
2677 multiple paths through the pattern tree. More workspace will be needed for
2678 patterns and subjects where there are a lot of potential matches.
2679 .P
2680 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2681 .sp
2682 int rc;
2683 int ovector[10];
2684 int wspace[20];
2685 rc = pcre_dfa_exec(
2686 re, /* result of pcre_compile() */
2687 NULL, /* we didn't study the pattern */
2688 "some string", /* the subject string */
2689 11, /* the length of the subject string */
2690 0, /* start at offset 0 in the subject */
2691 0, /* default options */
2692 ovector, /* vector of integers for substring information */
2693 10, /* number of elements (NOT size in bytes) */
2694 wspace, /* working space vector */
2695 20); /* number of elements (NOT size in bytes) */
2696 .
2697 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2698 .rs
2699 .sp
2700 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2701 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2702 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2703 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2704 PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2705 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2706 so their description is not repeated here.
2707 .sp
2708 PCRE_PARTIAL_HARD
2709 PCRE_PARTIAL_SOFT
2710 .sp
2711 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2712 details are slightly different. When PCRE_PARTIAL_HARD is set for
2713 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2714 is reached and there is still at least one matching possibility that requires
2715 additional characters. This happens even if some complete matches have also
2716 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2717 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2718 there have been no complete matches, but there is still at least one matching
2719 possibility. The portion of the string that was inspected when the longest
2720 partial match was found is set as the first matching string in both cases.
2721 There is a more detailed discussion of partial and multi-segment matching, with
2722 examples, in the
2723 .\" HREF
2724 \fBpcrepartial\fP
2725 .\"
2726 documentation.
2727 .sp
2728 PCRE_DFA_SHORTEST
2729 .sp
2730 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2731 soon as it has found one match. Because of the way the alternative algorithm
2732 works, this is necessarily the shortest possible match at the first possible
2733 matching point in the subject string.
2734 .sp
2735 PCRE_DFA_RESTART
2736 .sp
2737 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2738 again, with additional subject characters, and have it continue with the same
2739 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2740 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2741 before because data about the match so far is left in them after a partial
2742 match. There is more discussion of this facility in the
2743 .\" HREF
2744 \fBpcrepartial\fP
2745 .\"
2746 documentation.
2747 .
2748 .
2749 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2750 .rs
2751 .sp
2752 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2753 substring in the subject. Note, however, that all the matches from one run of
2754 the function start at the same point in the subject. The shorter matches are
2755 all initial substrings of the longer matches. For example, if the pattern
2756 .sp
2757 <.*>
2758 .sp
2759 is matched against the string
2760 .sp
2761 This is <something> <something else> <something further> no more
2762 .sp
2763 the three matched strings are
2764 .sp
2765 <something>
2766 <something> <something else>
2767 <something> <something else> <something further>
2768 .sp
2769 On success, the yield of the function is a number greater than zero, which is
2770 the number of matched substrings. The substrings themselves are returned in
2771 \fIovector\fP. Each string uses two elements; the first is the offset to the
2772 start, and the second is the offset to the end. In fact, all the strings have
2773 the same start offset. (Space could have been saved by giving this only once,
2774 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2775 returns data, even though the meaning of the strings is different.)
2776 .P
2777 The strings are returned in reverse order of length; that is, the longest
2778 matching string is given first. If there were too many matches to fit into
2779 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2780 the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2781 the entire \fIovector\fP for returning matched strings.
2782
2783 NOTE: PCRE's "auto-possessification" optimization usually applies to character
2784 repeats at the end of a pattern (as well as internally). For example, the
2785 pattern "a\ed+" is compiled as if it were "a\ed++" because there is no point
2786 even considering the possibility of backtracking into the repeated digits. For
2787 DFA matching, this means that only one possible match is found. If you really
2788 do want multiple matches in such cases, either use an ungreedy repeat
2789 ("a\ed+?") or set the PCRE_NO_AUTO_POSSESSIFY option when compiling.
2790 .
2791 .
2792 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2793 .rs
2794 .sp
2795 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2796 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2797 described
2798 .\" HTML <a href="#errorlist">
2799 .\" </a>
2800 above.
2801 .\"
2802 There are in addition the following errors that are specific to
2803 \fBpcre_dfa_exec()\fP:
2804 .sp
2805 PCRE_ERROR_DFA_UITEM (-16)
2806 .sp
2807 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2808 that it does not support, for instance, the use of \eC or a back reference.
2809 .sp
2810 PCRE_ERROR_DFA_UCOND (-17)
2811 .sp
2812 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2813 uses a back reference for the condition, or a test for recursion in a specific
2814 group. These are not supported.
2815 .sp
2816 PCRE_ERROR_DFA_UMLIMIT (-18)
2817 .sp
2818 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2819 block that contains a setting of the \fImatch_limit\fP or
2820 \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2821 meaningless for DFA matching).
2822 .sp
2823 PCRE_ERROR_DFA_WSSIZE (-19)
2824 .sp
2825 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2826 \fIworkspace\fP vector.
2827 .sp
2828 PCRE_ERROR_DFA_RECURSE (-20)
2829 .sp
2830 When a recursive subpattern is processed, the matching function calls itself
2831 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2832 error is given if the output vector is not large enough. This should be
2833 extremely rare, as a vector of size 1000 is used.
2834 .sp
2835 PCRE_ERROR_DFA_BADRESTART (-30)
2836 .sp
2837 When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2838 some plausibility checks are made on the contents of the workspace, which
2839 should contain data about the previous partial match. If any of these checks
2840 fail, this error is given.
2841 .
2842 .
2843 .SH "SEE ALSO"
2844 .rs
2845 .sp
2846 \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2847 \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2848 \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2849 \fBpcrestack\fP(3).
2850 .
2851 .
2852 .SH AUTHOR
2853 .rs
2854 .sp
2855 .nf
2856 Philip Hazel
2857 University Computing Service
2858 Cambridge CB2 3QH, England.
2859 .fi
2860 .
2861 .
2862 .SH REVISION
2863 .rs
2864 .sp
2865 .nf
2866 Last updated: 01 October 2013
2867 Copyright (c) 1997-2013 University of Cambridge.
2868 .fi

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5