ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log

Revision 960 - (show annotations)
Thu Apr 19 17:30:38 2012 UTC (9 years, 5 months ago) by ph10
File size: 115777 byte(s)
Fix long-standing DFA testing restart bug in pcretest, and add some 
plausibility checks when restarting in pcre_dfa_exec().
1 .TH PCREAPI 3 "19 April 2012" "PCRE 8.31"
3 PCRE - Perl-compatible regular expressions
4 .sp
5 .B #include <pcre.h>
6 .
7 .
9 .rs
10 .sp
11 .SM
12 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
13 .ti +5n
14 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
15 .ti +5n
16 .B const unsigned char *\fItableptr\fP);
17 .PP
18 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
19 .ti +5n
20 .B int *\fIerrorcodeptr\fP,
21 .ti +5n
22 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
23 .ti +5n
24 .B const unsigned char *\fItableptr\fP);
25 .PP
26 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
27 .ti +5n
28 .B const char **\fIerrptr\fP);
29 .PP
30 .B void pcre_free_study(pcre_extra *\fIextra\fP);
31 .PP
32 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
33 .ti +5n
34 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
35 .ti +5n
36 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
37 .PP
38 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
39 .ti +5n
40 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
41 .ti +5n
42 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
43 .ti +5n
44 .B int *\fIworkspace\fP, int \fIwscount\fP);
45 .
46 .
48 .rs
49 .sp
50 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
51 .ti +5n
52 .B const char *\fIsubject\fP, int *\fIovector\fP,
53 .ti +5n
54 .B int \fIstringcount\fP, const char *\fIstringname\fP,
55 .ti +5n
56 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
57 .PP
58 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
59 .ti +5n
60 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
61 .ti +5n
62 .B int \fIbuffersize\fP);
63 .PP
64 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
65 .ti +5n
66 .B const char *\fIsubject\fP, int *\fIovector\fP,
67 .ti +5n
68 .B int \fIstringcount\fP, const char *\fIstringname\fP,
69 .ti +5n
70 .B const char **\fIstringptr\fP);
71 .PP
72 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
73 .ti +5n
74 .B const char *\fIname\fP);
75 .PP
76 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
77 .ti +5n
78 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
79 .PP
80 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
81 .ti +5n
82 .B int \fIstringcount\fP, int \fIstringnumber\fP,
83 .ti +5n
84 .B const char **\fIstringptr\fP);
85 .PP
86 .B int pcre_get_substring_list(const char *\fIsubject\fP,
87 .ti +5n
88 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
89 .PP
90 .B void pcre_free_substring(const char *\fIstringptr\fP);
91 .PP
92 .B void pcre_free_substring_list(const char **\fIstringptr\fP);
93 .
94 .
96 .rs
97 .sp
98 .B pcre_jit_stack *pcre_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
99 .PP
100 .B void pcre_jit_stack_free(pcre_jit_stack *\fIstack\fP);
101 .PP
102 .B void pcre_assign_jit_stack(pcre_extra *\fIextra\fP,
103 .ti +5n
104 .B pcre_jit_callback \fIcallback\fP, void *\fIdata\fP);
105 .PP
106 .B const unsigned char *pcre_maketables(void);
107 .PP
108 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
109 .ti +5n
110 .B int \fIwhat\fP, void *\fIwhere\fP);
111 .PP
112 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
113 .PP
114 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
115 .PP
116 .B const char *pcre_version(void);
117 .PP
118 .B int pcre_pattern_to_host_byte_order(pcre *\fIcode\fP,
119 .ti +5n
120 .B pcre_extra *\fIextra\fP, const unsigned char *\fItables\fP);
121 .
122 .
124 .rs
125 .sp
126 .B void *(*pcre_malloc)(size_t);
127 .PP
128 .B void (*pcre_free)(void *);
129 .PP
130 .B void *(*pcre_stack_malloc)(size_t);
131 .PP
132 .B void (*pcre_stack_free)(void *);
133 .PP
134 .B int (*pcre_callout)(pcre_callout_block *);
135 .
136 .
138 .rs
139 .sp
140 From release 8.30, PCRE can be compiled as a library for handling 16-bit
141 character strings as well as, or instead of, the original library that handles
142 8-bit character strings. To avoid too much complication, this document
143 describes the 8-bit versions of the functions, with only occasional references
144 to the 16-bit library.
145 .P
146 The 16-bit functions operate in the same way as their 8-bit counterparts; they
147 just use different data types for their arguments and results, and their names
148 start with \fBpcre16_\fP instead of \fBpcre_\fP. For every option that has UTF8
149 in its name (for example, PCRE_UTF8), there is a corresponding 16-bit name with
150 UTF8 replaced by UTF16. This facility is in fact just cosmetic; the 16-bit
151 option names define the same bit values.
152 .P
153 References to bytes and UTF-8 in this document should be read as references to
154 16-bit data quantities and UTF-16 when using the 16-bit library, unless
155 specified otherwise. More details of the specific differences for the 16-bit
156 library are given in the
157 .\" HREF
158 \fBpcre16\fP
159 .\"
160 page.
161 .
162 .
164 .rs
165 .sp
166 PCRE has its own native API, which is described in this document. There are
167 also some wrapper functions (for the 8-bit library only) that correspond to the
168 POSIX regular expression API, but they do not give access to all the
169 functionality. They are described in the
170 .\" HREF
171 \fBpcreposix\fP
172 .\"
173 documentation. Both of these APIs define a set of C function calls. A C++
174 wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
175 documented in the
176 .\" HREF
177 \fBpcrecpp\fP
178 .\"
179 page.
180 .P
181 The native API C function prototypes are defined in the header file
182 \fBpcre.h\fP, and on Unix-like systems the (8-bit) library itself is called
183 \fBlibpcre\fP. It can normally be accessed by adding \fB-lpcre\fP to the
184 command for linking an application that uses PCRE. The header file defines the
185 macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
186 for the library. Applications can use these to include support for different
187 releases of PCRE.
188 .P
189 In a Windows environment, if you want to statically link an application program
190 against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
191 including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
192 \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
193 \fB__declspec(dllimport)\fP, with unwanted results.
194 .P
195 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
196 and \fBpcre_exec()\fP are used for compiling and matching regular expressions
197 in a Perl-compatible manner. A sample program that demonstrates the simplest
198 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
199 source distribution. A listing of this program is given in the
200 .\" HREF
201 \fBpcredemo\fP
202 .\"
203 documentation, and the
204 .\" HREF
205 \fBpcresample\fP
206 .\"
207 documentation describes how to compile and run it.
208 .P
209 Just-in-time compiler support is an optional feature of PCRE that can be built
210 in appropriate hardware environments. It greatly speeds up the matching
211 performance of many patterns. Simple programs can easily request that it be
212 used if available, by setting an option that is ignored when it is not
213 relevant. More complicated programs might need to make use of the functions
214 \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
215 \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
216 These functions are discussed in the
217 .\" HREF
218 \fBpcrejit\fP
219 .\"
220 documentation.
221 .P
222 A second matching function, \fBpcre_dfa_exec()\fP, which is not
223 Perl-compatible, is also provided. This uses a different algorithm for the
224 matching. The alternative algorithm finds all possible matches (at a given
225 point in the subject), and scans the subject just once (unless there are
226 lookbehind assertions). However, this algorithm does not return captured
227 substrings. A description of the two matching algorithms and their advantages
228 and disadvantages is given in the
229 .\" HREF
230 \fBpcrematching\fP
231 .\"
232 documentation.
233 .P
234 In addition to the main compiling and matching functions, there are convenience
235 functions for extracting captured substrings from a subject string that is
236 matched by \fBpcre_exec()\fP. They are:
237 .sp
238 \fBpcre_copy_substring()\fP
239 \fBpcre_copy_named_substring()\fP
240 \fBpcre_get_substring()\fP
241 \fBpcre_get_named_substring()\fP
242 \fBpcre_get_substring_list()\fP
243 \fBpcre_get_stringnumber()\fP
244 \fBpcre_get_stringtable_entries()\fP
245 .sp
246 \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
247 provided, to free the memory used for extracted strings.
248 .P
249 The function \fBpcre_maketables()\fP is used to build a set of character tables
250 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
251 or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
252 specialist use. Most commonly, no special tables are passed, in which case
253 internal tables that are generated when PCRE is built are used.
254 .P
255 The function \fBpcre_fullinfo()\fP is used to find out information about a
256 compiled pattern. The function \fBpcre_version()\fP returns a pointer to a
257 string containing the version of PCRE and its date of release.
258 .P
259 The function \fBpcre_refcount()\fP maintains a reference count in a data block
260 containing a compiled pattern. This is provided for the benefit of
261 object-oriented applications.
262 .P
263 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
264 the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
265 respectively. PCRE calls the memory management functions via these variables,
266 so a calling program can replace them if it wishes to intercept the calls. This
267 should be done before calling any PCRE functions.
268 .P
269 The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
270 indirections to memory management functions. These special functions are used
271 only when PCRE is compiled to use the heap for remembering data, instead of
272 recursive function calls, when running the \fBpcre_exec()\fP function. See the
273 .\" HREF
274 \fBpcrebuild\fP
275 .\"
276 documentation for details of how to do this. It is a non-standard way of
277 building PCRE, for use in environments that have limited stacks. Because of the
278 greater use of memory management, it runs more slowly. Separate functions are
279 provided so that special-purpose external code can be used for this case. When
280 used, these functions are always called in a stack-like manner (last obtained,
281 first freed), and always for memory blocks of the same size. There is a
282 discussion about PCRE's stack usage in the
283 .\" HREF
284 \fBpcrestack\fP
285 .\"
286 documentation.
287 .P
288 The global variable \fBpcre_callout\fP initially contains NULL. It can be set
289 by the caller to a "callout" function, which PCRE will then call at specified
290 points during a matching operation. Details are given in the
291 .\" HREF
292 \fBpcrecallout\fP
293 .\"
294 documentation.
295 .
296 .
297 .\" HTML <a name="newlines"></a>
299 .rs
300 .sp
301 PCRE supports five different conventions for indicating line breaks in
302 strings: a single CR (carriage return) character, a single LF (linefeed)
303 character, the two-character sequence CRLF, any of the three preceding, or any
304 Unicode newline sequence. The Unicode newline sequences are the three just
305 mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
306 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
307 (paragraph separator, U+2029).
308 .P
309 Each of the first three conventions is used by at least one operating system as
310 its standard newline sequence. When PCRE is built, a default can be specified.
311 The default default is LF, which is the Unix standard. When PCRE is run, the
312 default can be overridden, either when a pattern is compiled, or when it is
313 matched.
314 .P
315 At compile time, the newline convention can be specified by the \fIoptions\fP
316 argument of \fBpcre_compile()\fP, or it can be specified by special text at the
317 start of the pattern itself; this overrides any other settings. See the
318 .\" HREF
319 \fBpcrepattern\fP
320 .\"
321 page for details of the special character sequences.
322 .P
323 In the PCRE documentation the word "newline" is used to mean "the character or
324 pair of characters that indicate a line break". The choice of newline
325 convention affects the handling of the dot, circumflex, and dollar
326 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
327 recognized line ending sequence, the match position advancement for a
328 non-anchored pattern. There is more detail about this in the
329 .\" HTML <a href="#execoptions">
330 .\" </a>
331 section on \fBpcre_exec()\fP options
332 .\"
333 below.
334 .P
335 The choice of newline convention does not affect the interpretation of
336 the \en or \er escape sequences, nor does it affect what \eR matches, which is
337 controlled in a similar way, but by separate options.
338 .
339 .
341 .rs
342 .sp
343 The PCRE functions can be used in multi-threading applications, with the
344 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
345 \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
346 callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
347 .P
348 The compiled form of a regular expression is not altered during matching, so
349 the same compiled pattern can safely be used by several threads at once.
350 .P
351 If the just-in-time optimization feature is being used, it needs separate
352 memory stack areas for each thread. See the
353 .\" HREF
354 \fBpcrejit\fP
355 .\"
356 documentation for more details.
357 .
358 .
360 .rs
361 .sp
362 The compiled form of a regular expression can be saved and re-used at a later
363 time, possibly by a different program, and even on a host other than the one on
364 which it was compiled. Details are given in the
365 .\" HREF
366 \fBpcreprecompile\fP
367 .\"
368 documentation, which includes a description of the
369 \fBpcre_pattern_to_host_byte_order()\fP function. However, compiling a regular
370 expression with one version of PCRE for use with a different version is not
371 guaranteed to work and may cause crashes.
372 .
373 .
375 .rs
376 .sp
377 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
378 .PP
379 The function \fBpcre_config()\fP makes it possible for a PCRE client to
380 discover which optional features have been compiled into the PCRE library. The
381 .\" HREF
382 \fBpcrebuild\fP
383 .\"
384 documentation has more details about these optional features.
385 .P
386 The first argument for \fBpcre_config()\fP is an integer, specifying which
387 information is required; the second argument is a pointer to a variable into
388 which the information is placed. The returned value is zero on success, or the
389 negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
390 not recognized. The following information is available:
391 .sp
393 .sp
394 The output is an integer that is set to one if UTF-8 support is available;
395 otherwise it is set to zero. If this option is given to the 16-bit version of
396 this function, \fBpcre16_config()\fP, the result is PCRE_ERROR_BADOPTION.
397 .sp
399 .sp
400 The output is an integer that is set to one if UTF-16 support is available;
401 otherwise it is set to zero. This value should normally be given to the 16-bit
402 version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
403 version of this function, the result is PCRE_ERROR_BADOPTION.
404 .sp
406 .sp
407 The output is an integer that is set to one if support for Unicode character
408 properties is available; otherwise it is set to zero.
409 .sp
411 .sp
412 The output is an integer that is set to one if support for just-in-time
413 compiling is available; otherwise it is set to zero.
414 .sp
416 .sp
417 The output is a pointer to a zero-terminated "const char *" string. If JIT
418 support is available, the string contains the name of the architecture for
419 which the JIT compiler is configured, for example "x86 32bit (little endian +
420 unaligned)". If JIT support is not available, the result is NULL.
421 .sp
423 .sp
424 The output is an integer whose value specifies the default character sequence
425 that is recognized as meaning "newline". The four values that are supported
426 are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF, and -1 for ANY.
427 Though they are derived from ASCII, the same values are returned in EBCDIC
428 environments. The default should normally correspond to the standard sequence
429 for your operating system.
430 .sp
432 .sp
433 The output is an integer whose value indicates what character sequences the \eR
434 escape sequence matches by default. A value of 0 means that \eR matches any
435 Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
436 or CRLF. The default can be overridden when a pattern is compiled or matched.
437 .sp
439 .sp
440 The output is an integer that contains the number of bytes used for internal
441 linkage in compiled regular expressions. For the 8-bit library, the value can
442 be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
443 a number of bytes. The default value of 2 is sufficient for all but the most
444 massive patterns, since it allows the compiled pattern to be up to 64K in size.
445 Larger values allow larger regular expressions to be compiled, at the expense
446 of slower matching.
447 .sp
449 .sp
450 The output is an integer that contains the threshold above which the POSIX
451 interface uses \fBmalloc()\fP for output vectors. Further details are given in
452 the
453 .\" HREF
454 \fBpcreposix\fP
455 .\"
456 documentation.
457 .sp
459 .sp
460 The output is a long integer that gives the default limit for the number of
461 internal matching function calls in a \fBpcre_exec()\fP execution. Further
462 details are given with \fBpcre_exec()\fP below.
463 .sp
465 .sp
466 The output is a long integer that gives the default limit for the depth of
467 recursion when calling the internal matching function in a \fBpcre_exec()\fP
468 execution. Further details are given with \fBpcre_exec()\fP below.
469 .sp
471 .sp
472 The output is an integer that is set to one if internal recursion when running
473 \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
474 to remember their state. This is the usual way that PCRE is compiled. The
475 output is zero if PCRE was compiled to use blocks of data on the heap instead
476 of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
477 \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
478 avoiding the use of the stack.
479 .
480 .
482 .rs
483 .sp
484 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
485 .ti +5n
486 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
487 .ti +5n
488 .B const unsigned char *\fItableptr\fP);
489 .sp
490 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
491 .ti +5n
492 .B int *\fIerrorcodeptr\fP,
493 .ti +5n
494 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
495 .ti +5n
496 .B const unsigned char *\fItableptr\fP);
497 .P
498 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
499 called to compile a pattern into an internal form. The only difference between
500 the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
501 \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
502 too much repetition, we refer just to \fBpcre_compile()\fP below, but the
503 information applies equally to \fBpcre_compile2()\fP.
504 .P
505 The pattern is a C string terminated by a binary zero, and is passed in the
506 \fIpattern\fP argument. A pointer to a single block of memory that is obtained
507 via \fBpcre_malloc\fP is returned. This contains the compiled code and related
508 data. The \fBpcre\fP type is defined for the returned block; this is a typedef
509 for a structure whose contents are not externally defined. It is up to the
510 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
511 .P
512 Although the compiled code of a PCRE regex is relocatable, that is, it does not
513 depend on memory location, the complete \fBpcre\fP data block is not
514 fully relocatable, because it may contain a copy of the \fItableptr\fP
515 argument, which is an address (see below).
516 .P
517 The \fIoptions\fP argument contains various bit settings that affect the
518 compilation. It should be zero if no options are required. The available
519 options are described below. Some of them (in particular, those that are
520 compatible with Perl, but some others as well) can also be set and unset from
521 within the pattern (see the detailed description in the
522 .\" HREF
523 \fBpcrepattern\fP
524 .\"
525 documentation). For those options that can be different in different parts of
526 the pattern, the contents of the \fIoptions\fP argument specifies their
527 settings at the start of compilation and execution. The PCRE_ANCHORED,
529 PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
530 compile time.
531 .P
532 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
533 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
534 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
535 error message. This is a static string that is part of the library. You must
536 not try to free it. Normally, the offset from the start of the pattern to the
537 byte that was being processed when the error was discovered is placed in the
538 variable pointed to by \fIerroffset\fP, which must not be NULL (if it is, an
539 immediate error is given). However, for an invalid UTF-8 string, the offset is
540 that of the first byte of the failing character.
541 .P
542 Some errors are not detected until the whole pattern has been scanned; in these
543 cases, the offset passed back is the length of the pattern. Note that the
544 offset is in bytes, not characters, even in UTF-8 mode. It may sometimes point
545 into the middle of a UTF-8 character.
546 .P
547 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
548 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
549 returned via this argument in the event of an error. This is in addition to the
550 textual error message. Error codes and messages are listed below.
551 .P
552 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
553 character tables that are built when PCRE is compiled, using the default C
554 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
555 call to \fBpcre_maketables()\fP. This value is stored with the compiled
556 pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
557 passed to it. For more discussion, see the section on locale support below.
558 .P
559 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
560 .sp
561 pcre *re;
562 const char *error;
563 int erroffset;
564 re = pcre_compile(
565 "^A.*Z", /* the pattern */
566 0, /* default options */
567 &error, /* for error message */
568 &erroffset, /* for error offset */
569 NULL); /* use default character tables */
570 .sp
571 The following names for option bits are defined in the \fBpcre.h\fP header
572 file:
573 .sp
575 .sp
576 If this bit is set, the pattern is forced to be "anchored", that is, it is
577 constrained to match only at the first matching point in the string that is
578 being searched (the "subject string"). This effect can also be achieved by
579 appropriate constructs in the pattern itself, which is the only way to do it in
580 Perl.
581 .sp
583 .sp
584 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
585 all with number 255, before each pattern item. For discussion of the callout
586 facility, see the
587 .\" HREF
588 \fBpcrecallout\fP
589 .\"
590 documentation.
591 .sp
594 .sp
595 These options (which are mutually exclusive) control what the \eR escape
596 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
597 match any Unicode newline sequence. The default is specified when PCRE is
598 built. It can be overridden from within the pattern, or by setting an option
599 when a compiled pattern is matched.
600 .sp
602 .sp
603 If this bit is set, letters in the pattern match both upper and lower case
604 letters. It is equivalent to Perl's /i option, and it can be changed within a
605 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
606 concept of case for characters whose values are less than 128, so caseless
607 matching is always possible. For characters with higher values, the concept of
608 case is supported if PCRE is compiled with Unicode property support, but not
609 otherwise. If you want to use caseless matching for characters 128 and above,
610 you must ensure that PCRE is compiled with Unicode property support as well as
611 with UTF-8 support.
612 .sp
614 .sp
615 If this bit is set, a dollar metacharacter in the pattern matches only at the
616 end of the subject string. Without this option, a dollar also matches
617 immediately before a newline at the end of the string (but not before any other
618 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
619 There is no equivalent to this option in Perl, and no way to set it within a
620 pattern.
621 .sp
623 .sp
624 If this bit is set, a dot metacharacter in the pattern matches a character of
625 any value, including one that indicates a newline. However, it only ever
626 matches one character, even if newlines are coded as CRLF. Without this option,
627 a dot does not match when the current position is at a newline. This option is
628 equivalent to Perl's /s option, and it can be changed within a pattern by a
629 (?s) option setting. A negative class such as [^a] always matches newline
630 characters, independent of the setting of this option.
631 .sp
633 .sp
634 If this bit is set, names used to identify capturing subpatterns need not be
635 unique. This can be helpful for certain types of pattern when it is known that
636 only one instance of the named subpattern can ever be matched. There are more
637 details of named subpatterns below; see also the
638 .\" HREF
639 \fBpcrepattern\fP
640 .\"
641 documentation.
642 .sp
644 .sp
645 If this bit is set, whitespace data characters in the pattern are totally
646 ignored except when escaped or inside a character class. Whitespace does not
647 include the VT character (code 11). In addition, characters between an
648 unescaped # outside a character class and the next newline, inclusive, are also
649 ignored. This is equivalent to Perl's /x option, and it can be changed within a
650 pattern by a (?x) option setting.
651 .P
652 Which characters are interpreted as newlines is controlled by the options
653 passed to \fBpcre_compile()\fP or by a special sequence at the start of the
654 pattern, as described in the section entitled
655 .\" HTML <a href="pcrepattern.html#newlines">
656 .\" </a>
657 "Newline conventions"
658 .\"
659 in the \fBpcrepattern\fP documentation. Note that the end of this type of
660 comment is a literal newline sequence in the pattern; escape sequences that
661 happen to represent a newline do not count.
662 .P
663 This option makes it possible to include comments inside complicated patterns.
664 Note, however, that this applies only to data characters. Whitespace characters
665 may never appear within special character sequences in a pattern, for example
666 within the sequence (?( that introduces a conditional subpattern.
667 .sp
669 .sp
670 This option was invented in order to turn on additional functionality of PCRE
671 that is incompatible with Perl, but it is currently of very little use. When
672 set, any backslash in a pattern that is followed by a letter that has no
673 special meaning causes an error, thus reserving these combinations for future
674 expansion. By default, as in Perl, a backslash followed by a letter with no
675 special meaning is treated as a literal. (Perl can, however, be persuaded to
676 give an error for this, by running it with the -w option.) There are at present
677 no other features controlled by this option. It can also be set by a (?X)
678 option setting within a pattern.
679 .sp
681 .sp
682 If this option is set, an unanchored pattern is required to match before or at
683 the first newline in the subject string, though the matched text may continue
684 over the newline.
685 .sp
687 .sp
688 If this option is set, PCRE's behaviour is changed in some ways so that it is
689 compatible with JavaScript rather than Perl. The changes are as follows:
690 .P
691 (1) A lone closing square bracket in a pattern causes a compile-time error,
692 because this is illegal in JavaScript (by default it is treated as a data
693 character). Thus, the pattern AB]CD becomes illegal when this option is set.
694 .P
695 (2) At run time, a back reference to an unset subpattern group matches an empty
696 string (by default this causes the current matching alternative to fail). A
697 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
698 an "a" in the subject), whereas it fails by default, for Perl compatibility.
699 .P
700 (3) \eU matches an upper case "U" character; by default \eU causes a compile
701 time error (Perl uses \eU to upper case subsequent characters).
702 .P
703 (4) \eu matches a lower case "u" character unless it is followed by four
704 hexadecimal digits, in which case the hexadecimal number defines the code point
705 to match. By default, \eu causes a compile time error (Perl uses it to upper
706 case the following character).
707 .P
708 (5) \ex matches a lower case "x" character unless it is followed by two
709 hexadecimal digits, in which case the hexadecimal number defines the code point
710 to match. By default, as in Perl, a hexadecimal number is always expected after
711 \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
712 binary zero character followed by z).
713 .sp
715 .sp
716 By default, PCRE treats the subject string as consisting of a single line of
717 characters (even if it actually contains newlines). The "start of line"
718 metacharacter (^) matches only at the start of the string, while the "end of
719 line" metacharacter ($) matches only at the end of the string, or before a
720 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
721 Perl.
722 .P
723 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
724 match immediately following or immediately before internal newlines in the
725 subject string, respectively, as well as at the very start and end. This is
726 equivalent to Perl's /m option, and it can be changed within a pattern by a
727 (?m) option setting. If there are no newlines in a subject string, or no
728 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
729 .sp
735 .sp
736 These options override the default newline definition that was chosen when PCRE
737 was built. Setting the first or the second specifies that a newline is
738 indicated by a single character (CR or LF, respectively). Setting
739 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
740 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
741 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
742 that any Unicode newline sequence should be recognized. The Unicode newline
743 sequences are the three just mentioned, plus the single characters VT (vertical
744 tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
745 separator, U+2028), and PS (paragraph separator, U+2029). For the 8-bit
746 library, the last two are recognized only in UTF-8 mode.
747 .P
748 The newline setting in the options word uses three bits that are treated
749 as a number, giving eight possibilities. Currently only six are used (default
750 plus the five values above). This means that if you set more than one newline
751 option, the combination may or may not be sensible. For example,
753 other combinations may yield unused numbers and cause an error.
754 .P
755 The only time that a line break in a pattern is specially recognized when
756 compiling is when PCRE_EXTENDED is set. CR and LF are whitespace characters,
757 and so are ignored in this mode. Also, an unescaped # outside a character class
758 indicates a comment that lasts until after the next line break sequence. In
759 other circumstances, line break sequences in patterns are treated as literal
760 data.
761 .P
762 The newline option that is set at compile time becomes the default that is used
763 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
764 .sp
766 .sp
767 If this option is set, it disables the use of numbered capturing parentheses in
768 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
769 were followed by ?: but named parentheses can still be used for capturing (and
770 they acquire numbers in the usual way). There is no equivalent of this option
771 in Perl.
772 .sp
774 .sp
775 This is an option that acts at matching time; that is, it is really an option
776 for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
777 it is remembered with the compiled pattern and assumed at matching time. For
778 details see the discussion of PCRE_NO_START_OPTIMIZE
779 .\" HTML <a href="#execoptions">
780 .\" </a>
781 below.
782 .\"
783 .sp
785 .sp
786 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
787 \ew, and some of the POSIX character classes. By default, only ASCII characters
788 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
789 classify characters. More details are given in the section on
790 .\" HTML <a href="pcre.html#genericchartypes">
791 .\" </a>
792 generic character types
793 .\"
794 in the
795 .\" HREF
796 \fBpcrepattern\fP
797 .\"
798 page. If you set PCRE_UCP, matching one of the items it affects takes much
799 longer. The option is available only if PCRE has been compiled with Unicode
800 property support.
801 .sp
803 .sp
804 This option inverts the "greediness" of the quantifiers so that they are not
805 greedy by default, but become greedy if followed by "?". It is not compatible
806 with Perl. It can also be set by a (?U) option setting within the pattern.
807 .sp
809 .sp
810 This option causes PCRE to regard both the pattern and the subject as strings
811 of UTF-8 characters instead of single-byte strings. However, it is available
812 only when PCRE is built to include UTF support. If not, the use of this option
813 provokes an error. Details of how this option changes the behaviour of PCRE are
814 given in the
815 .\" HREF
816 \fBpcreunicode\fP
817 .\"
818 page.
819 .sp
821 .sp
822 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8
823 string is automatically checked. There is a discussion about the
824 .\" HTML <a href="pcreunicode.html#utf8strings">
825 .\" </a>
826 validity of UTF-8 strings
827 .\"
828 in the
829 .\" HREF
830 \fBpcreunicode\fP
831 .\"
832 page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
833 error. If you already know that your pattern is valid, and you want to skip
834 this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
835 When it is set, the effect of passing an invalid UTF-8 string as a pattern is
836 undefined. It may cause your program to crash. Note that this option can also
837 be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress the
838 validity checking of subject strings.
839 .
840 .
842 .rs
843 .sp
844 The following table lists the error codes than may be returned by
845 \fBpcre_compile2()\fP, along with the error messages that may be returned by
846 both compiling functions. Note that error messages are always 8-bit ASCII
847 strings, even in 16-bit mode. As PCRE has developed, some error codes have
848 fallen out of use. To avoid confusion, they have not been re-used.
849 .sp
850 0 no error
851 1 \e at end of pattern
852 2 \ec at end of pattern
853 3 unrecognized character follows \e
854 4 numbers out of order in {} quantifier
855 5 number too big in {} quantifier
856 6 missing terminating ] for character class
857 7 invalid escape sequence in character class
858 8 range out of order in character class
859 9 nothing to repeat
860 10 [this code is not in use]
861 11 internal error: unexpected repeat
862 12 unrecognized character after (? or (?-
863 13 POSIX named classes are supported only within a class
864 14 missing )
865 15 reference to non-existent subpattern
866 16 erroffset passed as NULL
867 17 unknown option bit(s) set
868 18 missing ) after comment
869 19 [this code is not in use]
870 20 regular expression is too large
871 21 failed to get memory
872 22 unmatched parentheses
873 23 internal error: code overflow
874 24 unrecognized character after (?<
875 25 lookbehind assertion is not fixed length
876 26 malformed number or name after (?(
877 27 conditional group contains more than two branches
878 28 assertion expected after (?(
879 29 (?R or (?[+-]digits must be followed by )
880 30 unknown POSIX class name
881 31 POSIX collating elements are not supported
882 32 this version of PCRE is compiled without UTF support
883 33 [this code is not in use]
884 34 character value in \ex{...} sequence is too large
885 35 invalid condition (?(0)
886 36 \eC not allowed in lookbehind assertion
887 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
888 38 number after (?C is > 255
889 39 closing ) for (?C expected
890 40 recursive call could loop indefinitely
891 41 unrecognized character after (?P
892 42 syntax error in subpattern name (missing terminator)
893 43 two named subpatterns have the same name
894 44 invalid UTF-8 string (specifically UTF-8)
895 45 support for \eP, \ep, and \eX has not been compiled
896 46 malformed \eP or \ep sequence
897 47 unknown property name after \eP or \ep
898 48 subpattern name is too long (maximum 32 characters)
899 49 too many named subpatterns (maximum 10000)
900 50 [this code is not in use]
901 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
902 52 internal error: overran compiling workspace
903 53 internal error: previously-checked referenced subpattern
904 not found
905 54 DEFINE group contains more than one branch
906 55 repeating a DEFINE group is not allowed
907 56 inconsistent NEWLINE options
908 57 \eg is not followed by a braced, angle-bracketed, or quoted
909 name/number or by a plain number
910 58 a numbered reference must not be zero
911 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
912 60 (*VERB) not recognized
913 61 number is too big
914 62 subpattern name expected
915 63 digit expected after (?+
916 64 ] is an invalid data character in JavaScript compatibility mode
917 65 different names for subpatterns of the same number are
918 not allowed
919 66 (*MARK) must have an argument
920 67 this version of PCRE is not compiled with Unicode property
921 support
922 68 \ec must be followed by an ASCII character
923 69 \ek is not followed by a braced, angle-bracketed, or quoted name
924 70 internal error: unknown opcode in find_fixedlength()
925 71 \eN is not supported in a class
926 72 too many forward references
927 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
928 74 invalid UTF-16 string (specifically UTF-16)
929 .sp
930 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
931 be used if the limits were changed when PCRE was built.
932 .
933 .
934 .\" HTML <a name="studyingapattern"></a>
936 .rs
937 .sp
938 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
939 .ti +5n
940 .B const char **\fIerrptr\fP);
941 .PP
942 If a compiled pattern is going to be used several times, it is worth spending
943 more time analyzing it in order to speed up the time taken for matching. The
944 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
945 argument. If studying the pattern produces additional information that will
946 help speed up matching, \fBpcre_study()\fP returns a pointer to a
947 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
948 results of the study.
949 .P
950 The returned value from \fBpcre_study()\fP can be passed directly to
951 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
952 also contains other fields that can be set by the caller before the block is
953 passed; these are described
954 .\" HTML <a href="#extradata">
955 .\" </a>
956 below
957 .\"
958 in the section on matching a pattern.
959 .P
960 If studying the pattern does not produce any useful information,
961 \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program
962 wants to pass any of the other fields to \fBpcre_exec()\fP or
963 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block.
964 .P
965 The second argument of \fBpcre_study()\fP contains option bits. There are three
966 options:
967 .sp
971 .sp
972 If any of these are set, and the just-in-time compiler is available, the
973 pattern is further compiled into machine code that executes much faster than
974 the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
975 compiler is not available, these options are ignored. All other bits in the
976 \fIoptions\fP argument must be zero.
977 .P
978 JIT compilation is a heavyweight optimization. It can take some time for
979 patterns to be analyzed, and for one-off matches and simple patterns the
980 benefit of faster execution might be offset by a much slower study time.
981 Not all patterns can be optimized by the JIT compiler. For those that cannot be
982 handled, matching automatically falls back to the \fBpcre_exec()\fP
983 interpreter. For more details, see the
984 .\" HREF
985 \fBpcrejit\fP
986 .\"
987 documentation.
988 .P
989 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
990 studying succeeds (even if no data is returned), the variable it points to is
991 set to NULL. Otherwise it is set to point to a textual error message. This is a
992 static string that is part of the library. You must not try to free it. You
993 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
994 sure that it has run successfully.
995 .P
996 When you are finished with a pattern, you can free the memory used for the
997 study data by calling \fBpcre_free_study()\fP. This function was added to the
998 API for release 8.20. For earlier versions, the memory could be freed with
999 \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1000 where JIT optimization is not used, but it is advisable to change to the new
1001 function when convenient.
1002 .P
1003 This is a typical way in which \fBpcre_study\fP() is used (except that in a
1004 real application there should be tests for errors):
1005 .sp
1006 int rc;
1007 pcre *re;
1008 pcre_extra *sd;
1009 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1010 sd = pcre_study(
1011 re, /* result of pcre_compile() */
1012 0, /* no options */
1013 &error); /* set to NULL or points to a message */
1014 rc = pcre_exec( /* see below for details of pcre_exec() options */
1015 re, sd, "subject", 7, 0, 0, ovector, 30);
1016 ...
1017 pcre_free_study(sd);
1018 pcre_free(re);
1019 .sp
1020 Studying a pattern does two things: first, a lower bound for the length of
1021 subject string that is needed to match the pattern is computed. This does not
1022 mean that there are any strings of that length that match, but it does
1023 guarantee that no shorter strings match. The value is used by
1024 \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP to avoid wasting time by trying to
1025 match strings that are shorter than the lower bound. You can find out the value
1026 in a calling program via the \fBpcre_fullinfo()\fP function.
1027 .P
1028 Studying a pattern is also useful for non-anchored patterns that do not have a
1029 single fixed starting character. A bitmap of possible starting bytes is
1030 created. This speeds up finding a position in the subject at which to start
1031 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.)
1032 .P
1033 These two optimizations apply to both \fBpcre_exec()\fP and
1034 \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1035 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option
1036 when calling \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP, but if this is done,
1037 JIT execution is also disabled. You might want to do this if your pattern
1038 contains callouts or (*MARK) and you want to make use of these facilities in
1039 cases where matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
1040 .\" HTML <a href="#execoptions">
1041 .\" </a>
1042 below.
1043 .\"
1044 .
1045 .
1046 .\" HTML <a name="localesupport"></a>
1048 .rs
1049 .sp
1050 PCRE handles caseless matching, and determines whether characters are letters,
1051 digits, or whatever, by reference to a set of tables, indexed by character
1052 value. When running in UTF-8 mode, this applies only to characters
1053 with codes less than 128. By default, higher-valued codes never match escapes
1054 such as \ew or \ed, but they can be tested with \ep if PCRE is built with
1055 Unicode character property support. Alternatively, the PCRE_UCP option can be
1056 set at compile time; this causes \ew and friends to use Unicode property
1057 support instead of built-in tables. The use of locales with Unicode is
1058 discouraged. If you are handling characters with codes greater than 128, you
1059 should either use UTF-8 and Unicode, or use locales, but not try to mix the
1060 two.
1061 .P
1062 PCRE contains an internal set of tables that are used when the final argument
1063 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1064 Normally, the internal tables recognize only ASCII characters. However, when
1065 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1066 default "C" locale of the local system, which may cause them to be different.
1067 .P
1068 The internal tables can always be overridden by tables supplied by the
1069 application that calls PCRE. These may be created in a different locale from
1070 the default. As more and more applications change to using Unicode, the need
1071 for this locale support is expected to die away.
1072 .P
1073 External tables are built by calling the \fBpcre_maketables()\fP function,
1074 which has no arguments, in the relevant locale. The result can then be passed
1075 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
1076 example, to build and use tables that are appropriate for the French locale
1077 (where accented characters with values greater than 128 are treated as letters),
1078 the following code could be used:
1079 .sp
1080 setlocale(LC_CTYPE, "fr_FR");
1081 tables = pcre_maketables();
1082 re = pcre_compile(..., tables);
1083 .sp
1084 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1085 are using Windows, the name for the French locale is "french".
1086 .P
1087 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1088 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1089 that the memory containing the tables remains available for as long as it is
1090 needed.
1091 .P
1092 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1093 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1094 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
1095 pattern, compilation, studying and matching all happen in the same locale, but
1096 different patterns can be compiled in different locales.
1097 .P
1098 It is possible to pass a table pointer or NULL (indicating the use of the
1099 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
1100 this facility could be used to match a pattern in a different locale from the
1101 one in which it was compiled. Passing table pointers at run time is discussed
1102 below in the section on matching a pattern.
1103 .
1104 .
1105 .\" HTML <a name="infoaboutpattern"></a>
1107 .rs
1108 .sp
1109 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1110 .ti +5n
1111 .B int \fIwhat\fP, void *\fIwhere\fP);
1112 .PP
1113 The \fBpcre_fullinfo()\fP function returns information about a compiled
1114 pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1115 library at version 8.30, after more than 10 years of obsolescence.
1116 .P
1117 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1118 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1119 the pattern was not studied. The third argument specifies which piece of
1120 information is required, and the fourth argument is a pointer to a variable
1121 to receive the data. The yield of the function is zero for success, or one of
1122 the following negative numbers:
1123 .sp
1124 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1125 the argument \fIwhere\fP was NULL
1126 PCRE_ERROR_BADMAGIC the "magic number" was not found
1127 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1128 endianness
1129 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1130 .sp
1131 The "magic number" is placed at the start of each compiled pattern as an simple
1132 check against passing an arbitrary memory pointer. The endianness error can
1133 occur if a compiled pattern is saved and reloaded on a different host. Here is
1134 a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1135 pattern:
1136 .sp
1137 int rc;
1138 size_t length;
1139 rc = pcre_fullinfo(
1140 re, /* result of pcre_compile() */
1141 sd, /* result of pcre_study(), or NULL */
1142 PCRE_INFO_SIZE, /* what is required */
1143 &length); /* where to put the data */
1144 .sp
1145 The possible values for the third argument are defined in \fBpcre.h\fP, and are
1146 as follows:
1147 .sp
1149 .sp
1150 Return the number of the highest back reference in the pattern. The fourth
1151 argument should point to an \fBint\fP variable. Zero is returned if there are
1152 no back references.
1153 .sp
1155 .sp
1156 Return the number of capturing subpatterns in the pattern. The fourth argument
1157 should point to an \fBint\fP variable.
1158 .sp
1160 .sp
1161 Return a pointer to the internal default character tables within PCRE. The
1162 fourth argument should point to an \fBunsigned char *\fP variable. This
1163 information call is provided for internal use by the \fBpcre_study()\fP
1164 function. External callers can cause PCRE to use its internal tables by passing
1165 a NULL table pointer.
1166 .sp
1168 .sp
1169 Return information about the first data unit of any matched string, for a
1170 non-anchored pattern. (The name of this option refers to the 8-bit library,
1171 where data units are bytes.) The fourth argument should point to an \fBint\fP
1172 variable.
1173 .P
1174 If there is a fixed first value, for example, the letter "c" from a pattern
1175 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1176 value is always less than 256; in the 16-bit library the value can be up to
1177 0xffff.
1178 .P
1179 If there is no fixed first value, and if either
1180 .sp
1181 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1182 starts with "^", or
1183 .sp
1184 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1185 (if it were set, the pattern would be anchored),
1186 .sp
1187 -1 is returned, indicating that the pattern matches only at the start of a
1188 subject string or after any newline within the string. Otherwise -2 is
1189 returned. For anchored patterns, -2 is returned.
1190 .sp
1192 .sp
1193 If the pattern was studied, and this resulted in the construction of a 256-bit
1194 table indicating a fixed set of values for the first data unit in any matching
1195 string, a pointer to the table is returned. Otherwise NULL is returned. The
1196 fourth argument should point to an \fBunsigned char *\fP variable.
1197 .sp
1199 .sp
1200 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1201 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1202 explicit match is either a literal CR or LF character, or \er or \en.
1203 .sp
1205 .sp
1206 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1207 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1208 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1209 .sp
1211 .sp
1212 Return 1 if the pattern was studied with one of the JIT options, and
1213 just-in-time compiling was successful. The fourth argument should point to an
1214 \fBint\fP variable. A return value of 0 means that JIT support is not available
1215 in this version of PCRE, or that the pattern was not studied with a JIT option,
1216 or that the JIT compiler could not handle this particular pattern. See the
1217 .\" HREF
1218 \fBpcrejit\fP
1219 .\"
1220 documentation for details of what can and cannot be handled.
1221 .sp
1223 .sp
1224 If the pattern was successfully studied with a JIT option, return the size of
1225 the JIT compiled code, otherwise return zero. The fourth argument should point
1226 to a \fBsize_t\fP variable.
1227 .sp
1229 .sp
1230 Return the value of the rightmost literal data unit that must exist in any
1231 matched string, other than at its start, if such a value has been recorded. The
1232 fourth argument should point to an \fBint\fP variable. If there is no such
1233 value, -1 is returned. For anchored patterns, a last literal value is recorded
1234 only if it follows something of variable length. For example, for the pattern
1235 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1236 is -1.
1237 .sp
1239 .sp
1240 Return the number of characters (NB not bytes) in the longest lookbehind
1241 assertion in the pattern. Note that the simple assertions \eb and \eB require a
1242 one-character lookbehind. This information is useful when doing multi-segment
1243 matching using the partial matching facilities.
1244 .sp
1246 .sp
1247 If the pattern was studied and a minimum length for matching subject strings
1248 was computed, its value is returned. Otherwise the returned value is -1. The
1249 value is a number of characters, which in UTF-8 mode may be different from the
1250 number of bytes. The fourth argument should point to an \fBint\fP variable. A
1251 non-negative value is a lower bound to the length of any matching string. There
1252 may not be any strings of that length that do actually match, but every string
1253 that does match is at least that long.
1254 .sp
1258 .sp
1259 PCRE supports the use of named as well as numbered capturing parentheses. The
1260 names are just an additional way of identifying the parentheses, which still
1261 acquire numbers. Several convenience functions such as
1262 \fBpcre_get_named_substring()\fP are provided for extracting captured
1263 substrings by name. It is also possible to extract the data directly, by first
1264 converting the name to a number in order to access the correct pointers in the
1265 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1266 you need to use the name-to-number map, which is described by these three
1267 values.
1268 .P
1269 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1270 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1271 entry; both of these return an \fBint\fP value. The entry size depends on the
1272 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1273 entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1274 the first two bytes of each entry are the number of the capturing parenthesis,
1275 most significant byte first. In the 16-bit library, the pointer points to
1276 16-bit data units, the first of which contains the parenthesis number. The rest
1277 of the entry is the corresponding name, zero terminated.
1278 .P
1279 The names are in alphabetical order. Duplicate names may appear if (?| is used
1280 to create multiple groups with the same number, as described in the
1281 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1282 .\" </a>
1283 section on duplicate subpattern numbers
1284 .\"
1285 in the
1286 .\" HREF
1287 \fBpcrepattern\fP
1288 .\"
1289 page. Duplicate names for subpatterns with different numbers are permitted only
1290 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1291 table in the order in which they were found in the pattern. In the absence of
1292 (?| this is the order of increasing number; when (?| is used this is not
1293 necessarily the case because later subpatterns may have lower numbers.
1294 .P
1295 As a simple example of the name/number table, consider the following pattern
1296 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1297 space - including newlines - is ignored):
1298 .sp
1299 .\" JOIN
1300 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1301 (?<month>\ed\ed) - (?<day>\ed\ed) )
1302 .sp
1303 There are four named subpatterns, so the table has four entries, and each entry
1304 in the table is eight bytes long. The table is as follows, with non-printing
1305 bytes shows in hexadecimal, and undefined bytes shown as ??:
1306 .sp
1307 00 01 d a t e 00 ??
1308 00 05 d a y 00 ?? ??
1309 00 04 m o n t h 00
1310 00 02 y e a r 00 ??
1311 .sp
1312 When writing code to extract data from named subpatterns using the
1313 name-to-number map, remember that the length of the entries is likely to be
1314 different for each compiled pattern.
1315 .sp
1317 .sp
1318 Return 1 if the pattern can be used for partial matching with
1319 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1320 \fBint\fP variable. From release 8.00, this always returns 1, because the
1321 restrictions that previously applied to partial matching have been lifted. The
1322 .\" HREF
1323 \fBpcrepartial\fP
1324 .\"
1325 documentation gives details of partial matching.
1326 .sp
1328 .sp
1329 Return a copy of the options with which the pattern was compiled. The fourth
1330 argument should point to an \fBunsigned long int\fP variable. These option bits
1331 are those specified in the call to \fBpcre_compile()\fP, modified by any
1332 top-level option settings at the start of the pattern itself. In other words,
1333 they are the options that will be in force when matching starts. For example,
1334 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1336 .P
1337 A pattern is automatically anchored by PCRE if all of its top-level
1338 alternatives begin with one of the following:
1339 .sp
1340 ^ unless PCRE_MULTILINE is set
1341 \eA always
1342 \eG always
1343 .\" JOIN
1344 .* if PCRE_DOTALL is set and there are no back
1345 references to the subpattern in which .* appears
1346 .sp
1347 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1348 \fBpcre_fullinfo()\fP.
1349 .sp
1351 .sp
1352 Return the size of the compiled pattern in bytes (for both libraries). The
1353 fourth argument should point to a \fBsize_t\fP variable. This value does not
1354 include the size of the \fBpcre\fP structure that is returned by
1355 \fBpcre_compile()\fP. The value that is passed as the argument to
1356 \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1357 place the compiled data is the value returned by this option plus the size of
1358 the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1359 does not alter the value returned by this option.
1360 .sp
1362 .sp
1363 Return the size in bytes of the data block pointed to by the \fIstudy_data\fP
1364 field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP is NULL, or there is no
1365 study data, zero is returned. The fourth argument should point to a
1366 \fBsize_t\fP variable. The \fIstudy_data\fP field is set by \fBpcre_study()\fP
1367 to record information that will speed up matching (see the section entitled
1368 .\" HTML <a href="#studyingapattern">
1369 .\" </a>
1370 "Studying a pattern"
1371 .\"
1372 above). The format of the \fIstudy_data\fP block is private, but its length
1373 is made available via this option so that it can be saved and restored (see the
1374 .\" HREF
1375 \fBpcreprecompile\fP
1376 .\"
1377 documentation for details).
1378 .
1379 .
1381 .rs
1382 .sp
1383 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1384 .PP
1385 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1386 data block that contains a compiled pattern. It is provided for the benefit of
1387 applications that operate in an object-oriented manner, where different parts
1388 of the application may be using the same compiled pattern, but you want to free
1389 the block when they are all done.
1390 .P
1391 When a pattern is compiled, the reference count field is initialized to zero.
1392 It is changed only by calling this function, whose action is to add the
1393 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1394 function is the new value. However, the value of the count is constrained to
1395 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1396 it is forced to the appropriate limit value.
1397 .P
1398 Except when it is zero, the reference count is not correctly preserved if a
1399 pattern is compiled on one host and then transferred to a host whose byte-order
1400 is different. (This seems a highly unlikely scenario.)
1401 .
1402 .
1404 .rs
1405 .sp
1406 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1407 .ti +5n
1408 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1409 .ti +5n
1410 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1411 .P
1412 The function \fBpcre_exec()\fP is called to match a subject string against a
1413 compiled pattern, which is passed in the \fIcode\fP argument. If the
1414 pattern was studied, the result of the study should be passed in the
1415 \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1416 and \fIextra\fP arguments as many times as you like, in order to match
1417 different subject strings with the same pattern.
1418 .P
1419 This function is the main matching facility of the library, and it operates in
1420 a Perl-like manner. For specialist use there is also an alternative matching
1421 function, which is described
1422 .\" HTML <a href="#dfamatch">
1423 .\" </a>
1424 below
1425 .\"
1426 in the section about the \fBpcre_dfa_exec()\fP function.
1427 .P
1428 In most applications, the pattern will have been compiled (and optionally
1429 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1430 possible to save compiled patterns and study data, and then use them later
1431 in different processes, possibly even on different hosts. For a discussion
1432 about this, see the
1433 .\" HREF
1434 \fBpcreprecompile\fP
1435 .\"
1436 documentation.
1437 .P
1438 Here is an example of a simple call to \fBpcre_exec()\fP:
1439 .sp
1440 int rc;
1441 int ovector[30];
1442 rc = pcre_exec(
1443 re, /* result of pcre_compile() */
1444 NULL, /* we didn't study the pattern */
1445 "some string", /* the subject string */
1446 11, /* the length of the subject string */
1447 0, /* start at offset 0 in the subject */
1448 0, /* default options */
1449 ovector, /* vector of integers for substring information */
1450 30); /* number of elements (NOT size in bytes) */
1451 .
1452 .
1453 .\" HTML <a name="extradata"></a>
1454 .SS "Extra data for \fBpcre_exec()\fR"
1455 .rs
1456 .sp
1457 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1458 data block. The \fBpcre_study()\fP function returns such a block (when it
1459 doesn't return NULL), but you can also create one for yourself, and pass
1460 additional information in it. The \fBpcre_extra\fP block contains the following
1461 fields (not necessarily in this order):
1462 .sp
1463 unsigned long int \fIflags\fP;
1464 void *\fIstudy_data\fP;
1465 void *\fIexecutable_jit\fP;
1466 unsigned long int \fImatch_limit\fP;
1467 unsigned long int \fImatch_limit_recursion\fP;
1468 void *\fIcallout_data\fP;
1469 const unsigned char *\fItables\fP;
1470 unsigned char **\fImark\fP;
1471 .sp
1472 In the 16-bit version of this structure, the \fImark\fP field has type
1473 "PCRE_UCHAR16 **".
1474 .P
1475 The \fIflags\fP field is used to specify which of the other fields are set. The
1476 flag bits are:
1477 .sp
1485 .sp
1486 Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1487 the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1488 returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1489 should not set these yourself, but you may add to the block by setting other
1490 fields and their corresponding flag bits.
1491 .P
1492 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1493 vast amount of resources when running patterns that are not going to match,
1494 but which have a very large number of possibilities in their search trees. The
1495 classic example is a pattern that uses nested unlimited repeats.
1496 .P
1497 Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1498 calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1499 imposed on the number of times this function is called during a match, which
1500 has the effect of limiting the amount of backtracking that can take place. For
1501 patterns that are not anchored, the count restarts from zero for each position
1502 in the subject string.
1503 .P
1504 When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1505 with a JIT option, the way that the matching is executed is entirely different.
1506 However, there is still the possibility of runaway matching that goes on for a
1507 very long time, and so the \fImatch_limit\fP value is also used in this case
1508 (but in a different way) to limit how long the matching can continue.
1509 .P
1510 The default value for the limit can be set when PCRE is built; the default
1511 default is 10 million, which handles all but the most extreme cases. You can
1512 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1513 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1514 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1516 .P
1517 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1518 instead of limiting the total number of times that \fBmatch()\fP is called, it
1519 limits the depth of recursion. The recursion depth is a smaller number than the
1520 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1521 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1522 .P
1523 Limiting the recursion depth limits the amount of machine stack that can be
1524 used, or, when PCRE has been compiled to use memory on the heap instead of the
1525 stack, the amount of heap memory that can be used. This limit is not relevant,
1526 and is ignored, when matching is done using JIT compiled code.
1527 .P
1528 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1529 built; the default default is the same value as the default for
1530 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1531 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1532 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1533 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1534 .P
1535 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1536 and is described in the
1537 .\" HREF
1538 \fBpcrecallout\fP
1539 .\"
1540 documentation.
1541 .P
1542 The \fItables\fP field is used to pass a character tables pointer to
1543 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1544 pattern. A non-NULL value is stored with the compiled pattern only if custom
1545 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1546 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1547 internal tables to be used. This facility is helpful when re-using patterns
1548 that have been saved after compiling with an external set of tables, because
1549 the external tables might be at a different address when \fBpcre_exec()\fP is
1550 called. See the
1551 .\" HREF
1552 \fBpcreprecompile\fP
1553 .\"
1554 documentation for a discussion of saving compiled patterns for later use.
1555 .P
1556 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1557 be set to point to a suitable variable. If the pattern contains any
1558 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1559 a name to pass back, a pointer to the name string (zero terminated) is placed
1560 in the variable pointed to by the \fImark\fP field. The names are within the
1561 compiled pattern; if you wish to retain such a name you must copy it before
1562 freeing the memory of a compiled pattern. If there is no name to pass back, the
1563 variable pointed to by the \fImark\fP field is set to NULL. For details of the
1564 backtracking control verbs, see the section entitled
1565 .\" HTML <a href="pcrepattern#backtrackcontrol">
1566 .\" </a>
1567 "Backtracking control"
1568 .\"
1569 in the
1570 .\" HREF
1571 \fBpcrepattern\fP
1572 .\"
1573 documentation.
1574 .
1575 .
1576 .\" HTML <a name="execoptions"></a>
1577 .SS "Option bits for \fBpcre_exec()\fP"
1578 .rs
1579 .sp
1580 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1581 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1585 .P
1586 If the pattern was successfully studied with one of the just-in-time (JIT)
1587 compile options, the only supported options for JIT execution are
1590 unsupported option is used, JIT execution is disabled and the normal
1591 interpretive code in \fBpcre_exec()\fP is run.
1592 .sp
1594 .sp
1595 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1596 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1597 to be anchored by virtue of its contents, it cannot be made unachored at
1598 matching time.
1599 .sp
1602 .sp
1603 These options (which are mutually exclusive) control what the \eR escape
1604 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1605 match any Unicode newline sequence. These options override the choice that was
1606 made or defaulted when the pattern was compiled.
1607 .sp
1613 .sp
1614 These options override the newline definition that was chosen or defaulted when
1615 the pattern was compiled. For details, see the description of
1616 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1617 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1618 the way the match position is advanced after a match failure for an unanchored
1619 pattern.
1620 .P
1622 match attempt for an unanchored pattern fails when the current position is at a
1623 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1624 characters, the match position is advanced by two characters instead of one, in
1625 other words, to after the CRLF.
1626 .P
1627 The above rule is a compromise that makes the most common cases work as
1628 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1629 set), it does not match the string "\er\enA" because, after failing at the
1630 start, it skips both the CR and the LF before retrying. However, the pattern
1631 [\er\en]A does match that string, because it contains an explicit CR or LF
1632 reference, and so advances only by one character after the first failure.
1633 .P
1634 An explicit match for CR of LF is either a literal appearance of one of those
1635 characters, or one of the \er or \en escape sequences. Implicit matches such as
1636 [^X] do not count, nor does \es (which includes CR and LF in the characters
1637 that it matches).
1638 .P
1639 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1640 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1641 .sp
1643 .sp
1644 This option specifies that first character of the subject string is not the
1645 beginning of a line, so the circumflex metacharacter should not match before
1646 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1647 never to match. This option affects only the behaviour of the circumflex
1648 metacharacter. It does not affect \eA.
1649 .sp
1651 .sp
1652 This option specifies that the end of the subject string is not the end of a
1653 line, so the dollar metacharacter should not match it nor (except in multiline
1654 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1655 compile time) causes dollar never to match. This option affects only the
1656 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1657 .sp
1659 .sp
1660 An empty string is not considered to be a valid match if this option is set. If
1661 there are alternatives in the pattern, they are tried. If all the alternatives
1662 match the empty string, the entire match fails. For example, if the pattern
1663 .sp
1664 a?b?
1665 .sp
1666 is applied to a string not beginning with "a" or "b", it matches an empty
1667 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1668 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1669 .sp
1671 .sp
1672 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1673 the start of the subject is permitted. If the pattern is anchored, such a match
1674 can occur only if the pattern contains \eK.
1675 .P
1676 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1677 does make a special case of a pattern match of the empty string within its
1678 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1679 emulate Perl's behaviour after matching a null string by first trying the match
1680 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1681 if that fails, by advancing the starting offset (see below) and trying an
1682 ordinary match again. There is some code that demonstrates how to do this in
1683 the
1684 .\" HREF
1685 \fBpcredemo\fP
1686 .\"
1687 sample program. In the most general case, you have to check to see if the
1688 newline convention recognizes CRLF as a newline, and if so, and the current
1689 character is CR followed by LF, advance the starting offset by two characters
1690 instead of one.
1691 .sp
1693 .sp
1694 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1695 a match, in order to speed up the process. For example, if it is known that an
1696 unanchored match must start with a specific character, it searches the subject
1697 for that character, and fails immediately if it cannot find it, without
1698 actually running the main matching function. This means that a special item
1699 such as (*COMMIT) at the start of a pattern is not considered until after a
1700 suitable starting point for the match has been found. When callouts or (*MARK)
1701 items are in use, these "start-up" optimizations can cause them to be skipped
1702 if the pattern is never actually used. The start-up optimizations are in effect
1703 a pre-scan of the subject that takes place before the pattern is run.
1704 .P
1705 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1706 causing performance to suffer, but ensuring that in cases where the result is
1707 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1708 are considered at every possible starting position in the subject string. If
1709 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1710 time. The use of PCRE_NO_START_OPTIMIZE disables JIT execution; when it is set,
1711 matching is always done using interpretively.
1712 .P
1713 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1714 Consider the pattern
1715 .sp
1717 .sp
1718 When this is compiled, PCRE records the fact that a match must start with the
1719 character "A". Suppose the subject string is "DEFABC". The start-up
1720 optimization scans along the subject, finds "A" and runs the first match
1721 attempt from there. The (*COMMIT) item means that the pattern must match the
1722 current starting position, which in this case, it does. However, if the same
1723 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1724 subject string does not happen. The first match attempt is run starting from
1725 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1726 the overall result is "no match". If the pattern is studied, more start-up
1727 optimizations may be used. For example, a minimum length for the subject may be
1728 recorded. Consider the pattern
1729 .sp
1730 (*MARK:A)(X|Y)
1731 .sp
1732 The minimum length for a match is one character. If the subject is "ABC", there
1733 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1734 If the pattern is studied, the final attempt does not take place, because PCRE
1735 knows that the subject is too short, and so the (*MARK) is never encountered.
1736 In this case, studying the pattern does not affect the overall match result,
1737 which is still "no match", but it does affect the auxiliary information that is
1738 returned.
1739 .sp
1741 .sp
1742 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1743 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1744 The entire string is checked before any other processing takes place. The value
1745 of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1746 UTF-8 character. There is a discussion about the
1747 .\" HTML <a href="pcreunicode.html#utf8strings">
1748 .\" </a>
1749 validity of UTF-8 strings
1750 .\"
1751 in the
1752 .\" HREF
1753 \fBpcreunicode\fP
1754 .\"
1755 page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1756 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1757 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1758 cases, information about the precise nature of the error may also be returned
1759 (see the descriptions of these errors in the section entitled \fIError return
1760 values from\fP \fBpcre_exec()\fP
1761 .\" HTML <a href="#errorlist">
1762 .\" </a>
1763 below).
1764 .\"
1765 If \fIstartoffset\fP contains a value that does not point to the start of a
1766 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1767 returned.
1768 .P
1769 If you already know that your subject is valid, and you want to skip these
1770 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1771 calling \fBpcre_exec()\fP. You might want to do this for the second and
1772 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1773 all the matches in a single subject string. However, you should be sure that
1774 the value of \fIstartoffset\fP points to the start of a character (or the end
1775 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1776 invalid string as a subject or an invalid value of \fIstartoffset\fP is
1777 undefined. Your program may crash.
1778 .sp
1781 .sp
1782 These options turn on the partial matching feature. For backwards
1783 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1784 occurs if the end of the subject string is reached successfully, but there are
1785 not enough subject characters to complete the match. If this happens when
1786 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1787 testing any remaining alternatives. Only if no complete match can be found is
1788 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1789 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1790 but only if no complete match can be found.
1791 .P
1792 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1793 partial match is found, \fBpcre_exec()\fP immediately returns
1794 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1795 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1796 important that an alternative complete match.
1797 .P
1798 In both cases, the portion of the string that was inspected when the partial
1799 match was found is set as the first matching string. There is a more detailed
1800 discussion of partial and multi-segment matching, with examples, in the
1801 .\" HREF
1802 \fBpcrepartial\fP
1803 .\"
1804 documentation.
1805 .
1806 .
1807 .SS "The string to be matched by \fBpcre_exec()\fP"
1808 .rs
1809 .sp
1810 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1811 \fIsubject\fP, a length in bytes in \fIlength\fP, and a starting byte offset
1812 in \fIstartoffset\fP. If this is negative or greater than the length of the
1813 subject, \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting
1814 offset is zero, the search for a match starts at the beginning of the subject,
1815 and this is by far the most common case. In UTF-8 mode, the byte offset must
1816 point to the start of a UTF-8 character (or the end of the subject). Unlike the
1817 pattern string, the subject may contain binary zero bytes.
1818 .P
1819 A non-zero starting offset is useful when searching for another match in the
1820 same subject by calling \fBpcre_exec()\fP again after a previous success.
1821 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1822 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1823 lookbehind. For example, consider the pattern
1824 .sp
1825 \eBiss\eB
1826 .sp
1827 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1828 the current position in the subject is not a word boundary.) When applied to
1829 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1830 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1831 subject, namely "issipi", it does not match, because \eB is always false at the
1832 start of the subject, which is deemed to be a word boundary. However, if
1833 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1834 set to 4, it finds the second occurrence of "iss" because it is able to look
1835 behind the starting point to discover that it is preceded by a letter.
1836 .P
1837 Finding all the matches in a subject is tricky when the pattern can match an
1838 empty string. It is possible to emulate Perl's /g behaviour by first trying the
1839 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1840 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1841 and trying an ordinary match again. There is some code that demonstrates how to
1842 do this in the
1843 .\" HREF
1844 \fBpcredemo\fP
1845 .\"
1846 sample program. In the most general case, you have to check to see if the
1847 newline convention recognizes CRLF as a newline, and if so, and the current
1848 character is CR followed by LF, advance the starting offset by two characters
1849 instead of one.
1850 .P
1851 If a non-zero starting offset is passed when the pattern is anchored, one
1852 attempt to match at the given offset is made. This can only succeed if the
1853 pattern does not require the match to be at the start of the subject.
1854 .
1855 .
1856 .SS "How \fBpcre_exec()\fP returns captured substrings"
1857 .rs
1858 .sp
1859 In general, a pattern matches a certain portion of the subject, and in
1860 addition, further substrings from the subject may be picked out by parts of the
1861 pattern. Following the usage in Jeffrey Friedl's book, this is called
1862 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1863 a fragment of a pattern that picks out a substring. PCRE supports several other
1864 kinds of parenthesized subpattern that do not cause substrings to be captured.
1865 .P
1866 Captured substrings are returned to the caller via a vector of integers whose
1867 address is passed in \fIovector\fP. The number of elements in the vector is
1868 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
1869 argument is NOT the size of \fIovector\fP in bytes.
1870 .P
1871 The first two-thirds of the vector is used to pass back captured substrings,
1872 each substring using a pair of integers. The remaining third of the vector is
1873 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1874 and is not available for passing back information. The number passed in
1875 \fIovecsize\fP should always be a multiple of three. If it is not, it is
1876 rounded down.
1877 .P
1878 When a match is successful, information about captured substrings is returned
1879 in pairs of integers, starting at the beginning of \fIovector\fP, and
1880 continuing up to two-thirds of its length at the most. The first element of
1881 each pair is set to the byte offset of the first character in a substring, and
1882 the second is set to the byte offset of the first character after the end of a
1883 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
1884 mode. They are not character counts.
1885 .P
1886 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
1887 portion of the subject string matched by the entire pattern. The next pair is
1888 used for the first capturing subpattern, and so on. The value returned by
1889 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
1890 For example, if two substrings have been captured, the returned value is 3. If
1891 there are no capturing subpatterns, the return value from a successful match is
1892 1, indicating that just the first pair of offsets has been set.
1893 .P
1894 If a capturing subpattern is matched repeatedly, it is the last portion of the
1895 string that it matched that is returned.
1896 .P
1897 If the vector is too small to hold all the captured substring offsets, it is
1898 used as far as possible (up to two-thirds of its length), and the function
1899 returns a value of zero. If neither the actual string matched nor any captured
1900 substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
1901 passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
1902 back references and the \fIovector\fP is not big enough to remember the related
1903 substrings, PCRE has to get additional memory for use during matching. Thus it
1904 is usually advisable to supply an \fIovector\fP of reasonable size.
1905 .P
1906 There are some cases where zero is returned (indicating vector overflow) when
1907 in fact the vector is exactly the right size for the final match. For example,
1908 consider the pattern
1909 .sp
1910 (a)(?:(b)c|bd)
1911 .sp
1912 If a vector of 6 elements (allowing for only 1 captured substring) is given
1913 with subject string "abd", \fBpcre_exec()\fP will try to set the second
1914 captured string, thereby recording a vector overflow, before failing to match
1915 "c" and backing up to try the second alternative. The zero return, however,
1916 does correctly indicate that the maximum number of slots (namely 2) have been
1917 filled. In similar cases where there is temporary overflow, but the final
1918 number of used slots is actually less than the maximum, a non-zero value is
1919 returned.
1920 .P
1921 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
1922 subpatterns there are in a compiled pattern. The smallest size for
1923 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
1924 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
1925 .P
1926 It is possible for capturing subpattern number \fIn+1\fP to match some part of
1927 the subject when subpattern \fIn\fP has not been used at all. For example, if
1928 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1929 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1930 happens, both values in the offset pairs corresponding to unused subpatterns
1931 are set to -1.
1932 .P
1933 Offset values that correspond to unused subpatterns at the end of the
1934 expression are also set to -1. For example, if the string "abc" is matched
1935 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1936 return from the function is 2, because the highest used capturing subpattern
1937 number is 1, and the offsets for for the second and third capturing subpatterns
1938 (assuming the vector is large enough, of course) are set to -1.
1939 .P
1940 \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
1941 correspond to capturing parentheses in the pattern are never changed. That is,
1942 if a pattern contains \fIn\fP capturing parentheses, no more than
1943 \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
1944 elements (in the first two-thirds) retain whatever values they previously had.
1945 .P
1946 Some convenience functions are provided for extracting the captured substrings
1947 as separate strings. These are described below.
1948 .
1949 .
1950 .\" HTML <a name="errorlist"></a>
1951 .SS "Error return values from \fBpcre_exec()\fP"
1952 .rs
1953 .sp
1954 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1955 defined in the header file:
1956 .sp
1958 .sp
1959 The subject string did not match the pattern.
1960 .sp
1962 .sp
1963 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
1964 NULL and \fIovecsize\fP was not zero.
1965 .sp
1967 .sp
1968 An unrecognized bit was set in the \fIoptions\fP argument.
1969 .sp
1971 .sp
1972 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1973 the case when it is passed a junk pointer and to detect when a pattern that was
1974 compiled in an environment of one endianness is run in an environment with the
1975 other endianness. This is the error that PCRE gives when the magic number is
1976 not present.
1977 .sp
1979 .sp
1980 While running the pattern match, an unknown item was encountered in the
1981 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1982 of the compiled pattern.
1983 .sp
1985 .sp
1986 If a pattern contains back references, but the \fIovector\fP that is passed to
1987 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
1988 gets a block of memory at the start of matching to use for this purpose. If the
1989 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
1990 automatically freed at the end of matching.
1991 .P
1992 This error is also given if \fBpcre_stack_malloc()\fP fails in
1993 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
1994 \fB--disable-stack-for-recursion\fP.
1995 .sp
1997 .sp
1998 This error is used by the \fBpcre_copy_substring()\fP,
1999 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2000 below). It is never returned by \fBpcre_exec()\fP.
2001 .sp
2003 .sp
2004 The backtracking limit, as specified by the \fImatch_limit\fP field in a
2005 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2006 above.
2007 .sp
2009 .sp
2010 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2011 use by callout functions that want to yield a distinctive error code. See the
2012 .\" HREF
2013 \fBpcrecallout\fP
2014 .\"
2015 documentation for details.
2016 .sp
2018 .sp
2019 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2020 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2021 (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2022 UTF-8 character is placed in the first element, and a reason code is placed in
2023 the second element. The reason codes are listed in the
2024 .\" HTML <a href="#badutf8reasons">
2025 .\" </a>
2026 following section.
2027 .\"
2028 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2029 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2030 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2031 .sp
2033 .sp
2034 The UTF-8 byte sequence that was passed as a subject was checked and found to
2035 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2036 \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2037 end of the subject.
2038 .sp
2040 .sp
2041 The subject string did not match, but it did match partially. See the
2042 .\" HREF
2043 \fBpcrepartial\fP
2044 .\"
2045 documentation for details of partial matching.
2046 .sp
2048 .sp
2049 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2050 option was used with a compiled pattern containing items that were not
2051 supported for partial matching. From release 8.00 onwards, there are no
2052 restrictions on partial matching.
2053 .sp
2055 .sp
2056 An unexpected internal error has occurred. This error could be caused by a bug
2057 in PCRE or by overwriting of the compiled pattern.
2058 .sp
2060 .sp
2061 This error is given if the value of the \fIovecsize\fP argument is negative.
2062 .sp
2064 .sp
2065 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2066 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2067 description above.
2068 .sp
2070 .sp
2071 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2072 .sp
2074 .sp
2075 The value of \fIstartoffset\fP was negative or greater than the length of the
2076 subject, that is, the value in \fIlength\fP.
2077 .sp
2079 .sp
2080 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2081 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2082 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2083 fact sufficient to detect this case, but this special error code for
2084 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2085 retained for backwards compatibility.
2086 .sp
2088 .sp
2089 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2090 the pattern. Specifically, it means that either the whole pattern or a
2091 subpattern has been called recursively for the second time at the same position
2092 in the subject string. Some simple patterns that might do this are detected and
2093 faulted at compile time, but more complicated cases, in particular mutual
2094 recursions between two different subpatterns, cannot be detected until run
2095 time.
2096 .sp
2098 .sp
2099 This error is returned when a pattern that was successfully studied using a
2100 JIT compile option is being matched, but the memory available for the
2101 just-in-time processing stack is not large enough. See the
2102 .\" HREF
2103 \fBpcrejit\fP
2104 .\"
2105 documentation for more details.
2106 .sp
2108 .sp
2109 This error is given if a pattern that was compiled by the 8-bit library is
2110 passed to a 16-bit library function, or vice versa.
2111 .sp
2113 .sp
2114 This error is given if a pattern that was compiled and saved is reloaded on a
2115 host with different endianness. The utility function
2116 \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2117 so that it runs on the new host.
2118 .P
2119 Error numbers -16 to -20, -22, and -30 are not used by \fBpcre_exec()\fP.
2120 .
2121 .
2122 .\" HTML <a name="badutf8reasons"></a>
2123 .SS "Reason codes for invalid UTF-8 strings"
2124 .rs
2125 .sp
2126 This section applies only to the 8-bit library. The corresponding information
2127 for the 16-bit library is given in the
2128 .\" HREF
2129 \fBpcre16\fP
2130 .\"
2131 page.
2132 .P
2133 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2134 PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2135 least 2, the offset of the start of the invalid UTF-8 character is placed in
2136 the first output vector element (\fIovector[0]\fP) and a reason code is placed
2137 in the second element (\fIovector[1]\fP). The reason codes are given names in
2138 the \fBpcre.h\fP header file:
2139 .sp
2145 .sp
2146 The string ends with a truncated UTF-8 character; the code specifies how many
2147 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2148 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2149 allows for up to 6 bytes, and this is checked first; hence the possibility of
2150 4 or 5 missing bytes.
2151 .sp
2156 PCRE_UTF8_ERR10
2157 .sp
2158 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2159 character do not have the binary value 0b10 (that is, either the most
2160 significant bit is 0, or the next bit is 1).
2161 .sp
2162 PCRE_UTF8_ERR11
2163 PCRE_UTF8_ERR12
2164 .sp
2165 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2166 these code points are excluded by RFC 3629.
2167 .sp
2168 PCRE_UTF8_ERR13
2169 .sp
2170 A 4-byte character has a value greater than 0x10fff; these code points are
2171 excluded by RFC 3629.
2172 .sp
2173 PCRE_UTF8_ERR14
2174 .sp
2175 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2176 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2177 from UTF-8.
2178 .sp
2179 PCRE_UTF8_ERR15
2180 PCRE_UTF8_ERR16
2181 PCRE_UTF8_ERR17
2182 PCRE_UTF8_ERR18
2183 PCRE_UTF8_ERR19
2184 .sp
2185 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2186 value that can be represented by fewer bytes, which is invalid. For example,
2187 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2188 one byte.
2189 .sp
2190 PCRE_UTF8_ERR20
2191 .sp
2192 The two most significant bits of the first byte of a character have the binary
2193 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2194 byte can only validly occur as the second or subsequent byte of a multi-byte
2195 character.
2196 .sp
2197 PCRE_UTF8_ERR21
2198 .sp
2199 The first byte of a character has the value 0xfe or 0xff. These values can
2200 never occur in a valid UTF-8 string.
2201 .
2202 .
2204 .rs
2205 .sp
2206 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2207 .ti +5n
2208 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
2209 .ti +5n
2210 .B int \fIbuffersize\fP);
2211 .PP
2212 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2213 .ti +5n
2214 .B int \fIstringcount\fP, int \fIstringnumber\fP,
2215 .ti +5n
2216 .B const char **\fIstringptr\fP);
2217 .PP
2218 .B int pcre_get_substring_list(const char *\fIsubject\fP,
2219 .ti +5n
2220 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
2221 .PP
2222 Captured substrings can be accessed directly by using the offsets returned by
2223 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2224 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2225 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2226 as new, separate, zero-terminated strings. These functions identify substrings
2227 by number. The next section describes functions for extracting named
2228 substrings.
2229 .P
2230 A substring that contains a binary zero is correctly extracted and has a
2231 further zero added on the end, but the result is not, of course, a C string.
2232 However, you can process such a string by referring to the length that is
2233 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2234 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2235 for handling strings containing binary zeros, because the end of the final
2236 string is not independently indicated.
2237 .P
2238 The first three arguments are the same for all three of these functions:
2239 \fIsubject\fP is the subject string that has just been successfully matched,
2240 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2241 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2242 captured by the match, including the substring that matched the entire regular
2243 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2244 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2245 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2246 number of elements in the vector divided by three.
2247 .P
2248 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2249 extract a single substring, whose number is given as \fIstringnumber\fP. A
2250 value of zero extracts the substring that matched the entire pattern, whereas
2251 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2252 the string is placed in \fIbuffer\fP, whose length is given by
2253 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2254 obtained via \fBpcre_malloc\fP, and its address is returned via
2255 \fIstringptr\fP. The yield of the function is the length of the string, not
2256 including the terminating zero, or one of these error codes:
2257 .sp
2259 .sp
2260 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2261 memory failed for \fBpcre_get_substring()\fP.
2262 .sp
2264 .sp
2265 There is no substring whose number is \fIstringnumber\fP.
2266 .P
2267 The \fBpcre_get_substring_list()\fP function extracts all available substrings
2268 and builds a list of pointers to them. All this is done in a single block of
2269 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2270 is returned via \fIlistptr\fP, which is also the start of the list of string
2271 pointers. The end of the list is marked by a NULL pointer. The yield of the
2272 function is zero if all went well, or the error code
2273 .sp
2275 .sp
2276 if the attempt to get the memory block failed.
2277 .P
2278 When any of these functions encounter a substring that is unset, which can
2279 happen when capturing subpattern number \fIn+1\fP matches some part of the
2280 subject, but subpattern \fIn\fP has not been used at all, they return an empty
2281 string. This can be distinguished from a genuine zero-length substring by
2282 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2283 substrings.
2284 .P
2285 The two convenience functions \fBpcre_free_substring()\fP and
2286 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2287 a previous call of \fBpcre_get_substring()\fP or
2288 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2289 the function pointed to by \fBpcre_free\fP, which of course could be called
2290 directly from a C program. However, PCRE is used in some situations where it is
2291 linked via a special interface to another programming language that cannot use
2292 \fBpcre_free\fP directly; it is for these cases that the functions are
2293 provided.
2294 .
2295 .
2297 .rs
2298 .sp
2299 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2300 .ti +5n
2301 .B const char *\fIname\fP);
2302 .PP
2303 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2304 .ti +5n
2305 .B const char *\fIsubject\fP, int *\fIovector\fP,
2306 .ti +5n
2307 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2308 .ti +5n
2309 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
2310 .PP
2311 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2312 .ti +5n
2313 .B const char *\fIsubject\fP, int *\fIovector\fP,
2314 .ti +5n
2315 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2316 .ti +5n
2317 .B const char **\fIstringptr\fP);
2318 .PP
2319 To extract a substring by name, you first have to find associated number.
2320 For example, for this pattern
2321 .sp
2322 (a+)b(?<xxx>\ed+)...
2323 .sp
2324 the number of the subpattern called "xxx" is 2. If the name is known to be
2325 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2326 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2327 pattern, and the second is the name. The yield of the function is the
2328 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2329 that name.
2330 .P
2331 Given the number, you can extract the substring directly, or use one of the
2332 functions described in the previous section. For convenience, there are also
2333 two functions that do the whole job.
2334 .P
2335 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2336 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2337 functions that extract by number. As these are described in the previous
2338 section, they are not re-described here. There are just two differences:
2339 .P
2340 First, instead of a substring number, a substring name is given. Second, there
2341 is an extra argument, given at the start, which is a pointer to the compiled
2342 pattern. This is needed in order to gain access to the name-to-number
2343 translation table.
2344 .P
2345 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2346 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2347 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2348 the behaviour may not be what you want (see the next section).
2349 .P
2350 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2351 subpatterns with the same number, as described in the
2352 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2353 .\" </a>
2354 section on duplicate subpattern numbers
2355 .\"
2356 in the
2357 .\" HREF
2358 \fBpcrepattern\fP
2359 .\"
2360 page, you cannot use names to distinguish the different subpatterns, because
2361 names are not included in the compiled code. The matching process uses only
2362 numbers. For this reason, the use of different names for subpatterns of the
2363 same number causes an error at compile time.
2364 .
2365 .
2367 .rs
2368 .sp
2369 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2370 .ti +5n
2371 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
2372 .PP
2373 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2374 are not required to be unique. (Duplicate names are always allowed for
2375 subpatterns with the same number, created by using the (?| feature. Indeed, if
2376 such subpatterns are named, they are required to use the same names.)
2377 .P
2378 Normally, patterns with duplicate names are such that in any one match, only
2379 one of the named subpatterns participates. An example is shown in the
2380 .\" HREF
2381 \fBpcrepattern\fP
2382 .\"
2383 documentation.
2384 .P
2385 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2386 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2387 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2388 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2389 returns one of the numbers that are associated with the name, but it is not
2390 defined which it is.
2391 .P
2392 If you want to get full details of all captured substrings for a given name,
2393 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2394 argument is the compiled pattern, and the second is the name. The third and
2395 fourth are pointers to variables which are updated by the function. After it
2396 has run, they point to the first and last entries in the name-to-number table
2397 for the given name. The function itself returns the length of each entry, or
2398 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2399 described above in the section entitled \fIInformation about a pattern\fP
2400 .\" HTML <a href="#infoaboutpattern">
2401 .\" </a>
2402 above.
2403 .\"
2404 Given all the relevant entries for the name, you can extract each of their
2405 numbers, and hence the captured data, if any.
2406 .
2407 .
2409 .rs
2410 .sp
2411 The traditional matching function uses a similar algorithm to Perl, which stops
2412 when it finds the first match, starting at a given point in the subject. If you
2413 want to find all possible matches, or the longest possible match, consider
2414 using the alternative matching function (see below) instead. If you cannot use
2415 the alternative function, but still need to find all possible matches, you
2416 can kludge it up by making use of the callout facility, which is described in
2417 the
2418 .\" HREF
2419 \fBpcrecallout\fP
2420 .\"
2421 documentation.
2422 .P
2423 What you have to do is to insert a callout right at the end of the pattern.
2424 When your callout function is called, extract and save the current matched
2425 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2426 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2427 will yield PCRE_ERROR_NOMATCH.
2428 .
2429 .
2431 .rs
2432 .sp
2433 Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2434 stack, which in certain environments can be rather limited in size. Some users
2435 find it helpful to have an estimate of the amount of stack that is used by
2436 \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2437 .\" HREF
2438 \fBpcrestack\fP
2439 .\"
2440 documentation. The estimate that is output by \fBpcretest\fP when called with
2441 the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2442 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2443 .P
2444 Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2445 the negative error code PCRE_ERROR_NULL, but with this special combination of
2446 arguments, it returns instead a negative number whose absolute value is the
2447 approximate stack frame size in bytes. (A negative number is used so that it is
2448 clear that no match has happened.) The value is approximate because in some
2449 cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2450 additional variables on the stack.
2451 .P
2452 If PCRE has been compiled to use the heap instead of the stack for recursion,
2453 the value returned is the size of each block that is obtained from the heap.
2454 .
2455 .
2456 .\" HTML <a name="dfamatch"></a>
2458 .rs
2459 .sp
2460 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2461 .ti +5n
2462 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2463 .ti +5n
2464 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2465 .ti +5n
2466 .B int *\fIworkspace\fP, int \fIwscount\fP);
2467 .P
2468 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2469 a compiled pattern, using a matching algorithm that scans the subject string
2470 just once, and does not backtrack. This has different characteristics to the
2471 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2472 patterns are not supported. Nevertheless, there are times when this kind of
2473 matching can be useful. For a discussion of the two matching algorithms, and a
2474 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2475 .\" HREF
2476 \fBpcrematching\fP
2477 .\"
2478 documentation.
2479 .P
2480 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2481 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2482 different way, and this is described below. The other common arguments are used
2483 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2484 here.
2485 .P
2486 The two additional arguments provide workspace for the function. The workspace
2487 vector should contain at least 20 elements. It is used for keeping track of
2488 multiple paths through the pattern tree. More workspace will be needed for
2489 patterns and subjects where there are a lot of potential matches.
2490 .P
2491 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2492 .sp
2493 int rc;
2494 int ovector[10];
2495 int wspace[20];
2496 rc = pcre_dfa_exec(
2497 re, /* result of pcre_compile() */
2498 NULL, /* we didn't study the pattern */
2499 "some string", /* the subject string */
2500 11, /* the length of the subject string */
2501 0, /* start at offset 0 in the subject */
2502 0, /* default options */
2503 ovector, /* vector of integers for substring information */
2504 10, /* number of elements (NOT size in bytes) */
2505 wspace, /* working space vector */
2506 20); /* number of elements (NOT size in bytes) */
2507 .
2508 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2509 .rs
2510 .sp
2511 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2512 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2516 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2517 so their description is not repeated here.
2518 .sp
2521 .sp
2522 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2523 details are slightly different. When PCRE_PARTIAL_HARD is set for
2524 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2525 is reached and there is still at least one matching possibility that requires
2526 additional characters. This happens even if some complete matches have also
2527 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2528 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2529 there have been no complete matches, but there is still at least one matching
2530 possibility. The portion of the string that was inspected when the longest
2531 partial match was found is set as the first matching string in both cases.
2532 There is a more detailed discussion of partial and multi-segment matching, with
2533 examples, in the
2534 .\" HREF
2535 \fBpcrepartial\fP
2536 .\"
2537 documentation.
2538 .sp
2540 .sp
2541 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2542 soon as it has found one match. Because of the way the alternative algorithm
2543 works, this is necessarily the shortest possible match at the first possible
2544 matching point in the subject string.
2545 .sp
2547 .sp
2548 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2549 again, with additional subject characters, and have it continue with the same
2550 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2551 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2552 before because data about the match so far is left in them after a partial
2553 match. There is more discussion of this facility in the
2554 .\" HREF
2555 \fBpcrepartial\fP
2556 .\"
2557 documentation.
2558 .
2559 .
2560 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2561 .rs
2562 .sp
2563 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2564 substring in the subject. Note, however, that all the matches from one run of
2565 the function start at the same point in the subject. The shorter matches are
2566 all initial substrings of the longer matches. For example, if the pattern
2567 .sp
2568 <.*>
2569 .sp
2570 is matched against the string
2571 .sp
2572 This is <something> <something else> <something further> no more
2573 .sp
2574 the three matched strings are
2575 .sp
2576 <something>
2577 <something> <something else>
2578 <something> <something else> <something further>
2579 .sp
2580 On success, the yield of the function is a number greater than zero, which is
2581 the number of matched substrings. The substrings themselves are returned in
2582 \fIovector\fP. Each string uses two elements; the first is the offset to the
2583 start, and the second is the offset to the end. In fact, all the strings have
2584 the same start offset. (Space could have been saved by giving this only once,
2585 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2586 returns data, even though the meaning of the strings is different.)
2587 .P
2588 The strings are returned in reverse order of length; that is, the longest
2589 matching string is given first. If there were too many matches to fit into
2590 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2591 the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2592 the entire \fIovector\fP for returning matched strings.
2593 .
2594 .
2595 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2596 .rs
2597 .sp
2598 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2599 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2600 described
2601 .\" HTML <a href="#errorlist">
2602 .\" </a>
2603 above.
2604 .\"
2605 There are in addition the following errors that are specific to
2606 \fBpcre_dfa_exec()\fP:
2607 .sp
2609 .sp
2610 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2611 that it does not support, for instance, the use of \eC or a back reference.
2612 .sp
2614 .sp
2615 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2616 uses a back reference for the condition, or a test for recursion in a specific
2617 group. These are not supported.
2618 .sp
2620 .sp
2621 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2622 block that contains a setting of the \fImatch_limit\fP or
2623 \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2624 meaningless for DFA matching).
2625 .sp
2627 .sp
2628 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2629 \fIworkspace\fP vector.
2630 .sp
2632 .sp
2633 When a recursive subpattern is processed, the matching function calls itself
2634 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2635 error is given if the output vector is not large enough. This should be
2636 extremely rare, as a vector of size 1000 is used.
2637 .sp
2639 .sp
2640 When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2641 some plausibility checks are made on the contents of the workspace, which
2642 should contain data about the previous partial match. If any of these checks
2643 fail, this error is given.
2644 .
2645 .
2646 .SH "SEE ALSO"
2647 .rs
2648 .sp
2649 \fBpcre16\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
2650 \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
2651 \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
2652 .
2653 .
2655 .rs
2656 .sp
2657 .nf
2658 Philip Hazel
2659 University Computing Service
2660 Cambridge CB2 3QH, England.
2661 .fi
2662 .
2663 .
2665 .rs
2666 .sp
2667 .nf
2668 Last updated: 19 April 2012
2669 Copyright (c) 1997-2012 University of Cambridge.
2670 .fi


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

  ViewVC Help
Powered by ViewVC 1.1.5