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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1339 - (show annotations)
Fri Jun 14 09:09:28 2013 UTC (6 years ago) by ph10
File size: 125756 byte(s)
Update man pages to avoid the use of .ti (which causes difficulties for some 
translation software).
1 .TH PCREAPI 3 "12 June 2013" "PCRE 8.33"
2 .SH NAME
3 PCRE - Perl-compatible regular expressions
4 .sp
5 .B #include <pcre.h>
6 .
7 .
8 .SH "PCRE NATIVE API BASIC FUNCTIONS"
9 .rs
10 .sp
11 .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_START_OPTIMIZE
799 .sp
800 This is an option that acts at matching time; that is, it is really an option
801 for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
802 it is remembered with the compiled pattern and assumed at matching time. This
803 is necessary if you want to use JIT execution, because the JIT compiler needs
804 to know whether or not this option is set. For details see the discussion of
805 PCRE_NO_START_OPTIMIZE
806 .\" HTML <a href="#execoptions">
807 .\" </a>
808 below.
809 .\"
810 .sp
811 PCRE_UCP
812 .sp
813 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
814 \ew, and some of the POSIX character classes. By default, only ASCII characters
815 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
816 classify characters. More details are given in the section on
817 .\" HTML <a href="pcre.html#genericchartypes">
818 .\" </a>
819 generic character types
820 .\"
821 in the
822 .\" HREF
823 \fBpcrepattern\fP
824 .\"
825 page. If you set PCRE_UCP, matching one of the items it affects takes much
826 longer. The option is available only if PCRE has been compiled with Unicode
827 property support.
828 .sp
829 PCRE_UNGREEDY
830 .sp
831 This option inverts the "greediness" of the quantifiers so that they are not
832 greedy by default, but become greedy if followed by "?". It is not compatible
833 with Perl. It can also be set by a (?U) option setting within the pattern.
834 .sp
835 PCRE_UTF8
836 .sp
837 This option causes PCRE to regard both the pattern and the subject as strings
838 of UTF-8 characters instead of single-byte strings. However, it is available
839 only when PCRE is built to include UTF support. If not, the use of this option
840 provokes an error. Details of how this option changes the behaviour of PCRE are
841 given in the
842 .\" HREF
843 \fBpcreunicode\fP
844 .\"
845 page.
846 .sp
847 PCRE_NO_UTF8_CHECK
848 .sp
849 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
850 automatically checked. There is a discussion about the
851 .\" HTML <a href="pcreunicode.html#utf8strings">
852 .\" </a>
853 validity of UTF-8 strings
854 .\"
855 in the
856 .\" HREF
857 \fBpcreunicode\fP
858 .\"
859 page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
860 error. If you already know that your pattern is valid, and you want to skip
861 this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
862 When it is set, the effect of passing an invalid UTF-8 string as a pattern is
863 undefined. It may cause your program to crash. Note that this option can also
864 be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress the
865 validity checking of subject strings only. If the same string is being matched
866 many times, the option can be safely set for the second and subsequent
867 matchings to improve performance.
868 .
869 .
870 .SH "COMPILATION ERROR CODES"
871 .rs
872 .sp
873 The following table lists the error codes than may be returned by
874 \fBpcre_compile2()\fP, along with the error messages that may be returned by
875 both compiling functions. Note that error messages are always 8-bit ASCII
876 strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
877 have fallen out of use. To avoid confusion, they have not been re-used.
878 .sp
879 0 no error
880 1 \e at end of pattern
881 2 \ec at end of pattern
882 3 unrecognized character follows \e
883 4 numbers out of order in {} quantifier
884 5 number too big in {} quantifier
885 6 missing terminating ] for character class
886 7 invalid escape sequence in character class
887 8 range out of order in character class
888 9 nothing to repeat
889 10 [this code is not in use]
890 11 internal error: unexpected repeat
891 12 unrecognized character after (? or (?-
892 13 POSIX named classes are supported only within a class
893 14 missing )
894 15 reference to non-existent subpattern
895 16 erroffset passed as NULL
896 17 unknown option bit(s) set
897 18 missing ) after comment
898 19 [this code is not in use]
899 20 regular expression is too large
900 21 failed to get memory
901 22 unmatched parentheses
902 23 internal error: code overflow
903 24 unrecognized character after (?<
904 25 lookbehind assertion is not fixed length
905 26 malformed number or name after (?(
906 27 conditional group contains more than two branches
907 28 assertion expected after (?(
908 29 (?R or (?[+-]digits must be followed by )
909 30 unknown POSIX class name
910 31 POSIX collating elements are not supported
911 32 this version of PCRE is compiled without UTF support
912 33 [this code is not in use]
913 34 character value in \ex{...} sequence is too large
914 35 invalid condition (?(0)
915 36 \eC not allowed in lookbehind assertion
916 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
917 38 number after (?C is > 255
918 39 closing ) for (?C expected
919 40 recursive call could loop indefinitely
920 41 unrecognized character after (?P
921 42 syntax error in subpattern name (missing terminator)
922 43 two named subpatterns have the same name
923 44 invalid UTF-8 string (specifically UTF-8)
924 45 support for \eP, \ep, and \eX has not been compiled
925 46 malformed \eP or \ep sequence
926 47 unknown property name after \eP or \ep
927 48 subpattern name is too long (maximum 32 characters)
928 49 too many named subpatterns (maximum 10000)
929 50 [this code is not in use]
930 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
931 52 internal error: overran compiling workspace
932 53 internal error: previously-checked referenced subpattern
933 not found
934 54 DEFINE group contains more than one branch
935 55 repeating a DEFINE group is not allowed
936 56 inconsistent NEWLINE options
937 57 \eg is not followed by a braced, angle-bracketed, or quoted
938 name/number or by a plain number
939 58 a numbered reference must not be zero
940 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
941 60 (*VERB) not recognized or malformed
942 61 number is too big
943 62 subpattern name expected
944 63 digit expected after (?+
945 64 ] is an invalid data character in JavaScript compatibility mode
946 65 different names for subpatterns of the same number are
947 not allowed
948 66 (*MARK) must have an argument
949 67 this version of PCRE is not compiled with Unicode property
950 support
951 68 \ec must be followed by an ASCII character
952 69 \ek is not followed by a braced, angle-bracketed, or quoted name
953 70 internal error: unknown opcode in find_fixedlength()
954 71 \eN is not supported in a class
955 72 too many forward references
956 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
957 74 invalid UTF-16 string (specifically UTF-16)
958 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
959 76 character value in \eu.... sequence is too large
960 77 invalid UTF-32 string (specifically UTF-32)
961 .sp
962 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
963 be used if the limits were changed when PCRE was built.
964 .
965 .
966 .\" HTML <a name="studyingapattern"></a>
967 .SH "STUDYING A PATTERN"
968 .rs
969 .sp
970 .nf
971 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
972 .B " const char **\fIerrptr\fP);"
973 .fi
974 .PP
975 If a compiled pattern is going to be used several times, it is worth spending
976 more time analyzing it in order to speed up the time taken for matching. The
977 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
978 argument. If studying the pattern produces additional information that will
979 help speed up matching, \fBpcre_study()\fP returns a pointer to a
980 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
981 results of the study.
982 .P
983 The returned value from \fBpcre_study()\fP can be passed directly to
984 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
985 also contains other fields that can be set by the caller before the block is
986 passed; these are described
987 .\" HTML <a href="#extradata">
988 .\" </a>
989 below
990 .\"
991 in the section on matching a pattern.
992 .P
993 If studying the pattern does not produce any useful information,
994 \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
995 calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
996 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
997 if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
998 returns a \fBpcre_extra\fP block even if studying did not find any additional
999 information. It may still return NULL, however, if an error occurs in
1000 \fBpcre_study()\fP.
1001 .P
1002 The second argument of \fBpcre_study()\fP contains option bits. There are three
1003 further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1004 .sp
1005 PCRE_STUDY_JIT_COMPILE
1006 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1007 PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1008 .sp
1009 If any of these are set, and the just-in-time compiler is available, the
1010 pattern is further compiled into machine code that executes much faster than
1011 the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
1012 compiler is not available, these options are ignored. All undefined bits in the
1013 \fIoptions\fP argument must be zero.
1014 .P
1015 JIT compilation is a heavyweight optimization. It can take some time for
1016 patterns to be analyzed, and for one-off matches and simple patterns the
1017 benefit of faster execution might be offset by a much slower study time.
1018 Not all patterns can be optimized by the JIT compiler. For those that cannot be
1019 handled, matching automatically falls back to the \fBpcre_exec()\fP
1020 interpreter. For more details, see the
1021 .\" HREF
1022 \fBpcrejit\fP
1023 .\"
1024 documentation.
1025 .P
1026 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
1027 studying succeeds (even if no data is returned), the variable it points to is
1028 set to NULL. Otherwise it is set to point to a textual error message. This is a
1029 static string that is part of the library. You must not try to free it. You
1030 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
1031 sure that it has run successfully.
1032 .P
1033 When you are finished with a pattern, you can free the memory used for the
1034 study data by calling \fBpcre_free_study()\fP. This function was added to the
1035 API for release 8.20. For earlier versions, the memory could be freed with
1036 \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1037 where JIT optimization is not used, but it is advisable to change to the new
1038 function when convenient.
1039 .P
1040 This is a typical way in which \fBpcre_study\fP() is used (except that in a
1041 real application there should be tests for errors):
1042 .sp
1043 int rc;
1044 pcre *re;
1045 pcre_extra *sd;
1046 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1047 sd = pcre_study(
1048 re, /* result of pcre_compile() */
1049 0, /* no options */
1050 &error); /* set to NULL or points to a message */
1051 rc = pcre_exec( /* see below for details of pcre_exec() options */
1052 re, sd, "subject", 7, 0, 0, ovector, 30);
1053 ...
1054 pcre_free_study(sd);
1055 pcre_free(re);
1056 .sp
1057 Studying a pattern does two things: first, a lower bound for the length of
1058 subject string that is needed to match the pattern is computed. This does not
1059 mean that there are any strings of that length that match, but it does
1060 guarantee that no shorter strings match. The value is used to avoid wasting
1061 time by trying to match strings that are shorter than the lower bound. You can
1062 find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
1063 .P
1064 Studying a pattern is also useful for non-anchored patterns that do not have a
1065 single fixed starting character. A bitmap of possible starting bytes is
1066 created. This speeds up finding a position in the subject at which to start
1067 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1068 In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1069 .P
1070 These two optimizations apply to both \fBpcre_exec()\fP and
1071 \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1072 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option.
1073 You might want to do this if your pattern contains callouts or (*MARK) and you
1074 want to make use of these facilities in cases where matching fails.
1075 .P
1076 PCRE_NO_START_OPTIMIZE can be specified at either compile time or execution
1077 time. However, if PCRE_NO_START_OPTIMIZE is passed to \fBpcre_exec()\fP, (that
1078 is, after any JIT compilation has happened) JIT execution is disabled. For JIT
1079 execution to work with PCRE_NO_START_OPTIMIZE, the option must be set at
1080 compile time.
1081 .P
1082 There is a longer discussion of PCRE_NO_START_OPTIMIZE
1083 .\" HTML <a href="#execoptions">
1084 .\" </a>
1085 below.
1086 .\"
1087 .
1088 .
1089 .\" HTML <a name="localesupport"></a>
1090 .SH "LOCALE SUPPORT"
1091 .rs
1092 .sp
1093 PCRE handles caseless matching, and determines whether characters are letters,
1094 digits, or whatever, by reference to a set of tables, indexed by character
1095 value. When running in UTF-8 mode, this applies only to characters
1096 with codes less than 128. By default, higher-valued codes never match escapes
1097 such as \ew or \ed, but they can be tested with \ep if PCRE is built with
1098 Unicode character property support. Alternatively, the PCRE_UCP option can be
1099 set at compile time; this causes \ew and friends to use Unicode property
1100 support instead of built-in tables. The use of locales with Unicode is
1101 discouraged. If you are handling characters with codes greater than 128, you
1102 should either use UTF-8 and Unicode, or use locales, but not try to mix the
1103 two.
1104 .P
1105 PCRE contains an internal set of tables that are used when the final argument
1106 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1107 Normally, the internal tables recognize only ASCII characters. However, when
1108 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1109 default "C" locale of the local system, which may cause them to be different.
1110 .P
1111 The internal tables can always be overridden by tables supplied by the
1112 application that calls PCRE. These may be created in a different locale from
1113 the default. As more and more applications change to using Unicode, the need
1114 for this locale support is expected to die away.
1115 .P
1116 External tables are built by calling the \fBpcre_maketables()\fP function,
1117 which has no arguments, in the relevant locale. The result can then be passed
1118 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
1119 example, to build and use tables that are appropriate for the French locale
1120 (where accented characters with values greater than 128 are treated as letters),
1121 the following code could be used:
1122 .sp
1123 setlocale(LC_CTYPE, "fr_FR");
1124 tables = pcre_maketables();
1125 re = pcre_compile(..., tables);
1126 .sp
1127 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1128 are using Windows, the name for the French locale is "french".
1129 .P
1130 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1131 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1132 that the memory containing the tables remains available for as long as it is
1133 needed.
1134 .P
1135 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1136 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1137 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
1138 pattern, compilation, studying and matching all happen in the same locale, but
1139 different patterns can be compiled in different locales.
1140 .P
1141 It is possible to pass a table pointer or NULL (indicating the use of the
1142 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
1143 this facility could be used to match a pattern in a different locale from the
1144 one in which it was compiled. Passing table pointers at run time is discussed
1145 below in the section on matching a pattern.
1146 .
1147 .
1148 .\" HTML <a name="infoaboutpattern"></a>
1149 .SH "INFORMATION ABOUT A PATTERN"
1150 .rs
1151 .sp
1152 .nf
1153 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1154 .B " int \fIwhat\fP, void *\fIwhere\fP);"
1155 .fi
1156 .PP
1157 The \fBpcre_fullinfo()\fP function returns information about a compiled
1158 pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1159 library at version 8.30, after more than 10 years of obsolescence.
1160 .P
1161 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1162 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1163 the pattern was not studied. The third argument specifies which piece of
1164 information is required, and the fourth argument is a pointer to a variable
1165 to receive the data. The yield of the function is zero for success, or one of
1166 the following negative numbers:
1167 .sp
1168 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1169 the argument \fIwhere\fP was NULL
1170 PCRE_ERROR_BADMAGIC the "magic number" was not found
1171 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1172 endianness
1173 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1174 PCRE_ERROR_UNSET the requested field is not set
1175 .sp
1176 The "magic number" is placed at the start of each compiled pattern as an simple
1177 check against passing an arbitrary memory pointer. The endianness error can
1178 occur if a compiled pattern is saved and reloaded on a different host. Here is
1179 a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1180 pattern:
1181 .sp
1182 int rc;
1183 size_t length;
1184 rc = pcre_fullinfo(
1185 re, /* result of pcre_compile() */
1186 sd, /* result of pcre_study(), or NULL */
1187 PCRE_INFO_SIZE, /* what is required */
1188 &length); /* where to put the data */
1189 .sp
1190 The possible values for the third argument are defined in \fBpcre.h\fP, and are
1191 as follows:
1192 .sp
1193 PCRE_INFO_BACKREFMAX
1194 .sp
1195 Return the number of the highest back reference in the pattern. The fourth
1196 argument should point to an \fBint\fP variable. Zero is returned if there are
1197 no back references.
1198 .sp
1199 PCRE_INFO_CAPTURECOUNT
1200 .sp
1201 Return the number of capturing subpatterns in the pattern. The fourth argument
1202 should point to an \fBint\fP variable.
1203 .sp
1204 PCRE_INFO_DEFAULT_TABLES
1205 .sp
1206 Return a pointer to the internal default character tables within PCRE. The
1207 fourth argument should point to an \fBunsigned char *\fP variable. This
1208 information call is provided for internal use by the \fBpcre_study()\fP
1209 function. External callers can cause PCRE to use its internal tables by passing
1210 a NULL table pointer.
1211 .sp
1212 PCRE_INFO_FIRSTBYTE
1213 .sp
1214 Return information about the first data unit of any matched string, for a
1215 non-anchored pattern. (The name of this option refers to the 8-bit library,
1216 where data units are bytes.) The fourth argument should point to an \fBint\fP
1217 variable.
1218 .P
1219 If there is a fixed first value, for example, the letter "c" from a pattern
1220 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1221 value is always less than 256. In the 16-bit library the value can be up to
1222 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1223 .P
1224 If there is no fixed first value, and if either
1225 .sp
1226 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1227 starts with "^", or
1228 .sp
1229 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1230 (if it were set, the pattern would be anchored),
1231 .sp
1232 -1 is returned, indicating that the pattern matches only at the start of a
1233 subject string or after any newline within the string. Otherwise -2 is
1234 returned. For anchored patterns, -2 is returned.
1235 .P
1236 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1237 to return the full 32-bit range of the character, this value is deprecated;
1238 instead the PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER values
1239 should be used.
1240 .sp
1241 PCRE_INFO_FIRSTTABLE
1242 .sp
1243 If the pattern was studied, and this resulted in the construction of a 256-bit
1244 table indicating a fixed set of values for the first data unit in any matching
1245 string, a pointer to the table is returned. Otherwise NULL is returned. The
1246 fourth argument should point to an \fBunsigned char *\fP variable.
1247 .sp
1248 PCRE_INFO_HASCRORLF
1249 .sp
1250 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1251 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1252 explicit match is either a literal CR or LF character, or \er or \en.
1253 .sp
1254 PCRE_INFO_JCHANGED
1255 .sp
1256 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1257 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1258 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1259 .sp
1260 PCRE_INFO_JIT
1261 .sp
1262 Return 1 if the pattern was studied with one of the JIT options, and
1263 just-in-time compiling was successful. The fourth argument should point to an
1264 \fBint\fP variable. A return value of 0 means that JIT support is not available
1265 in this version of PCRE, or that the pattern was not studied with a JIT option,
1266 or that the JIT compiler could not handle this particular pattern. See the
1267 .\" HREF
1268 \fBpcrejit\fP
1269 .\"
1270 documentation for details of what can and cannot be handled.
1271 .sp
1272 PCRE_INFO_JITSIZE
1273 .sp
1274 If the pattern was successfully studied with a JIT option, return the size of
1275 the JIT compiled code, otherwise return zero. The fourth argument should point
1276 to a \fBsize_t\fP variable.
1277 .sp
1278 PCRE_INFO_LASTLITERAL
1279 .sp
1280 Return the value of the rightmost literal data unit that must exist in any
1281 matched string, other than at its start, if such a value has been recorded. The
1282 fourth argument should point to an \fBint\fP variable. If there is no such
1283 value, -1 is returned. For anchored patterns, a last literal value is recorded
1284 only if it follows something of variable length. For example, for the pattern
1285 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1286 is -1.
1287 .P
1288 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1289 to return the full 32-bit range of the character, this value is deprecated;
1290 instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1291 be used.
1292 .sp
1293 PCRE_INFO_MATCHLIMIT
1294 .sp
1295 If the pattern set a match limit by including an item of the form
1296 (*LIMIT_MATCH=nnnn) at the start, the value is returned. The fourth argument
1297 should point to an unsigned 32-bit integer. If no such value has been set, the
1298 call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1299 .sp
1300 PCRE_INFO_MAXLOOKBEHIND
1301 .sp
1302 Return the number of characters (NB not data units) in the longest lookbehind
1303 assertion in the pattern. This information is useful when doing multi-segment
1304 matching using the partial matching facilities. Note that the simple assertions
1305 \eb and \eB require a one-character lookbehind. \eA also registers a
1306 one-character lookbehind, though it does not actually inspect the previous
1307 character. This is to ensure that at least one character from the old segment
1308 is retained when a new segment is processed. Otherwise, if there are no
1309 lookbehinds in the pattern, \eA might match incorrectly at the start of a new
1310 segment.
1311 .sp
1312 PCRE_INFO_MINLENGTH
1313 .sp
1314 If the pattern was studied and a minimum length for matching subject strings
1315 was computed, its value is returned. Otherwise the returned value is -1. The
1316 value is a number of characters, which in UTF mode may be different from the
1317 number of data units. The fourth argument should point to an \fBint\fP
1318 variable. A non-negative value is a lower bound to the length of any matching
1319 string. There may not be any strings of that length that do actually match, but
1320 every string that does match is at least that long.
1321 .sp
1322 PCRE_INFO_NAMECOUNT
1323 PCRE_INFO_NAMEENTRYSIZE
1324 PCRE_INFO_NAMETABLE
1325 .sp
1326 PCRE supports the use of named as well as numbered capturing parentheses. The
1327 names are just an additional way of identifying the parentheses, which still
1328 acquire numbers. Several convenience functions such as
1329 \fBpcre_get_named_substring()\fP are provided for extracting captured
1330 substrings by name. It is also possible to extract the data directly, by first
1331 converting the name to a number in order to access the correct pointers in the
1332 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1333 you need to use the name-to-number map, which is described by these three
1334 values.
1335 .P
1336 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1337 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1338 entry; both of these return an \fBint\fP value. The entry size depends on the
1339 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1340 entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1341 the first two bytes of each entry are the number of the capturing parenthesis,
1342 most significant byte first. In the 16-bit library, the pointer points to
1343 16-bit data units, the first of which contains the parenthesis number. In the
1344 32-bit library, the pointer points to 32-bit data units, the first of which
1345 contains the parenthesis number. The rest of the entry is the corresponding
1346 name, zero terminated.
1347 .P
1348 The names are in alphabetical order. Duplicate names may appear if (?| is used
1349 to create multiple groups with the same number, as described in the
1350 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1351 .\" </a>
1352 section on duplicate subpattern numbers
1353 .\"
1354 in the
1355 .\" HREF
1356 \fBpcrepattern\fP
1357 .\"
1358 page. Duplicate names for subpatterns with different numbers are permitted only
1359 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1360 table in the order in which they were found in the pattern. In the absence of
1361 (?| this is the order of increasing number; when (?| is used this is not
1362 necessarily the case because later subpatterns may have lower numbers.
1363 .P
1364 As a simple example of the name/number table, consider the following pattern
1365 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1366 space - including newlines - is ignored):
1367 .sp
1368 .\" JOIN
1369 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1370 (?<month>\ed\ed) - (?<day>\ed\ed) )
1371 .sp
1372 There are four named subpatterns, so the table has four entries, and each entry
1373 in the table is eight bytes long. The table is as follows, with non-printing
1374 bytes shows in hexadecimal, and undefined bytes shown as ??:
1375 .sp
1376 00 01 d a t e 00 ??
1377 00 05 d a y 00 ?? ??
1378 00 04 m o n t h 00
1379 00 02 y e a r 00 ??
1380 .sp
1381 When writing code to extract data from named subpatterns using the
1382 name-to-number map, remember that the length of the entries is likely to be
1383 different for each compiled pattern.
1384 .sp
1385 PCRE_INFO_OKPARTIAL
1386 .sp
1387 Return 1 if the pattern can be used for partial matching with
1388 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1389 \fBint\fP variable. From release 8.00, this always returns 1, because the
1390 restrictions that previously applied to partial matching have been lifted. The
1391 .\" HREF
1392 \fBpcrepartial\fP
1393 .\"
1394 documentation gives details of partial matching.
1395 .sp
1396 PCRE_INFO_OPTIONS
1397 .sp
1398 Return a copy of the options with which the pattern was compiled. The fourth
1399 argument should point to an \fBunsigned long int\fP variable. These option bits
1400 are those specified in the call to \fBpcre_compile()\fP, modified by any
1401 top-level option settings at the start of the pattern itself. In other words,
1402 they are the options that will be in force when matching starts. For example,
1403 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1404 result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1405 .P
1406 A pattern is automatically anchored by PCRE if all of its top-level
1407 alternatives begin with one of the following:
1408 .sp
1409 ^ unless PCRE_MULTILINE is set
1410 \eA always
1411 \eG always
1412 .\" JOIN
1413 .* if PCRE_DOTALL is set and there are no back
1414 references to the subpattern in which .* appears
1415 .sp
1416 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1417 \fBpcre_fullinfo()\fP.
1418 .sp
1419 PCRE_INFO_RECURSIONLIMIT
1420 .sp
1421 If the pattern set a recursion limit by including an item of the form
1422 (*LIMIT_RECURSION=nnnn) at the start, the value is returned. The fourth
1423 argument should point to an unsigned 32-bit integer. If no such value has been
1424 set, the call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1425 .sp
1426 PCRE_INFO_SIZE
1427 .sp
1428 Return the size of the compiled pattern in bytes (for all three libraries). The
1429 fourth argument should point to a \fBsize_t\fP variable. This value does not
1430 include the size of the \fBpcre\fP structure that is returned by
1431 \fBpcre_compile()\fP. The value that is passed as the argument to
1432 \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1433 place the compiled data is the value returned by this option plus the size of
1434 the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1435 does not alter the value returned by this option.
1436 .sp
1437 PCRE_INFO_STUDYSIZE
1438 .sp
1439 Return the size in bytes (for all three libraries) of the data block pointed to
1440 by the \fIstudy_data\fP field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP
1441 is NULL, or there is no study data, zero is returned. The fourth argument
1442 should point to a \fBsize_t\fP variable. The \fIstudy_data\fP field is set by
1443 \fBpcre_study()\fP to record information that will speed up matching (see the
1444 section entitled
1445 .\" HTML <a href="#studyingapattern">
1446 .\" </a>
1447 "Studying a pattern"
1448 .\"
1449 above). The format of the \fIstudy_data\fP block is private, but its length
1450 is made available via this option so that it can be saved and restored (see the
1451 .\" HREF
1452 \fBpcreprecompile\fP
1453 .\"
1454 documentation for details).
1455 .sp
1456 PCRE_INFO_FIRSTCHARACTERFLAGS
1457 .sp
1458 Return information about the first data unit of any matched string, for a
1459 non-anchored pattern. The fourth argument should point to an \fBint\fP
1460 variable.
1461 .P
1462 If there is a fixed first value, for example, the letter "c" from a pattern
1463 such as (cat|cow|coyote), 1 is returned, and the character value can be
1464 retrieved using PCRE_INFO_FIRSTCHARACTER.
1465 .P
1466 If there is no fixed first value, and if either
1467 .sp
1468 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1469 starts with "^", or
1470 .sp
1471 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1472 (if it were set, the pattern would be anchored),
1473 .sp
1474 2 is returned, indicating that the pattern matches only at the start of a
1475 subject string or after any newline within the string. Otherwise 0 is
1476 returned. For anchored patterns, 0 is returned.
1477 .sp
1478 PCRE_INFO_FIRSTCHARACTER
1479 .sp
1480 Return the fixed first character value, if PCRE_INFO_FIRSTCHARACTERFLAGS
1481 returned 1; otherwise returns 0. The fourth argument should point to an
1482 \fBuint_t\fP variable.
1483 .P
1484 In the 8-bit library, the value is always less than 256. In the 16-bit library
1485 the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1486 can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1487 .P
1488 If there is no fixed first value, and if either
1489 .sp
1490 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1491 starts with "^", or
1492 .sp
1493 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1494 (if it were set, the pattern would be anchored),
1495 .sp
1496 -1 is returned, indicating that the pattern matches only at the start of a
1497 subject string or after any newline within the string. Otherwise -2 is
1498 returned. For anchored patterns, -2 is returned.
1499 .sp
1500 PCRE_INFO_REQUIREDCHARFLAGS
1501 .sp
1502 Returns 1 if there is a rightmost literal data unit that must exist in any
1503 matched string, other than at its start. The fourth argument should point to
1504 an \fBint\fP variable. If there is no such value, 0 is returned. If returning
1505 1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1506 .P
1507 For anchored patterns, a last literal value is recorded only if it follows
1508 something of variable length. For example, for the pattern /^a\ed+z\ed+/ the
1509 returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1510 /^a\edz\ed/ the returned value is 0.
1511 .sp
1512 PCRE_INFO_REQUIREDCHAR
1513 .sp
1514 Return the value of the rightmost literal data unit that must exist in any
1515 matched string, other than at its start, if such a value has been recorded. The
1516 fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1517 value, 0 is returned.
1518 .
1519 .
1520 .SH "REFERENCE COUNTS"
1521 .rs
1522 .sp
1523 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1524 .PP
1525 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1526 data block that contains a compiled pattern. It is provided for the benefit of
1527 applications that operate in an object-oriented manner, where different parts
1528 of the application may be using the same compiled pattern, but you want to free
1529 the block when they are all done.
1530 .P
1531 When a pattern is compiled, the reference count field is initialized to zero.
1532 It is changed only by calling this function, whose action is to add the
1533 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1534 function is the new value. However, the value of the count is constrained to
1535 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1536 it is forced to the appropriate limit value.
1537 .P
1538 Except when it is zero, the reference count is not correctly preserved if a
1539 pattern is compiled on one host and then transferred to a host whose byte-order
1540 is different. (This seems a highly unlikely scenario.)
1541 .
1542 .
1543 .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1544 .rs
1545 .sp
1546 .nf
1547 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1548 .B " const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1549 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);"
1550 .fi
1551 .P
1552 The function \fBpcre_exec()\fP is called to match a subject string against a
1553 compiled pattern, which is passed in the \fIcode\fP argument. If the
1554 pattern was studied, the result of the study should be passed in the
1555 \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1556 and \fIextra\fP arguments as many times as you like, in order to match
1557 different subject strings with the same pattern.
1558 .P
1559 This function is the main matching facility of the library, and it operates in
1560 a Perl-like manner. For specialist use there is also an alternative matching
1561 function, which is described
1562 .\" HTML <a href="#dfamatch">
1563 .\" </a>
1564 below
1565 .\"
1566 in the section about the \fBpcre_dfa_exec()\fP function.
1567 .P
1568 In most applications, the pattern will have been compiled (and optionally
1569 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1570 possible to save compiled patterns and study data, and then use them later
1571 in different processes, possibly even on different hosts. For a discussion
1572 about this, see the
1573 .\" HREF
1574 \fBpcreprecompile\fP
1575 .\"
1576 documentation.
1577 .P
1578 Here is an example of a simple call to \fBpcre_exec()\fP:
1579 .sp
1580 int rc;
1581 int ovector[30];
1582 rc = pcre_exec(
1583 re, /* result of pcre_compile() */
1584 NULL, /* we didn't study the pattern */
1585 "some string", /* the subject string */
1586 11, /* the length of the subject string */
1587 0, /* start at offset 0 in the subject */
1588 0, /* default options */
1589 ovector, /* vector of integers for substring information */
1590 30); /* number of elements (NOT size in bytes) */
1591 .
1592 .
1593 .\" HTML <a name="extradata"></a>
1594 .SS "Extra data for \fBpcre_exec()\fR"
1595 .rs
1596 .sp
1597 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1598 data block. The \fBpcre_study()\fP function returns such a block (when it
1599 doesn't return NULL), but you can also create one for yourself, and pass
1600 additional information in it. The \fBpcre_extra\fP block contains the following
1601 fields (not necessarily in this order):
1602 .sp
1603 unsigned long int \fIflags\fP;
1604 void *\fIstudy_data\fP;
1605 void *\fIexecutable_jit\fP;
1606 unsigned long int \fImatch_limit\fP;
1607 unsigned long int \fImatch_limit_recursion\fP;
1608 void *\fIcallout_data\fP;
1609 const unsigned char *\fItables\fP;
1610 unsigned char **\fImark\fP;
1611 .sp
1612 In the 16-bit version of this structure, the \fImark\fP field has type
1613 "PCRE_UCHAR16 **".
1614 .sp
1615 In the 32-bit version of this structure, the \fImark\fP field has type
1616 "PCRE_UCHAR32 **".
1617 .P
1618 The \fIflags\fP field is used to specify which of the other fields are set. The
1619 flag bits are:
1620 .sp
1621 PCRE_EXTRA_CALLOUT_DATA
1622 PCRE_EXTRA_EXECUTABLE_JIT
1623 PCRE_EXTRA_MARK
1624 PCRE_EXTRA_MATCH_LIMIT
1625 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1626 PCRE_EXTRA_STUDY_DATA
1627 PCRE_EXTRA_TABLES
1628 .sp
1629 Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1630 the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1631 returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1632 should not set these yourself, but you may add to the block by setting other
1633 fields and their corresponding flag bits.
1634 .P
1635 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1636 vast amount of resources when running patterns that are not going to match,
1637 but which have a very large number of possibilities in their search trees. The
1638 classic example is a pattern that uses nested unlimited repeats.
1639 .P
1640 Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1641 calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1642 imposed on the number of times this function is called during a match, which
1643 has the effect of limiting the amount of backtracking that can take place. For
1644 patterns that are not anchored, the count restarts from zero for each position
1645 in the subject string.
1646 .P
1647 When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1648 with a JIT option, the way that the matching is executed is entirely different.
1649 However, there is still the possibility of runaway matching that goes on for a
1650 very long time, and so the \fImatch_limit\fP value is also used in this case
1651 (but in a different way) to limit how long the matching can continue.
1652 .P
1653 The default value for the limit can be set when PCRE is built; the default
1654 default is 10 million, which handles all but the most extreme cases. You can
1655 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1656 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1657 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1658 PCRE_ERROR_MATCHLIMIT.
1659 .P
1660 A value for the match limit may also be supplied by an item at the start of a
1661 pattern of the form
1662 .sp
1663 (*LIMIT_MATCH=d)
1664 .sp
1665 where d is a decimal number. However, such a setting is ignored unless d is
1666 less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1667 is set, less than the default.
1668 .P
1669 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1670 instead of limiting the total number of times that \fBmatch()\fP is called, it
1671 limits the depth of recursion. The recursion depth is a smaller number than the
1672 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1673 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1674 .P
1675 Limiting the recursion depth limits the amount of machine stack that can be
1676 used, or, when PCRE has been compiled to use memory on the heap instead of the
1677 stack, the amount of heap memory that can be used. This limit is not relevant,
1678 and is ignored, when matching is done using JIT compiled code.
1679 .P
1680 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1681 built; the default default is the same value as the default for
1682 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1683 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1684 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1685 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1686 .P
1687 A value for the recursion limit may also be supplied by an item at the start of
1688 a pattern of the form
1689 .sp
1690 (*LIMIT_RECURSION=d)
1691 .sp
1692 where d is a decimal number. However, such a setting is ignored unless d is
1693 less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1694 is set, less than the default.
1695 .P
1696 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1697 and is described in the
1698 .\" HREF
1699 \fBpcrecallout\fP
1700 .\"
1701 documentation.
1702 .P
1703 The \fItables\fP field is used to pass a character tables pointer to
1704 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1705 pattern. A non-NULL value is stored with the compiled pattern only if custom
1706 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1707 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1708 internal tables to be used. This facility is helpful when re-using patterns
1709 that have been saved after compiling with an external set of tables, because
1710 the external tables might be at a different address when \fBpcre_exec()\fP is
1711 called. See the
1712 .\" HREF
1713 \fBpcreprecompile\fP
1714 .\"
1715 documentation for a discussion of saving compiled patterns for later use.
1716 .P
1717 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1718 be set to point to a suitable variable. If the pattern contains any
1719 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1720 a name to pass back, a pointer to the name string (zero terminated) is placed
1721 in the variable pointed to by the \fImark\fP field. The names are within the
1722 compiled pattern; if you wish to retain such a name you must copy it before
1723 freeing the memory of a compiled pattern. If there is no name to pass back, the
1724 variable pointed to by the \fImark\fP field is set to NULL. For details of the
1725 backtracking control verbs, see the section entitled
1726 .\" HTML <a href="pcrepattern#backtrackcontrol">
1727 .\" </a>
1728 "Backtracking control"
1729 .\"
1730 in the
1731 .\" HREF
1732 \fBpcrepattern\fP
1733 .\"
1734 documentation.
1735 .
1736 .
1737 .\" HTML <a name="execoptions"></a>
1738 .SS "Option bits for \fBpcre_exec()\fP"
1739 .rs
1740 .sp
1741 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1742 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1743 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1744 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1745 PCRE_PARTIAL_SOFT.
1746 .P
1747 If the pattern was successfully studied with one of the just-in-time (JIT)
1748 compile options, the only supported options for JIT execution are
1749 PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1750 PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1751 unsupported option is used, JIT execution is disabled and the normal
1752 interpretive code in \fBpcre_exec()\fP is run.
1753 .sp
1754 PCRE_ANCHORED
1755 .sp
1756 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1757 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1758 to be anchored by virtue of its contents, it cannot be made unachored at
1759 matching time.
1760 .sp
1761 PCRE_BSR_ANYCRLF
1762 PCRE_BSR_UNICODE
1763 .sp
1764 These options (which are mutually exclusive) control what the \eR escape
1765 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1766 match any Unicode newline sequence. These options override the choice that was
1767 made or defaulted when the pattern was compiled.
1768 .sp
1769 PCRE_NEWLINE_CR
1770 PCRE_NEWLINE_LF
1771 PCRE_NEWLINE_CRLF
1772 PCRE_NEWLINE_ANYCRLF
1773 PCRE_NEWLINE_ANY
1774 .sp
1775 These options override the newline definition that was chosen or defaulted when
1776 the pattern was compiled. For details, see the description of
1777 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1778 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1779 the way the match position is advanced after a match failure for an unanchored
1780 pattern.
1781 .P
1782 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1783 match attempt for an unanchored pattern fails when the current position is at a
1784 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1785 characters, the match position is advanced by two characters instead of one, in
1786 other words, to after the CRLF.
1787 .P
1788 The above rule is a compromise that makes the most common cases work as
1789 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1790 set), it does not match the string "\er\enA" because, after failing at the
1791 start, it skips both the CR and the LF before retrying. However, the pattern
1792 [\er\en]A does match that string, because it contains an explicit CR or LF
1793 reference, and so advances only by one character after the first failure.
1794 .P
1795 An explicit match for CR of LF is either a literal appearance of one of those
1796 characters, or one of the \er or \en escape sequences. Implicit matches such as
1797 [^X] do not count, nor does \es (which includes CR and LF in the characters
1798 that it matches).
1799 .P
1800 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1801 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1802 .sp
1803 PCRE_NOTBOL
1804 .sp
1805 This option specifies that first character of the subject string is not the
1806 beginning of a line, so the circumflex metacharacter should not match before
1807 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1808 never to match. This option affects only the behaviour of the circumflex
1809 metacharacter. It does not affect \eA.
1810 .sp
1811 PCRE_NOTEOL
1812 .sp
1813 This option specifies that the end of the subject string is not the end of a
1814 line, so the dollar metacharacter should not match it nor (except in multiline
1815 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1816 compile time) causes dollar never to match. This option affects only the
1817 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1818 .sp
1819 PCRE_NOTEMPTY
1820 .sp
1821 An empty string is not considered to be a valid match if this option is set. If
1822 there are alternatives in the pattern, they are tried. If all the alternatives
1823 match the empty string, the entire match fails. For example, if the pattern
1824 .sp
1825 a?b?
1826 .sp
1827 is applied to a string not beginning with "a" or "b", it matches an empty
1828 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1829 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1830 .sp
1831 PCRE_NOTEMPTY_ATSTART
1832 .sp
1833 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1834 the start of the subject is permitted. If the pattern is anchored, such a match
1835 can occur only if the pattern contains \eK.
1836 .P
1837 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1838 does make a special case of a pattern match of the empty string within its
1839 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1840 emulate Perl's behaviour after matching a null string by first trying the match
1841 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1842 if that fails, by advancing the starting offset (see below) and trying an
1843 ordinary match again. There is some code that demonstrates how to do this in
1844 the
1845 .\" HREF
1846 \fBpcredemo\fP
1847 .\"
1848 sample program. In the most general case, you have to check to see if the
1849 newline convention recognizes CRLF as a newline, and if so, and the current
1850 character is CR followed by LF, advance the starting offset by two characters
1851 instead of one.
1852 .sp
1853 PCRE_NO_START_OPTIMIZE
1854 .sp
1855 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1856 a match, in order to speed up the process. For example, if it is known that an
1857 unanchored match must start with a specific character, it searches the subject
1858 for that character, and fails immediately if it cannot find it, without
1859 actually running the main matching function. This means that a special item
1860 such as (*COMMIT) at the start of a pattern is not considered until after a
1861 suitable starting point for the match has been found. Also, when callouts or
1862 (*MARK) items are in use, these "start-up" optimizations can cause them to be
1863 skipped if the pattern is never actually used. The start-up optimizations are
1864 in effect a pre-scan of the subject that takes place before the pattern is run.
1865 .P
1866 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1867 causing performance to suffer, but ensuring that in cases where the result is
1868 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1869 are considered at every possible starting position in the subject string. If
1870 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1871 time. The use of PCRE_NO_START_OPTIMIZE at matching time (that is, passing it
1872 to \fBpcre_exec()\fP) disables JIT execution; in this situation, matching is
1873 always done using interpretively.
1874 .P
1875 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1876 Consider the pattern
1877 .sp
1878 (*COMMIT)ABC
1879 .sp
1880 When this is compiled, PCRE records the fact that a match must start with the
1881 character "A". Suppose the subject string is "DEFABC". The start-up
1882 optimization scans along the subject, finds "A" and runs the first match
1883 attempt from there. The (*COMMIT) item means that the pattern must match the
1884 current starting position, which in this case, it does. However, if the same
1885 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1886 subject string does not happen. The first match attempt is run starting from
1887 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1888 the overall result is "no match". If the pattern is studied, more start-up
1889 optimizations may be used. For example, a minimum length for the subject may be
1890 recorded. Consider the pattern
1891 .sp
1892 (*MARK:A)(X|Y)
1893 .sp
1894 The minimum length for a match is one character. If the subject is "ABC", there
1895 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1896 If the pattern is studied, the final attempt does not take place, because PCRE
1897 knows that the subject is too short, and so the (*MARK) is never encountered.
1898 In this case, studying the pattern does not affect the overall match result,
1899 which is still "no match", but it does affect the auxiliary information that is
1900 returned.
1901 .sp
1902 PCRE_NO_UTF8_CHECK
1903 .sp
1904 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1905 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1906 The entire string is checked before any other processing takes place. The value
1907 of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1908 UTF-8 character. There is a discussion about the
1909 .\" HTML <a href="pcreunicode.html#utf8strings">
1910 .\" </a>
1911 validity of UTF-8 strings
1912 .\"
1913 in the
1914 .\" HREF
1915 \fBpcreunicode\fP
1916 .\"
1917 page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1918 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1919 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1920 cases, information about the precise nature of the error may also be returned
1921 (see the descriptions of these errors in the section entitled \fIError return
1922 values from\fP \fBpcre_exec()\fP
1923 .\" HTML <a href="#errorlist">
1924 .\" </a>
1925 below).
1926 .\"
1927 If \fIstartoffset\fP contains a value that does not point to the start of a
1928 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1929 returned.
1930 .P
1931 If you already know that your subject is valid, and you want to skip these
1932 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1933 calling \fBpcre_exec()\fP. You might want to do this for the second and
1934 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1935 all the matches in a single subject string. However, you should be sure that
1936 the value of \fIstartoffset\fP points to the start of a character (or the end
1937 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1938 invalid string as a subject or an invalid value of \fIstartoffset\fP is
1939 undefined. Your program may crash.
1940 .sp
1941 PCRE_PARTIAL_HARD
1942 PCRE_PARTIAL_SOFT
1943 .sp
1944 These options turn on the partial matching feature. For backwards
1945 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1946 occurs if the end of the subject string is reached successfully, but there are
1947 not enough subject characters to complete the match. If this happens when
1948 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1949 testing any remaining alternatives. Only if no complete match can be found is
1950 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1951 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1952 but only if no complete match can be found.
1953 .P
1954 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1955 partial match is found, \fBpcre_exec()\fP immediately returns
1956 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1957 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1958 important that an alternative complete match.
1959 .P
1960 In both cases, the portion of the string that was inspected when the partial
1961 match was found is set as the first matching string. There is a more detailed
1962 discussion of partial and multi-segment matching, with examples, in the
1963 .\" HREF
1964 \fBpcrepartial\fP
1965 .\"
1966 documentation.
1967 .
1968 .
1969 .SS "The string to be matched by \fBpcre_exec()\fP"
1970 .rs
1971 .sp
1972 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1973 \fIsubject\fP, a length in \fIlength\fP, and a starting offset in
1974 \fIstartoffset\fP. The units for \fIlength\fP and \fIstartoffset\fP are bytes
1975 for the 8-bit library, 16-bit data items for the 16-bit library, and 32-bit
1976 data items for the 32-bit library.
1977 .P
1978 If \fIstartoffset\fP is negative or greater than the length of the subject,
1979 \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting offset is
1980 zero, the search for a match starts at the beginning of the subject, and this
1981 is by far the most common case. In UTF-8 or UTF-16 mode, the offset must point
1982 to the start of a character, or the end of the subject (in UTF-32 mode, one
1983 data unit equals one character, so all offsets are valid). Unlike the pattern
1984 string, the subject may contain binary zeroes.
1985 .P
1986 A non-zero starting offset is useful when searching for another match in the
1987 same subject by calling \fBpcre_exec()\fP again after a previous success.
1988 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1989 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1990 lookbehind. For example, consider the pattern
1991 .sp
1992 \eBiss\eB
1993 .sp
1994 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1995 the current position in the subject is not a word boundary.) When applied to
1996 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1997 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1998 subject, namely "issipi", it does not match, because \eB is always false at the
1999 start of the subject, which is deemed to be a word boundary. However, if
2000 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
2001 set to 4, it finds the second occurrence of "iss" because it is able to look
2002 behind the starting point to discover that it is preceded by a letter.
2003 .P
2004 Finding all the matches in a subject is tricky when the pattern can match an
2005 empty string. It is possible to emulate Perl's /g behaviour by first trying the
2006 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
2007 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
2008 and trying an ordinary match again. There is some code that demonstrates how to
2009 do this in the
2010 .\" HREF
2011 \fBpcredemo\fP
2012 .\"
2013 sample program. In the most general case, you have to check to see if the
2014 newline convention recognizes CRLF as a newline, and if so, and the current
2015 character is CR followed by LF, advance the starting offset by two characters
2016 instead of one.
2017 .P
2018 If a non-zero starting offset is passed when the pattern is anchored, one
2019 attempt to match at the given offset is made. This can only succeed if the
2020 pattern does not require the match to be at the start of the subject.
2021 .
2022 .
2023 .SS "How \fBpcre_exec()\fP returns captured substrings"
2024 .rs
2025 .sp
2026 In general, a pattern matches a certain portion of the subject, and in
2027 addition, further substrings from the subject may be picked out by parts of the
2028 pattern. Following the usage in Jeffrey Friedl's book, this is called
2029 "capturing" in what follows, and the phrase "capturing subpattern" is used for
2030 a fragment of a pattern that picks out a substring. PCRE supports several other
2031 kinds of parenthesized subpattern that do not cause substrings to be captured.
2032 .P
2033 Captured substrings are returned to the caller via a vector of integers whose
2034 address is passed in \fIovector\fP. The number of elements in the vector is
2035 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
2036 argument is NOT the size of \fIovector\fP in bytes.
2037 .P
2038 The first two-thirds of the vector is used to pass back captured substrings,
2039 each substring using a pair of integers. The remaining third of the vector is
2040 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
2041 and is not available for passing back information. The number passed in
2042 \fIovecsize\fP should always be a multiple of three. If it is not, it is
2043 rounded down.
2044 .P
2045 When a match is successful, information about captured substrings is returned
2046 in pairs of integers, starting at the beginning of \fIovector\fP, and
2047 continuing up to two-thirds of its length at the most. The first element of
2048 each pair is set to the offset of the first character in a substring, and the
2049 second is set to the offset of the first character after the end of a
2050 substring. These values are always data unit offsets, even in UTF mode. They
2051 are byte offsets in the 8-bit library, 16-bit data item offsets in the 16-bit
2052 library, and 32-bit data item offsets in the 32-bit library. \fBNote\fP: they
2053 are not character counts.
2054 .P
2055 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
2056 portion of the subject string matched by the entire pattern. The next pair is
2057 used for the first capturing subpattern, and so on. The value returned by
2058 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
2059 For example, if two substrings have been captured, the returned value is 3. If
2060 there are no capturing subpatterns, the return value from a successful match is
2061 1, indicating that just the first pair of offsets has been set.
2062 .P
2063 If a capturing subpattern is matched repeatedly, it is the last portion of the
2064 string that it matched that is returned.
2065 .P
2066 If the vector is too small to hold all the captured substring offsets, it is
2067 used as far as possible (up to two-thirds of its length), and the function
2068 returns a value of zero. If neither the actual string matched nor any captured
2069 substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2070 passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2071 back references and the \fIovector\fP is not big enough to remember the related
2072 substrings, PCRE has to get additional memory for use during matching. Thus it
2073 is usually advisable to supply an \fIovector\fP of reasonable size.
2074 .P
2075 There are some cases where zero is returned (indicating vector overflow) when
2076 in fact the vector is exactly the right size for the final match. For example,
2077 consider the pattern
2078 .sp
2079 (a)(?:(b)c|bd)
2080 .sp
2081 If a vector of 6 elements (allowing for only 1 captured substring) is given
2082 with subject string "abd", \fBpcre_exec()\fP will try to set the second
2083 captured string, thereby recording a vector overflow, before failing to match
2084 "c" and backing up to try the second alternative. The zero return, however,
2085 does correctly indicate that the maximum number of slots (namely 2) have been
2086 filled. In similar cases where there is temporary overflow, but the final
2087 number of used slots is actually less than the maximum, a non-zero value is
2088 returned.
2089 .P
2090 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
2091 subpatterns there are in a compiled pattern. The smallest size for
2092 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
2093 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
2094 .P
2095 It is possible for capturing subpattern number \fIn+1\fP to match some part of
2096 the subject when subpattern \fIn\fP has not been used at all. For example, if
2097 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2098 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2099 happens, both values in the offset pairs corresponding to unused subpatterns
2100 are set to -1.
2101 .P
2102 Offset values that correspond to unused subpatterns at the end of the
2103 expression are also set to -1. For example, if the string "abc" is matched
2104 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2105 return from the function is 2, because the highest used capturing subpattern
2106 number is 1, and the offsets for for the second and third capturing subpatterns
2107 (assuming the vector is large enough, of course) are set to -1.
2108 .P
2109 \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
2110 correspond to capturing parentheses in the pattern are never changed. That is,
2111 if a pattern contains \fIn\fP capturing parentheses, no more than
2112 \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
2113 elements (in the first two-thirds) retain whatever values they previously had.
2114 .P
2115 Some convenience functions are provided for extracting the captured substrings
2116 as separate strings. These are described below.
2117 .
2118 .
2119 .\" HTML <a name="errorlist"></a>
2120 .SS "Error return values from \fBpcre_exec()\fP"
2121 .rs
2122 .sp
2123 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
2124 defined in the header file:
2125 .sp
2126 PCRE_ERROR_NOMATCH (-1)
2127 .sp
2128 The subject string did not match the pattern.
2129 .sp
2130 PCRE_ERROR_NULL (-2)
2131 .sp
2132 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
2133 NULL and \fIovecsize\fP was not zero.
2134 .sp
2135 PCRE_ERROR_BADOPTION (-3)
2136 .sp
2137 An unrecognized bit was set in the \fIoptions\fP argument.
2138 .sp
2139 PCRE_ERROR_BADMAGIC (-4)
2140 .sp
2141 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2142 the case when it is passed a junk pointer and to detect when a pattern that was
2143 compiled in an environment of one endianness is run in an environment with the
2144 other endianness. This is the error that PCRE gives when the magic number is
2145 not present.
2146 .sp
2147 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2148 .sp
2149 While running the pattern match, an unknown item was encountered in the
2150 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2151 of the compiled pattern.
2152 .sp
2153 PCRE_ERROR_NOMEMORY (-6)
2154 .sp
2155 If a pattern contains back references, but the \fIovector\fP that is passed to
2156 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2157 gets a block of memory at the start of matching to use for this purpose. If the
2158 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2159 automatically freed at the end of matching.
2160 .P
2161 This error is also given if \fBpcre_stack_malloc()\fP fails in
2162 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2163 \fB--disable-stack-for-recursion\fP.
2164 .sp
2165 PCRE_ERROR_NOSUBSTRING (-7)
2166 .sp
2167 This error is used by the \fBpcre_copy_substring()\fP,
2168 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2169 below). It is never returned by \fBpcre_exec()\fP.
2170 .sp
2171 PCRE_ERROR_MATCHLIMIT (-8)
2172 .sp
2173 The backtracking limit, as specified by the \fImatch_limit\fP field in a
2174 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2175 above.
2176 .sp
2177 PCRE_ERROR_CALLOUT (-9)
2178 .sp
2179 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2180 use by callout functions that want to yield a distinctive error code. See the
2181 .\" HREF
2182 \fBpcrecallout\fP
2183 .\"
2184 documentation for details.
2185 .sp
2186 PCRE_ERROR_BADUTF8 (-10)
2187 .sp
2188 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2189 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2190 (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2191 UTF-8 character is placed in the first element, and a reason code is placed in
2192 the second element. The reason codes are listed in the
2193 .\" HTML <a href="#badutf8reasons">
2194 .\" </a>
2195 following section.
2196 .\"
2197 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2198 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2199 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2200 .sp
2201 PCRE_ERROR_BADUTF8_OFFSET (-11)
2202 .sp
2203 The UTF-8 byte sequence that was passed as a subject was checked and found to
2204 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2205 \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2206 end of the subject.
2207 .sp
2208 PCRE_ERROR_PARTIAL (-12)
2209 .sp
2210 The subject string did not match, but it did match partially. See the
2211 .\" HREF
2212 \fBpcrepartial\fP
2213 .\"
2214 documentation for details of partial matching.
2215 .sp
2216 PCRE_ERROR_BADPARTIAL (-13)
2217 .sp
2218 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2219 option was used with a compiled pattern containing items that were not
2220 supported for partial matching. From release 8.00 onwards, there are no
2221 restrictions on partial matching.
2222 .sp
2223 PCRE_ERROR_INTERNAL (-14)
2224 .sp
2225 An unexpected internal error has occurred. This error could be caused by a bug
2226 in PCRE or by overwriting of the compiled pattern.
2227 .sp
2228 PCRE_ERROR_BADCOUNT (-15)
2229 .sp
2230 This error is given if the value of the \fIovecsize\fP argument is negative.
2231 .sp
2232 PCRE_ERROR_RECURSIONLIMIT (-21)
2233 .sp
2234 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2235 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2236 description above.
2237 .sp
2238 PCRE_ERROR_BADNEWLINE (-23)
2239 .sp
2240 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2241 .sp
2242 PCRE_ERROR_BADOFFSET (-24)
2243 .sp
2244 The value of \fIstartoffset\fP was negative or greater than the length of the
2245 subject, that is, the value in \fIlength\fP.
2246 .sp
2247 PCRE_ERROR_SHORTUTF8 (-25)
2248 .sp
2249 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2250 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2251 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2252 fact sufficient to detect this case, but this special error code for
2253 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2254 retained for backwards compatibility.
2255 .sp
2256 PCRE_ERROR_RECURSELOOP (-26)
2257 .sp
2258 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2259 the pattern. Specifically, it means that either the whole pattern or a
2260 subpattern has been called recursively for the second time at the same position
2261 in the subject string. Some simple patterns that might do this are detected and
2262 faulted at compile time, but more complicated cases, in particular mutual
2263 recursions between two different subpatterns, cannot be detected until run
2264 time.
2265 .sp
2266 PCRE_ERROR_JIT_STACKLIMIT (-27)
2267 .sp
2268 This error is returned when a pattern that was successfully studied using a
2269 JIT compile option is being matched, but the memory available for the
2270 just-in-time processing stack is not large enough. See the
2271 .\" HREF
2272 \fBpcrejit\fP
2273 .\"
2274 documentation for more details.
2275 .sp
2276 PCRE_ERROR_BADMODE (-28)
2277 .sp
2278 This error is given if a pattern that was compiled by the 8-bit library is
2279 passed to a 16-bit or 32-bit library function, or vice versa.
2280 .sp
2281 PCRE_ERROR_BADENDIANNESS (-29)
2282 .sp
2283 This error is given if a pattern that was compiled and saved is reloaded on a
2284 host with different endianness. The utility function
2285 \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2286 so that it runs on the new host.
2287 .sp
2288 PCRE_ERROR_JIT_BADOPTION
2289 .sp
2290 This error is returned when a pattern that was successfully studied using a JIT
2291 compile option is being matched, but the matching mode (partial or complete
2292 match) does not correspond to any JIT compilation mode. When the JIT fast path
2293 function is used, this error may be also given for invalid options. See the
2294 .\" HREF
2295 \fBpcrejit\fP
2296 .\"
2297 documentation for more details.
2298 .sp
2299 PCRE_ERROR_BADLENGTH (-32)
2300 .sp
2301 This error is given if \fBpcre_exec()\fP is called with a negative value for
2302 the \fIlength\fP argument.
2303 .P
2304 Error numbers -16 to -20, -22, and 30 are not used by \fBpcre_exec()\fP.
2305 .
2306 .
2307 .\" HTML <a name="badutf8reasons"></a>
2308 .SS "Reason codes for invalid UTF-8 strings"
2309 .rs
2310 .sp
2311 This section applies only to the 8-bit library. The corresponding information
2312 for the 16-bit and 32-bit libraries is given in the
2313 .\" HREF
2314 \fBpcre16\fP
2315 .\"
2316 and
2317 .\" HREF
2318 \fBpcre32\fP
2319 .\"
2320 pages.
2321 .P
2322 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2323 PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2324 least 2, the offset of the start of the invalid UTF-8 character is placed in
2325 the first output vector element (\fIovector[0]\fP) and a reason code is placed
2326 in the second element (\fIovector[1]\fP). The reason codes are given names in
2327 the \fBpcre.h\fP header file:
2328 .sp
2329 PCRE_UTF8_ERR1
2330 PCRE_UTF8_ERR2
2331 PCRE_UTF8_ERR3
2332 PCRE_UTF8_ERR4
2333 PCRE_UTF8_ERR5
2334 .sp
2335 The string ends with a truncated UTF-8 character; the code specifies how many
2336 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2337 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2338 allows for up to 6 bytes, and this is checked first; hence the possibility of
2339 4 or 5 missing bytes.
2340 .sp
2341 PCRE_UTF8_ERR6
2342 PCRE_UTF8_ERR7
2343 PCRE_UTF8_ERR8
2344 PCRE_UTF8_ERR9
2345 PCRE_UTF8_ERR10
2346 .sp
2347 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2348 character do not have the binary value 0b10 (that is, either the most
2349 significant bit is 0, or the next bit is 1).
2350 .sp
2351 PCRE_UTF8_ERR11
2352 PCRE_UTF8_ERR12
2353 .sp
2354 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2355 these code points are excluded by RFC 3629.
2356 .sp
2357 PCRE_UTF8_ERR13
2358 .sp
2359 A 4-byte character has a value greater than 0x10fff; these code points are
2360 excluded by RFC 3629.
2361 .sp
2362 PCRE_UTF8_ERR14
2363 .sp
2364 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2365 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2366 from UTF-8.
2367 .sp
2368 PCRE_UTF8_ERR15
2369 PCRE_UTF8_ERR16
2370 PCRE_UTF8_ERR17
2371 PCRE_UTF8_ERR18
2372 PCRE_UTF8_ERR19
2373 .sp
2374 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2375 value that can be represented by fewer bytes, which is invalid. For example,
2376 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2377 one byte.
2378 .sp
2379 PCRE_UTF8_ERR20
2380 .sp
2381 The two most significant bits of the first byte of a character have the binary
2382 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2383 byte can only validly occur as the second or subsequent byte of a multi-byte
2384 character.
2385 .sp
2386 PCRE_UTF8_ERR21
2387 .sp
2388 The first byte of a character has the value 0xfe or 0xff. These values can
2389 never occur in a valid UTF-8 string.
2390 .sp
2391 PCRE_UTF8_ERR22
2392 .sp
2393 This error code was formerly used when the presence of a so-called
2394 "non-character" caused an error. Unicode corrigendum #9 makes it clear that
2395 such characters should not cause a string to be rejected, and so this code is
2396 no longer in use and is never returned.
2397 .
2398 .
2399 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2400 .rs
2401 .sp
2402 .nf
2403 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2404 .B " int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,"
2405 .B " int \fIbuffersize\fP);"
2406 .sp
2407 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2408 .B " int \fIstringcount\fP, int \fIstringnumber\fP,"
2409 .B " const char **\fIstringptr\fP);"
2410 .sp
2411 .B int pcre_get_substring_list(const char *\fIsubject\fP,
2412 .B " int *\fIovector\fP, int \fIstringcount\fP, const char ***\fIlistptr\fP);"
2413 .fi
2414 .PP
2415 Captured substrings can be accessed directly by using the offsets returned by
2416 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2417 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2418 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2419 as new, separate, zero-terminated strings. These functions identify substrings
2420 by number. The next section describes functions for extracting named
2421 substrings.
2422 .P
2423 A substring that contains a binary zero is correctly extracted and has a
2424 further zero added on the end, but the result is not, of course, a C string.
2425 However, you can process such a string by referring to the length that is
2426 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2427 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2428 for handling strings containing binary zeros, because the end of the final
2429 string is not independently indicated.
2430 .P
2431 The first three arguments are the same for all three of these functions:
2432 \fIsubject\fP is the subject string that has just been successfully matched,
2433 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2434 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2435 captured by the match, including the substring that matched the entire regular
2436 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2437 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2438 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2439 number of elements in the vector divided by three.
2440 .P
2441 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2442 extract a single substring, whose number is given as \fIstringnumber\fP. A
2443 value of zero extracts the substring that matched the entire pattern, whereas
2444 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2445 the string is placed in \fIbuffer\fP, whose length is given by
2446 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2447 obtained via \fBpcre_malloc\fP, and its address is returned via
2448 \fIstringptr\fP. The yield of the function is the length of the string, not
2449 including the terminating zero, or one of these error codes:
2450 .sp
2451 PCRE_ERROR_NOMEMORY (-6)
2452 .sp
2453 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2454 memory failed for \fBpcre_get_substring()\fP.
2455 .sp
2456 PCRE_ERROR_NOSUBSTRING (-7)
2457 .sp
2458 There is no substring whose number is \fIstringnumber\fP.
2459 .P
2460 The \fBpcre_get_substring_list()\fP function extracts all available substrings
2461 and builds a list of pointers to them. All this is done in a single block of
2462 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2463 is returned via \fIlistptr\fP, which is also the start of the list of string
2464 pointers. The end of the list is marked by a NULL pointer. The yield of the
2465 function is zero if all went well, or the error code
2466 .sp
2467 PCRE_ERROR_NOMEMORY (-6)
2468 .sp
2469 if the attempt to get the memory block failed.
2470 .P
2471 When any of these functions encounter a substring that is unset, which can
2472 happen when capturing subpattern number \fIn+1\fP matches some part of the
2473 subject, but subpattern \fIn\fP has not been used at all, they return an empty
2474 string. This can be distinguished from a genuine zero-length substring by
2475 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2476 substrings.
2477 .P
2478 The two convenience functions \fBpcre_free_substring()\fP and
2479 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2480 a previous call of \fBpcre_get_substring()\fP or
2481 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2482 the function pointed to by \fBpcre_free\fP, which of course could be called
2483 directly from a C program. However, PCRE is used in some situations where it is
2484 linked via a special interface to another programming language that cannot use
2485 \fBpcre_free\fP directly; it is for these cases that the functions are
2486 provided.
2487 .
2488 .
2489 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2490 .rs
2491 .sp
2492 .nf
2493 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2494 .B " const char *\fIname\fP);"
2495 .sp
2496 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2497 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
2498 .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
2499 .B " char *\fIbuffer\fP, int \fIbuffersize\fP);"
2500 .sp
2501 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2502 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
2503 .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
2504 .B " const char **\fIstringptr\fP);"
2505 .fi
2506 .PP
2507 To extract a substring by name, you first have to find associated number.
2508 For example, for this pattern
2509 .sp
2510 (a+)b(?<xxx>\ed+)...
2511 .sp
2512 the number of the subpattern called "xxx" is 2. If the name is known to be
2513 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2514 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2515 pattern, and the second is the name. The yield of the function is the
2516 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2517 that name.
2518 .P
2519 Given the number, you can extract the substring directly, or use one of the
2520 functions described in the previous section. For convenience, there are also
2521 two functions that do the whole job.
2522 .P
2523 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2524 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2525 functions that extract by number. As these are described in the previous
2526 section, they are not re-described here. There are just two differences:
2527 .P
2528 First, instead of a substring number, a substring name is given. Second, there
2529 is an extra argument, given at the start, which is a pointer to the compiled
2530 pattern. This is needed in order to gain access to the name-to-number
2531 translation table.
2532 .P
2533 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2534 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2535 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2536 the behaviour may not be what you want (see the next section).
2537 .P
2538 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2539 subpatterns with the same number, as described in the
2540 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2541 .\" </a>
2542 section on duplicate subpattern numbers
2543 .\"
2544 in the
2545 .\" HREF
2546 \fBpcrepattern\fP
2547 .\"
2548 page, you cannot use names to distinguish the different subpatterns, because
2549 names are not included in the compiled code. The matching process uses only
2550 numbers. For this reason, the use of different names for subpatterns of the
2551 same number causes an error at compile time.
2552 .
2553 .
2554 .SH "DUPLICATE SUBPATTERN NAMES"
2555 .rs
2556 .sp
2557 .nf
2558 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2559 .B " const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);"
2560 .fi
2561 .PP
2562 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2563 are not required to be unique. (Duplicate names are always allowed for
2564 subpatterns with the same number, created by using the (?| feature. Indeed, if
2565 such subpatterns are named, they are required to use the same names.)
2566 .P
2567 Normally, patterns with duplicate names are such that in any one match, only
2568 one of the named subpatterns participates. An example is shown in the
2569 .\" HREF
2570 \fBpcrepattern\fP
2571 .\"
2572 documentation.
2573 .P
2574 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2575 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2576 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2577 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2578 returns one of the numbers that are associated with the name, but it is not
2579 defined which it is.
2580 .P
2581 If you want to get full details of all captured substrings for a given name,
2582 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2583 argument is the compiled pattern, and the second is the name. The third and
2584 fourth are pointers to variables which are updated by the function. After it
2585 has run, they point to the first and last entries in the name-to-number table
2586 for the given name. The function itself returns the length of each entry, or
2587 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2588 described above in the section entitled \fIInformation about a pattern\fP
2589 .\" HTML <a href="#infoaboutpattern">
2590 .\" </a>
2591 above.
2592 .\"
2593 Given all the relevant entries for the name, you can extract each of their
2594 numbers, and hence the captured data, if any.
2595 .
2596 .
2597 .SH "FINDING ALL POSSIBLE MATCHES"
2598 .rs
2599 .sp
2600 The traditional matching function uses a similar algorithm to Perl, which stops
2601 when it finds the first match, starting at a given point in the subject. If you
2602 want to find all possible matches, or the longest possible match, consider
2603 using the alternative matching function (see below) instead. If you cannot use
2604 the alternative function, but still need to find all possible matches, you
2605 can kludge it up by making use of the callout facility, which is described in
2606 the
2607 .\" HREF
2608 \fBpcrecallout\fP
2609 .\"
2610 documentation.
2611 .P
2612 What you have to do is to insert a callout right at the end of the pattern.
2613 When your callout function is called, extract and save the current matched
2614 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2615 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2616 will yield PCRE_ERROR_NOMATCH.
2617 .
2618 .
2619 .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2620 .rs
2621 .sp
2622 Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2623 stack, which in certain environments can be rather limited in size. Some users
2624 find it helpful to have an estimate of the amount of stack that is used by
2625 \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2626 .\" HREF
2627 \fBpcrestack\fP
2628 .\"
2629 documentation. The estimate that is output by \fBpcretest\fP when called with
2630 the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2631 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2632 .P
2633 Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2634 the negative error code PCRE_ERROR_NULL, but with this special combination of
2635 arguments, it returns instead a negative number whose absolute value is the
2636 approximate stack frame size in bytes. (A negative number is used so that it is
2637 clear that no match has happened.) The value is approximate because in some
2638 cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2639 additional variables on the stack.
2640 .P
2641 If PCRE has been compiled to use the heap instead of the stack for recursion,
2642 the value returned is the size of each block that is obtained from the heap.
2643 .
2644 .
2645 .\" HTML <a name="dfamatch"></a>
2646 .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2647 .rs
2648 .sp
2649 .nf
2650 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2651 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
2652 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
2653 .B " int *\fIworkspace\fP, int \fIwscount\fP);"
2654 .fi
2655 .P
2656 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2657 a compiled pattern, using a matching algorithm that scans the subject string
2658 just once, and does not backtrack. This has different characteristics to the
2659 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2660 patterns are not supported. Nevertheless, there are times when this kind of
2661 matching can be useful. For a discussion of the two matching algorithms, and a
2662 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2663 .\" HREF
2664 \fBpcrematching\fP
2665 .\"
2666 documentation.
2667 .P
2668 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2669 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2670 different way, and this is described below. The other common arguments are used
2671 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2672 here.
2673 .P
2674 The two additional arguments provide workspace for the function. The workspace
2675 vector should contain at least 20 elements. It is used for keeping track of
2676 multiple paths through the pattern tree. More workspace will be needed for
2677 patterns and subjects where there are a lot of potential matches.
2678 .P
2679 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2680 .sp
2681 int rc;
2682 int ovector[10];
2683 int wspace[20];
2684 rc = pcre_dfa_exec(
2685 re, /* result of pcre_compile() */
2686 NULL, /* we didn't study the pattern */
2687 "some string", /* the subject string */
2688 11, /* the length of the subject string */
2689 0, /* start at offset 0 in the subject */
2690 0, /* default options */
2691 ovector, /* vector of integers for substring information */
2692 10, /* number of elements (NOT size in bytes) */
2693 wspace, /* working space vector */
2694 20); /* number of elements (NOT size in bytes) */
2695 .
2696 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2697 .rs
2698 .sp
2699 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2700 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2701 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2702 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2703 PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2704 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2705 so their description is not repeated here.
2706 .sp
2707 PCRE_PARTIAL_HARD
2708 PCRE_PARTIAL_SOFT
2709 .sp
2710 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2711 details are slightly different. When PCRE_PARTIAL_HARD is set for
2712 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2713 is reached and there is still at least one matching possibility that requires
2714 additional characters. This happens even if some complete matches have also
2715 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2716 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2717 there have been no complete matches, but there is still at least one matching
2718 possibility. The portion of the string that was inspected when the longest
2719 partial match was found is set as the first matching string in both cases.
2720 There is a more detailed discussion of partial and multi-segment matching, with
2721 examples, in the
2722 .\" HREF
2723 \fBpcrepartial\fP
2724 .\"
2725 documentation.
2726 .sp
2727 PCRE_DFA_SHORTEST
2728 .sp
2729 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2730 soon as it has found one match. Because of the way the alternative algorithm
2731 works, this is necessarily the shortest possible match at the first possible
2732 matching point in the subject string.
2733 .sp
2734 PCRE_DFA_RESTART
2735 .sp
2736 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2737 again, with additional subject characters, and have it continue with the same
2738 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2739 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2740 before because data about the match so far is left in them after a partial
2741 match. There is more discussion of this facility in the
2742 .\" HREF
2743 \fBpcrepartial\fP
2744 .\"
2745 documentation.
2746 .
2747 .
2748 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2749 .rs
2750 .sp
2751 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2752 substring in the subject. Note, however, that all the matches from one run of
2753 the function start at the same point in the subject. The shorter matches are
2754 all initial substrings of the longer matches. For example, if the pattern
2755 .sp
2756 <.*>
2757 .sp
2758 is matched against the string
2759 .sp
2760 This is <something> <something else> <something further> no more
2761 .sp
2762 the three matched strings are
2763 .sp
2764 <something>
2765 <something> <something else>
2766 <something> <something else> <something further>
2767 .sp
2768 On success, the yield of the function is a number greater than zero, which is
2769 the number of matched substrings. The substrings themselves are returned in
2770 \fIovector\fP. Each string uses two elements; the first is the offset to the
2771 start, and the second is the offset to the end. In fact, all the strings have
2772 the same start offset. (Space could have been saved by giving this only once,
2773 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2774 returns data, even though the meaning of the strings is different.)
2775 .P
2776 The strings are returned in reverse order of length; that is, the longest
2777 matching string is given first. If there were too many matches to fit into
2778 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2779 the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2780 the entire \fIovector\fP for returning matched strings.
2781 .
2782 .
2783 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2784 .rs
2785 .sp
2786 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2787 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2788 described
2789 .\" HTML <a href="#errorlist">
2790 .\" </a>
2791 above.
2792 .\"
2793 There are in addition the following errors that are specific to
2794 \fBpcre_dfa_exec()\fP:
2795 .sp
2796 PCRE_ERROR_DFA_UITEM (-16)
2797 .sp
2798 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2799 that it does not support, for instance, the use of \eC or a back reference.
2800 .sp
2801 PCRE_ERROR_DFA_UCOND (-17)
2802 .sp
2803 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2804 uses a back reference for the condition, or a test for recursion in a specific
2805 group. These are not supported.
2806 .sp
2807 PCRE_ERROR_DFA_UMLIMIT (-18)
2808 .sp
2809 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2810 block that contains a setting of the \fImatch_limit\fP or
2811 \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2812 meaningless for DFA matching).
2813 .sp
2814 PCRE_ERROR_DFA_WSSIZE (-19)
2815 .sp
2816 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2817 \fIworkspace\fP vector.
2818 .sp
2819 PCRE_ERROR_DFA_RECURSE (-20)
2820 .sp
2821 When a recursive subpattern is processed, the matching function calls itself
2822 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2823 error is given if the output vector is not large enough. This should be
2824 extremely rare, as a vector of size 1000 is used.
2825 .sp
2826 PCRE_ERROR_DFA_BADRESTART (-30)
2827 .sp
2828 When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2829 some plausibility checks are made on the contents of the workspace, which
2830 should contain data about the previous partial match. If any of these checks
2831 fail, this error is given.
2832 .
2833 .
2834 .SH "SEE ALSO"
2835 .rs
2836 .sp
2837 \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2838 \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2839 \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2840 \fBpcrestack\fP(3).
2841 .
2842 .
2843 .SH AUTHOR
2844 .rs
2845 .sp
2846 .nf
2847 Philip Hazel
2848 University Computing Service
2849 Cambridge CB2 3QH, England.
2850 .fi
2851 .
2852 .
2853 .SH REVISION
2854 .rs
2855 .sp
2856 .nf
2857 Last updated: 12 June 2013
2858 Copyright (c) 1997-2013 University of Cambridge.
2859 .fi

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5