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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 569 - (show annotations)
Sun Nov 7 16:14:50 2010 UTC (8 years, 9 months ago) by ph10
File size: 97761 byte(s)
Add PCRE_ERROR_SHORTUTF8 to PCRE_PARTIAL_HARD processing.
1 .TH PCREAPI 3
2 .SH NAME
3 PCRE - Perl-compatible regular expressions
4 .SH "PCRE NATIVE API"
5 .rs
6 .sp
7 .B #include <pcre.h>
8 .PP
9 .SM
10 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
11 .ti +5n
12 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
13 .ti +5n
14 .B const unsigned char *\fItableptr\fP);
15 .PP
16 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
17 .ti +5n
18 .B int *\fIerrorcodeptr\fP,
19 .ti +5n
20 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
21 .ti +5n
22 .B const unsigned char *\fItableptr\fP);
23 .PP
24 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
25 .ti +5n
26 .B const char **\fIerrptr\fP);
27 .PP
28 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
29 .ti +5n
30 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
31 .ti +5n
32 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
33 .PP
34 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
35 .ti +5n
36 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
37 .ti +5n
38 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
39 .ti +5n
40 .B int *\fIworkspace\fP, int \fIwscount\fP);
41 .PP
42 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
43 .ti +5n
44 .B const char *\fIsubject\fP, int *\fIovector\fP,
45 .ti +5n
46 .B int \fIstringcount\fP, const char *\fIstringname\fP,
47 .ti +5n
48 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
49 .PP
50 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
51 .ti +5n
52 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
53 .ti +5n
54 .B int \fIbuffersize\fP);
55 .PP
56 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
57 .ti +5n
58 .B const char *\fIsubject\fP, int *\fIovector\fP,
59 .ti +5n
60 .B int \fIstringcount\fP, const char *\fIstringname\fP,
61 .ti +5n
62 .B const char **\fIstringptr\fP);
63 .PP
64 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
65 .ti +5n
66 .B const char *\fIname\fP);
67 .PP
68 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
69 .ti +5n
70 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
71 .PP
72 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
73 .ti +5n
74 .B int \fIstringcount\fP, int \fIstringnumber\fP,
75 .ti +5n
76 .B const char **\fIstringptr\fP);
77 .PP
78 .B int pcre_get_substring_list(const char *\fIsubject\fP,
79 .ti +5n
80 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
81 .PP
82 .B void pcre_free_substring(const char *\fIstringptr\fP);
83 .PP
84 .B void pcre_free_substring_list(const char **\fIstringptr\fP);
85 .PP
86 .B const unsigned char *pcre_maketables(void);
87 .PP
88 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
89 .ti +5n
90 .B int \fIwhat\fP, void *\fIwhere\fP);
91 .PP
92 .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
93 .B *\fIfirstcharptr\fP);
94 .PP
95 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
96 .PP
97 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
98 .PP
99 .B char *pcre_version(void);
100 .PP
101 .B void *(*pcre_malloc)(size_t);
102 .PP
103 .B void (*pcre_free)(void *);
104 .PP
105 .B void *(*pcre_stack_malloc)(size_t);
106 .PP
107 .B void (*pcre_stack_free)(void *);
108 .PP
109 .B int (*pcre_callout)(pcre_callout_block *);
110 .
111 .
112 .SH "PCRE API OVERVIEW"
113 .rs
114 .sp
115 PCRE has its own native API, which is described in this document. There are
116 also some wrapper functions that correspond to the POSIX regular expression
117 API. These are described in the
118 .\" HREF
119 \fBpcreposix\fP
120 .\"
121 documentation. Both of these APIs define a set of C function calls. A C++
122 wrapper is distributed with PCRE. It is documented in the
123 .\" HREF
124 \fBpcrecpp\fP
125 .\"
126 page.
127 .P
128 The native API C function prototypes are defined in the header file
129 \fBpcre.h\fP, and on Unix systems the library itself is called \fBlibpcre\fP.
130 It can normally be accessed by adding \fB-lpcre\fP to the command for linking
131 an application that uses PCRE. The header file defines the macros PCRE_MAJOR
132 and PCRE_MINOR to contain the major and minor release numbers for the library.
133 Applications can use these to include support for different releases of PCRE.
134 .P
135 In a Windows environment, if you want to statically link an application program
136 against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
137 including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
138 \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
139 \fB__declspec(dllimport)\fP, with unwanted results.
140 .P
141 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
142 and \fBpcre_exec()\fP are used for compiling and matching regular expressions
143 in a Perl-compatible manner. A sample program that demonstrates the simplest
144 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
145 source distribution. A listing of this program is given in the
146 .\" HREF
147 \fBpcredemo\fP
148 .\"
149 documentation, and the
150 .\" HREF
151 \fBpcresample\fP
152 .\"
153 documentation describes how to compile and run it.
154 .P
155 A second matching function, \fBpcre_dfa_exec()\fP, which is not
156 Perl-compatible, is also provided. This uses a different algorithm for the
157 matching. The alternative algorithm finds all possible matches (at a given
158 point in the subject), and scans the subject just once (unless there are
159 lookbehind assertions). However, this algorithm does not return captured
160 substrings. A description of the two matching algorithms and their advantages
161 and disadvantages is given in the
162 .\" HREF
163 \fBpcrematching\fP
164 .\"
165 documentation.
166 .P
167 In addition to the main compiling and matching functions, there are convenience
168 functions for extracting captured substrings from a subject string that is
169 matched by \fBpcre_exec()\fP. They are:
170 .sp
171 \fBpcre_copy_substring()\fP
172 \fBpcre_copy_named_substring()\fP
173 \fBpcre_get_substring()\fP
174 \fBpcre_get_named_substring()\fP
175 \fBpcre_get_substring_list()\fP
176 \fBpcre_get_stringnumber()\fP
177 \fBpcre_get_stringtable_entries()\fP
178 .sp
179 \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
180 provided, to free the memory used for extracted strings.
181 .P
182 The function \fBpcre_maketables()\fP is used to build a set of character tables
183 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
184 or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
185 specialist use. Most commonly, no special tables are passed, in which case
186 internal tables that are generated when PCRE is built are used.
187 .P
188 The function \fBpcre_fullinfo()\fP is used to find out information about a
189 compiled pattern; \fBpcre_info()\fP is an obsolete version that returns only
190 some of the available information, but is retained for backwards compatibility.
191 The function \fBpcre_version()\fP returns a pointer to a string containing the
192 version of PCRE and its date of release.
193 .P
194 The function \fBpcre_refcount()\fP maintains a reference count in a data block
195 containing a compiled pattern. This is provided for the benefit of
196 object-oriented applications.
197 .P
198 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
199 the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
200 respectively. PCRE calls the memory management functions via these variables,
201 so a calling program can replace them if it wishes to intercept the calls. This
202 should be done before calling any PCRE functions.
203 .P
204 The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
205 indirections to memory management functions. These special functions are used
206 only when PCRE is compiled to use the heap for remembering data, instead of
207 recursive function calls, when running the \fBpcre_exec()\fP function. See the
208 .\" HREF
209 \fBpcrebuild\fP
210 .\"
211 documentation for details of how to do this. It is a non-standard way of
212 building PCRE, for use in environments that have limited stacks. Because of the
213 greater use of memory management, it runs more slowly. Separate functions are
214 provided so that special-purpose external code can be used for this case. When
215 used, these functions are always called in a stack-like manner (last obtained,
216 first freed), and always for memory blocks of the same size. There is a
217 discussion about PCRE's stack usage in the
218 .\" HREF
219 \fBpcrestack\fP
220 .\"
221 documentation.
222 .P
223 The global variable \fBpcre_callout\fP initially contains NULL. It can be set
224 by the caller to a "callout" function, which PCRE will then call at specified
225 points during a matching operation. Details are given in the
226 .\" HREF
227 \fBpcrecallout\fP
228 .\"
229 documentation.
230 .
231 .
232 .\" HTML <a name="newlines"></a>
233 .SH NEWLINES
234 .rs
235 .sp
236 PCRE supports five different conventions for indicating line breaks in
237 strings: a single CR (carriage return) character, a single LF (linefeed)
238 character, the two-character sequence CRLF, any of the three preceding, or any
239 Unicode newline sequence. The Unicode newline sequences are the three just
240 mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
241 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
242 (paragraph separator, U+2029).
243 .P
244 Each of the first three conventions is used by at least one operating system as
245 its standard newline sequence. When PCRE is built, a default can be specified.
246 The default default is LF, which is the Unix standard. When PCRE is run, the
247 default can be overridden, either when a pattern is compiled, or when it is
248 matched.
249 .P
250 At compile time, the newline convention can be specified by the \fIoptions\fP
251 argument of \fBpcre_compile()\fP, or it can be specified by special text at the
252 start of the pattern itself; this overrides any other settings. See the
253 .\" HREF
254 \fBpcrepattern\fP
255 .\"
256 page for details of the special character sequences.
257 .P
258 In the PCRE documentation the word "newline" is used to mean "the character or
259 pair of characters that indicate a line break". The choice of newline
260 convention affects the handling of the dot, circumflex, and dollar
261 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
262 recognized line ending sequence, the match position advancement for a
263 non-anchored pattern. There is more detail about this in the
264 .\" HTML <a href="#execoptions">
265 .\" </a>
266 section on \fBpcre_exec()\fP options
267 .\"
268 below.
269 .P
270 The choice of newline convention does not affect the interpretation of
271 the \en or \er escape sequences, nor does it affect what \eR matches, which is
272 controlled in a similar way, but by separate options.
273 .
274 .
275 .SH MULTITHREADING
276 .rs
277 .sp
278 The PCRE functions can be used in multi-threading applications, with the
279 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
280 \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
281 callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
282 .P
283 The compiled form of a regular expression is not altered during matching, so
284 the same compiled pattern can safely be used by several threads at once.
285 .
286 .
287 .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
288 .rs
289 .sp
290 The compiled form of a regular expression can be saved and re-used at a later
291 time, possibly by a different program, and even on a host other than the one on
292 which it was compiled. Details are given in the
293 .\" HREF
294 \fBpcreprecompile\fP
295 .\"
296 documentation. However, compiling a regular expression with one version of PCRE
297 for use with a different version is not guaranteed to work and may cause
298 crashes.
299 .
300 .
301 .SH "CHECKING BUILD-TIME OPTIONS"
302 .rs
303 .sp
304 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
305 .PP
306 The function \fBpcre_config()\fP makes it possible for a PCRE client to
307 discover which optional features have been compiled into the PCRE library. The
308 .\" HREF
309 \fBpcrebuild\fP
310 .\"
311 documentation has more details about these optional features.
312 .P
313 The first argument for \fBpcre_config()\fP is an integer, specifying which
314 information is required; the second argument is a pointer to a variable into
315 which the information is placed. The following information is available:
316 .sp
317 PCRE_CONFIG_UTF8
318 .sp
319 The output is an integer that is set to one if UTF-8 support is available;
320 otherwise it is set to zero.
321 .sp
322 PCRE_CONFIG_UNICODE_PROPERTIES
323 .sp
324 The output is an integer that is set to one if support for Unicode character
325 properties is available; otherwise it is set to zero.
326 .sp
327 PCRE_CONFIG_NEWLINE
328 .sp
329 The output is an integer whose value specifies the default character sequence
330 that is recognized as meaning "newline". The four values that are supported
331 are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF, and -1 for ANY.
332 Though they are derived from ASCII, the same values are returned in EBCDIC
333 environments. The default should normally correspond to the standard sequence
334 for your operating system.
335 .sp
336 PCRE_CONFIG_BSR
337 .sp
338 The output is an integer whose value indicates what character sequences the \eR
339 escape sequence matches by default. A value of 0 means that \eR matches any
340 Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
341 or CRLF. The default can be overridden when a pattern is compiled or matched.
342 .sp
343 PCRE_CONFIG_LINK_SIZE
344 .sp
345 The output is an integer that contains the number of bytes used for internal
346 linkage in compiled regular expressions. The value is 2, 3, or 4. Larger values
347 allow larger regular expressions to be compiled, at the expense of slower
348 matching. The default value of 2 is sufficient for all but the most massive
349 patterns, since it allows the compiled pattern to be up to 64K in size.
350 .sp
351 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
352 .sp
353 The output is an integer that contains the threshold above which the POSIX
354 interface uses \fBmalloc()\fP for output vectors. Further details are given in
355 the
356 .\" HREF
357 \fBpcreposix\fP
358 .\"
359 documentation.
360 .sp
361 PCRE_CONFIG_MATCH_LIMIT
362 .sp
363 The output is a long integer that gives the default limit for the number of
364 internal matching function calls in a \fBpcre_exec()\fP execution. Further
365 details are given with \fBpcre_exec()\fP below.
366 .sp
367 PCRE_CONFIG_MATCH_LIMIT_RECURSION
368 .sp
369 The output is a long integer that gives the default limit for the depth of
370 recursion when calling the internal matching function in a \fBpcre_exec()\fP
371 execution. Further details are given with \fBpcre_exec()\fP below.
372 .sp
373 PCRE_CONFIG_STACKRECURSE
374 .sp
375 The output is an integer that is set to one if internal recursion when running
376 \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
377 to remember their state. This is the usual way that PCRE is compiled. The
378 output is zero if PCRE was compiled to use blocks of data on the heap instead
379 of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
380 \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
381 avoiding the use of the stack.
382 .
383 .
384 .SH "COMPILING A PATTERN"
385 .rs
386 .sp
387 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
388 .ti +5n
389 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
390 .ti +5n
391 .B const unsigned char *\fItableptr\fP);
392 .sp
393 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
394 .ti +5n
395 .B int *\fIerrorcodeptr\fP,
396 .ti +5n
397 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
398 .ti +5n
399 .B const unsigned char *\fItableptr\fP);
400 .P
401 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
402 called to compile a pattern into an internal form. The only difference between
403 the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
404 \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
405 too much repetition, we refer just to \fBpcre_compile()\fP below, but the
406 information applies equally to \fBpcre_compile2()\fP.
407 .P
408 The pattern is a C string terminated by a binary zero, and is passed in the
409 \fIpattern\fP argument. A pointer to a single block of memory that is obtained
410 via \fBpcre_malloc\fP is returned. This contains the compiled code and related
411 data. The \fBpcre\fP type is defined for the returned block; this is a typedef
412 for a structure whose contents are not externally defined. It is up to the
413 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
414 .P
415 Although the compiled code of a PCRE regex is relocatable, that is, it does not
416 depend on memory location, the complete \fBpcre\fP data block is not
417 fully relocatable, because it may contain a copy of the \fItableptr\fP
418 argument, which is an address (see below).
419 .P
420 The \fIoptions\fP argument contains various bit settings that affect the
421 compilation. It should be zero if no options are required. The available
422 options are described below. Some of them (in particular, those that are
423 compatible with Perl, but some others as well) can also be set and unset from
424 within the pattern (see the detailed description in the
425 .\" HREF
426 \fBpcrepattern\fP
427 .\"
428 documentation). For those options that can be different in different parts of
429 the pattern, the contents of the \fIoptions\fP argument specifies their
430 settings at the start of compilation and execution. The PCRE_ANCHORED,
431 PCRE_BSR_\fIxxx\fP, and PCRE_NEWLINE_\fIxxx\fP options can be set at the time
432 of matching as well as at compile time.
433 .P
434 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
435 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
436 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
437 error message. This is a static string that is part of the library. You must
438 not try to free it. The offset from the start of the pattern to the byte that
439 was being processed when the error was discovered is placed in the variable
440 pointed to by \fIerroffset\fP, which must not be NULL. If it is, an immediate
441 error is given. Some errors are not detected until checks are carried out when
442 the whole pattern has been scanned; in this case the offset is set to the end
443 of the pattern.
444 .P
445 Note that the offset is in bytes, not characters, even in UTF-8 mode. It may
446 point into the middle of a UTF-8 character (for example, when
447 PCRE_ERROR_BADUTF8 is returned for an invalid UTF-8 string).
448 .P
449 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
450 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
451 returned via this argument in the event of an error. This is in addition to the
452 textual error message. Error codes and messages are listed below.
453 .P
454 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
455 character tables that are built when PCRE is compiled, using the default C
456 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
457 call to \fBpcre_maketables()\fP. This value is stored with the compiled
458 pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
459 passed to it. For more discussion, see the section on locale support below.
460 .P
461 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
462 .sp
463 pcre *re;
464 const char *error;
465 int erroffset;
466 re = pcre_compile(
467 "^A.*Z", /* the pattern */
468 0, /* default options */
469 &error, /* for error message */
470 &erroffset, /* for error offset */
471 NULL); /* use default character tables */
472 .sp
473 The following names for option bits are defined in the \fBpcre.h\fP header
474 file:
475 .sp
476 PCRE_ANCHORED
477 .sp
478 If this bit is set, the pattern is forced to be "anchored", that is, it is
479 constrained to match only at the first matching point in the string that is
480 being searched (the "subject string"). This effect can also be achieved by
481 appropriate constructs in the pattern itself, which is the only way to do it in
482 Perl.
483 .sp
484 PCRE_AUTO_CALLOUT
485 .sp
486 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
487 all with number 255, before each pattern item. For discussion of the callout
488 facility, see the
489 .\" HREF
490 \fBpcrecallout\fP
491 .\"
492 documentation.
493 .sp
494 PCRE_BSR_ANYCRLF
495 PCRE_BSR_UNICODE
496 .sp
497 These options (which are mutually exclusive) control what the \eR escape
498 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
499 match any Unicode newline sequence. The default is specified when PCRE is
500 built. It can be overridden from within the pattern, or by setting an option
501 when a compiled pattern is matched.
502 .sp
503 PCRE_CASELESS
504 .sp
505 If this bit is set, letters in the pattern match both upper and lower case
506 letters. It is equivalent to Perl's /i option, and it can be changed within a
507 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
508 concept of case for characters whose values are less than 128, so caseless
509 matching is always possible. For characters with higher values, the concept of
510 case is supported if PCRE is compiled with Unicode property support, but not
511 otherwise. If you want to use caseless matching for characters 128 and above,
512 you must ensure that PCRE is compiled with Unicode property support as well as
513 with UTF-8 support.
514 .sp
515 PCRE_DOLLAR_ENDONLY
516 .sp
517 If this bit is set, a dollar metacharacter in the pattern matches only at the
518 end of the subject string. Without this option, a dollar also matches
519 immediately before a newline at the end of the string (but not before any other
520 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
521 There is no equivalent to this option in Perl, and no way to set it within a
522 pattern.
523 .sp
524 PCRE_DOTALL
525 .sp
526 If this bit is set, a dot metacharater in the pattern matches all characters,
527 including those that indicate newline. Without it, a dot does not match when
528 the current position is at a newline. This option is equivalent to Perl's /s
529 option, and it can be changed within a pattern by a (?s) option setting. A
530 negative class such as [^a] always matches newline characters, independent of
531 the setting of this option.
532 .sp
533 PCRE_DUPNAMES
534 .sp
535 If this bit is set, names used to identify capturing subpatterns need not be
536 unique. This can be helpful for certain types of pattern when it is known that
537 only one instance of the named subpattern can ever be matched. There are more
538 details of named subpatterns below; see also the
539 .\" HREF
540 \fBpcrepattern\fP
541 .\"
542 documentation.
543 .sp
544 PCRE_EXTENDED
545 .sp
546 If this bit is set, whitespace data characters in the pattern are totally
547 ignored except when escaped or inside a character class. Whitespace does not
548 include the VT character (code 11). In addition, characters between an
549 unescaped # outside a character class and the next newline, inclusive, are also
550 ignored. This is equivalent to Perl's /x option, and it can be changed within a
551 pattern by a (?x) option setting.
552 .P
553 This option makes it possible to include comments inside complicated patterns.
554 Note, however, that this applies only to data characters. Whitespace characters
555 may never appear within special character sequences in a pattern, for example
556 within the sequence (?( which introduces a conditional subpattern.
557 .sp
558 PCRE_EXTRA
559 .sp
560 This option was invented in order to turn on additional functionality of PCRE
561 that is incompatible with Perl, but it is currently of very little use. When
562 set, any backslash in a pattern that is followed by a letter that has no
563 special meaning causes an error, thus reserving these combinations for future
564 expansion. By default, as in Perl, a backslash followed by a letter with no
565 special meaning is treated as a literal. (Perl can, however, be persuaded to
566 give an error for this, by running it with the -w option.) There are at present
567 no other features controlled by this option. It can also be set by a (?X)
568 option setting within a pattern.
569 .sp
570 PCRE_FIRSTLINE
571 .sp
572 If this option is set, an unanchored pattern is required to match before or at
573 the first newline in the subject string, though the matched text may continue
574 over the newline.
575 .sp
576 PCRE_JAVASCRIPT_COMPAT
577 .sp
578 If this option is set, PCRE's behaviour is changed in some ways so that it is
579 compatible with JavaScript rather than Perl. The changes are as follows:
580 .P
581 (1) A lone closing square bracket in a pattern causes a compile-time error,
582 because this is illegal in JavaScript (by default it is treated as a data
583 character). Thus, the pattern AB]CD becomes illegal when this option is set.
584 .P
585 (2) At run time, a back reference to an unset subpattern group matches an empty
586 string (by default this causes the current matching alternative to fail). A
587 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
588 an "a" in the subject), whereas it fails by default, for Perl compatibility.
589 .sp
590 PCRE_MULTILINE
591 .sp
592 By default, PCRE treats the subject string as consisting of a single line of
593 characters (even if it actually contains newlines). The "start of line"
594 metacharacter (^) matches only at the start of the string, while the "end of
595 line" metacharacter ($) matches only at the end of the string, or before a
596 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
597 Perl.
598 .P
599 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
600 match immediately following or immediately before internal newlines in the
601 subject string, respectively, as well as at the very start and end. This is
602 equivalent to Perl's /m option, and it can be changed within a pattern by a
603 (?m) option setting. If there are no newlines in a subject string, or no
604 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
605 .sp
606 PCRE_NEWLINE_CR
607 PCRE_NEWLINE_LF
608 PCRE_NEWLINE_CRLF
609 PCRE_NEWLINE_ANYCRLF
610 PCRE_NEWLINE_ANY
611 .sp
612 These options override the default newline definition that was chosen when PCRE
613 was built. Setting the first or the second specifies that a newline is
614 indicated by a single character (CR or LF, respectively). Setting
615 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
616 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
617 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
618 that any Unicode newline sequence should be recognized. The Unicode newline
619 sequences are the three just mentioned, plus the single characters VT (vertical
620 tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
621 separator, U+2028), and PS (paragraph separator, U+2029). The last two are
622 recognized only in UTF-8 mode.
623 .P
624 The newline setting in the options word uses three bits that are treated
625 as a number, giving eight possibilities. Currently only six are used (default
626 plus the five values above). This means that if you set more than one newline
627 option, the combination may or may not be sensible. For example,
628 PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
629 other combinations may yield unused numbers and cause an error.
630 .P
631 The only time that a line break is specially recognized when compiling a
632 pattern is if PCRE_EXTENDED is set, and an unescaped # outside a character
633 class is encountered. This indicates a comment that lasts until after the next
634 line break sequence. In other circumstances, line break sequences are treated
635 as literal data, except that in PCRE_EXTENDED mode, both CR and LF are treated
636 as whitespace characters and are therefore ignored.
637 .P
638 The newline option that is set at compile time becomes the default that is used
639 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
640 .sp
641 PCRE_NO_AUTO_CAPTURE
642 .sp
643 If this option is set, it disables the use of numbered capturing parentheses in
644 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
645 were followed by ?: but named parentheses can still be used for capturing (and
646 they acquire numbers in the usual way). There is no equivalent of this option
647 in Perl.
648 .sp
649 PCRE_UCP
650 .sp
651 This option changes the way PCRE processes \eb, \ed, \es, \ew, and some of the
652 POSIX character classes. By default, only ASCII characters are recognized, but
653 if PCRE_UCP is set, Unicode properties are used instead to classify characters.
654 More details are given in the section on
655 .\" HTML <a href="pcre.html#genericchartypes">
656 .\" </a>
657 generic character types
658 .\"
659 in the
660 .\" HREF
661 \fBpcrepattern\fP
662 .\"
663 page. If you set PCRE_UCP, matching one of the items it affects takes much
664 longer. The option is available only if PCRE has been compiled with Unicode
665 property support.
666 .sp
667 PCRE_UNGREEDY
668 .sp
669 This option inverts the "greediness" of the quantifiers so that they are not
670 greedy by default, but become greedy if followed by "?". It is not compatible
671 with Perl. It can also be set by a (?U) option setting within the pattern.
672 .sp
673 PCRE_UTF8
674 .sp
675 This option causes PCRE to regard both the pattern and the subject as strings
676 of UTF-8 characters instead of single-byte character strings. However, it is
677 available only when PCRE is built to include UTF-8 support. If not, the use
678 of this option provokes an error. Details of how this option changes the
679 behaviour of PCRE are given in the
680 .\" HTML <a href="pcre.html#utf8support">
681 .\" </a>
682 section on UTF-8 support
683 .\"
684 in the main
685 .\" HREF
686 \fBpcre\fP
687 .\"
688 page.
689 .sp
690 PCRE_NO_UTF8_CHECK
691 .sp
692 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
693 automatically checked. There is a discussion about the
694 .\" HTML <a href="pcre.html#utf8strings">
695 .\" </a>
696 validity of UTF-8 strings
697 .\"
698 in the main
699 .\" HREF
700 \fBpcre\fP
701 .\"
702 page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_compile()\fP
703 returns an error. If you already know that your pattern is valid, and you want
704 to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
705 option. When it is set, the effect of passing an invalid UTF-8 string as a
706 pattern is undefined. It may cause your program to crash. Note that this option
707 can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
708 the UTF-8 validity checking of subject strings.
709 .
710 .
711 .SH "COMPILATION ERROR CODES"
712 .rs
713 .sp
714 The following table lists the error codes than may be returned by
715 \fBpcre_compile2()\fP, along with the error messages that may be returned by
716 both compiling functions. As PCRE has developed, some error codes have fallen
717 out of use. To avoid confusion, they have not been re-used.
718 .sp
719 0 no error
720 1 \e at end of pattern
721 2 \ec at end of pattern
722 3 unrecognized character follows \e
723 4 numbers out of order in {} quantifier
724 5 number too big in {} quantifier
725 6 missing terminating ] for character class
726 7 invalid escape sequence in character class
727 8 range out of order in character class
728 9 nothing to repeat
729 10 [this code is not in use]
730 11 internal error: unexpected repeat
731 12 unrecognized character after (? or (?-
732 13 POSIX named classes are supported only within a class
733 14 missing )
734 15 reference to non-existent subpattern
735 16 erroffset passed as NULL
736 17 unknown option bit(s) set
737 18 missing ) after comment
738 19 [this code is not in use]
739 20 regular expression is too large
740 21 failed to get memory
741 22 unmatched parentheses
742 23 internal error: code overflow
743 24 unrecognized character after (?<
744 25 lookbehind assertion is not fixed length
745 26 malformed number or name after (?(
746 27 conditional group contains more than two branches
747 28 assertion expected after (?(
748 29 (?R or (?[+-]digits must be followed by )
749 30 unknown POSIX class name
750 31 POSIX collating elements are not supported
751 32 this version of PCRE is not compiled with PCRE_UTF8 support
752 33 [this code is not in use]
753 34 character value in \ex{...} sequence is too large
754 35 invalid condition (?(0)
755 36 \eC not allowed in lookbehind assertion
756 37 PCRE does not support \eL, \el, \eN, \eU, or \eu
757 38 number after (?C is > 255
758 39 closing ) for (?C expected
759 40 recursive call could loop indefinitely
760 41 unrecognized character after (?P
761 42 syntax error in subpattern name (missing terminator)
762 43 two named subpatterns have the same name
763 44 invalid UTF-8 string
764 45 support for \eP, \ep, and \eX has not been compiled
765 46 malformed \eP or \ep sequence
766 47 unknown property name after \eP or \ep
767 48 subpattern name is too long (maximum 32 characters)
768 49 too many named subpatterns (maximum 10000)
769 50 [this code is not in use]
770 51 octal value is greater than \e377 (not in UTF-8 mode)
771 52 internal error: overran compiling workspace
772 53 internal error: previously-checked referenced subpattern
773 not found
774 54 DEFINE group contains more than one branch
775 55 repeating a DEFINE group is not allowed
776 56 inconsistent NEWLINE options
777 57 \eg is not followed by a braced, angle-bracketed, or quoted
778 name/number or by a plain number
779 58 a numbered reference must not be zero
780 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
781 60 (*VERB) not recognized
782 61 number is too big
783 62 subpattern name expected
784 63 digit expected after (?+
785 64 ] is an invalid data character in JavaScript compatibility mode
786 65 different names for subpatterns of the same number are
787 not allowed
788 66 (*MARK) must have an argument
789 67 this version of PCRE is not compiled with PCRE_UCP support
790 .sp
791 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
792 be used if the limits were changed when PCRE was built.
793 .
794 .
795 .SH "STUDYING A PATTERN"
796 .rs
797 .sp
798 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
799 .ti +5n
800 .B const char **\fIerrptr\fP);
801 .PP
802 If a compiled pattern is going to be used several times, it is worth spending
803 more time analyzing it in order to speed up the time taken for matching. The
804 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
805 argument. If studying the pattern produces additional information that will
806 help speed up matching, \fBpcre_study()\fP returns a pointer to a
807 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
808 results of the study.
809 .P
810 The returned value from \fBpcre_study()\fP can be passed directly to
811 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
812 also contains other fields that can be set by the caller before the block is
813 passed; these are described
814 .\" HTML <a href="#extradata">
815 .\" </a>
816 below
817 .\"
818 in the section on matching a pattern.
819 .P
820 If studying the pattern does not produce any useful information,
821 \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program
822 wants to pass any of the other fields to \fBpcre_exec()\fP or
823 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block.
824 .P
825 The second argument of \fBpcre_study()\fP contains option bits. At present, no
826 options are defined, and this argument should always be zero.
827 .P
828 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
829 studying succeeds (even if no data is returned), the variable it points to is
830 set to NULL. Otherwise it is set to point to a textual error message. This is a
831 static string that is part of the library. You must not try to free it. You
832 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
833 sure that it has run successfully.
834 .P
835 This is a typical call to \fBpcre_study\fP():
836 .sp
837 pcre_extra *pe;
838 pe = pcre_study(
839 re, /* result of pcre_compile() */
840 0, /* no options exist */
841 &error); /* set to NULL or points to a message */
842 .sp
843 Studying a pattern does two things: first, a lower bound for the length of
844 subject string that is needed to match the pattern is computed. This does not
845 mean that there are any strings of that length that match, but it does
846 guarantee that no shorter strings match. The value is used by
847 \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP to avoid wasting time by trying to
848 match strings that are shorter than the lower bound. You can find out the value
849 in a calling program via the \fBpcre_fullinfo()\fP function.
850 .P
851 Studying a pattern is also useful for non-anchored patterns that do not have a
852 single fixed starting character. A bitmap of possible starting bytes is
853 created. This speeds up finding a position in the subject at which to start
854 matching.
855 .P
856 The two optimizations just described can be disabled by setting the
857 PCRE_NO_START_OPTIMIZE option when calling \fBpcre_exec()\fP or
858 \fBpcre_dfa_exec()\fP. You might want to do this if your pattern contains
859 callouts, or make use of (*MARK), and you make use of these in cases where
860 matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
861 .\" HTML <a href="#execoptions">
862 .\" </a>
863 below.
864 .\"
865 .
866 .
867 .\" HTML <a name="localesupport"></a>
868 .SH "LOCALE SUPPORT"
869 .rs
870 .sp
871 PCRE handles caseless matching, and determines whether characters are letters,
872 digits, or whatever, by reference to a set of tables, indexed by character
873 value. When running in UTF-8 mode, this applies only to characters with codes
874 less than 128. By default, higher-valued codes never match escapes such as \ew
875 or \ed, but they can be tested with \ep if PCRE is built with Unicode character
876 property support. Alternatively, the PCRE_UCP option can be set at compile
877 time; this causes \ew and friends to use Unicode property support instead of
878 built-in tables. The use of locales with Unicode is discouraged. If you are
879 handling characters with codes greater than 128, you should either use UTF-8
880 and Unicode, or use locales, but not try to mix the two.
881 .P
882 PCRE contains an internal set of tables that are used when the final argument
883 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
884 Normally, the internal tables recognize only ASCII characters. However, when
885 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
886 default "C" locale of the local system, which may cause them to be different.
887 .P
888 The internal tables can always be overridden by tables supplied by the
889 application that calls PCRE. These may be created in a different locale from
890 the default. As more and more applications change to using Unicode, the need
891 for this locale support is expected to die away.
892 .P
893 External tables are built by calling the \fBpcre_maketables()\fP function,
894 which has no arguments, in the relevant locale. The result can then be passed
895 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
896 example, to build and use tables that are appropriate for the French locale
897 (where accented characters with values greater than 128 are treated as letters),
898 the following code could be used:
899 .sp
900 setlocale(LC_CTYPE, "fr_FR");
901 tables = pcre_maketables();
902 re = pcre_compile(..., tables);
903 .sp
904 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
905 are using Windows, the name for the French locale is "french".
906 .P
907 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
908 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
909 that the memory containing the tables remains available for as long as it is
910 needed.
911 .P
912 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
913 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
914 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
915 pattern, compilation, studying and matching all happen in the same locale, but
916 different patterns can be compiled in different locales.
917 .P
918 It is possible to pass a table pointer or NULL (indicating the use of the
919 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
920 this facility could be used to match a pattern in a different locale from the
921 one in which it was compiled. Passing table pointers at run time is discussed
922 below in the section on matching a pattern.
923 .
924 .
925 .SH "INFORMATION ABOUT A PATTERN"
926 .rs
927 .sp
928 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
929 .ti +5n
930 .B int \fIwhat\fP, void *\fIwhere\fP);
931 .PP
932 The \fBpcre_fullinfo()\fP function returns information about a compiled
933 pattern. It replaces the obsolete \fBpcre_info()\fP function, which is
934 nevertheless retained for backwards compability (and is documented below).
935 .P
936 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
937 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
938 the pattern was not studied. The third argument specifies which piece of
939 information is required, and the fourth argument is a pointer to a variable
940 to receive the data. The yield of the function is zero for success, or one of
941 the following negative numbers:
942 .sp
943 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
944 the argument \fIwhere\fP was NULL
945 PCRE_ERROR_BADMAGIC the "magic number" was not found
946 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
947 .sp
948 The "magic number" is placed at the start of each compiled pattern as an simple
949 check against passing an arbitrary memory pointer. Here is a typical call of
950 \fBpcre_fullinfo()\fP, to obtain the length of the compiled pattern:
951 .sp
952 int rc;
953 size_t length;
954 rc = pcre_fullinfo(
955 re, /* result of pcre_compile() */
956 pe, /* result of pcre_study(), or NULL */
957 PCRE_INFO_SIZE, /* what is required */
958 &length); /* where to put the data */
959 .sp
960 The possible values for the third argument are defined in \fBpcre.h\fP, and are
961 as follows:
962 .sp
963 PCRE_INFO_BACKREFMAX
964 .sp
965 Return the number of the highest back reference in the pattern. The fourth
966 argument should point to an \fBint\fP variable. Zero is returned if there are
967 no back references.
968 .sp
969 PCRE_INFO_CAPTURECOUNT
970 .sp
971 Return the number of capturing subpatterns in the pattern. The fourth argument
972 should point to an \fBint\fP variable.
973 .sp
974 PCRE_INFO_DEFAULT_TABLES
975 .sp
976 Return a pointer to the internal default character tables within PCRE. The
977 fourth argument should point to an \fBunsigned char *\fP variable. This
978 information call is provided for internal use by the \fBpcre_study()\fP
979 function. External callers can cause PCRE to use its internal tables by passing
980 a NULL table pointer.
981 .sp
982 PCRE_INFO_FIRSTBYTE
983 .sp
984 Return information about the first byte of any matched string, for a
985 non-anchored pattern. The fourth argument should point to an \fBint\fP
986 variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
987 still recognized for backwards compatibility.)
988 .P
989 If there is a fixed first byte, for example, from a pattern such as
990 (cat|cow|coyote), its value is returned. Otherwise, if either
991 .sp
992 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
993 starts with "^", or
994 .sp
995 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
996 (if it were set, the pattern would be anchored),
997 .sp
998 -1 is returned, indicating that the pattern matches only at the start of a
999 subject string or after any newline within the string. Otherwise -2 is
1000 returned. For anchored patterns, -2 is returned.
1001 .sp
1002 PCRE_INFO_FIRSTTABLE
1003 .sp
1004 If the pattern was studied, and this resulted in the construction of a 256-bit
1005 table indicating a fixed set of bytes for the first byte in any matching
1006 string, a pointer to the table is returned. Otherwise NULL is returned. The
1007 fourth argument should point to an \fBunsigned char *\fP variable.
1008 .sp
1009 PCRE_INFO_HASCRORLF
1010 .sp
1011 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1012 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1013 explicit match is either a literal CR or LF character, or \er or \en.
1014 .sp
1015 PCRE_INFO_JCHANGED
1016 .sp
1017 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1018 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1019 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1020 .sp
1021 PCRE_INFO_LASTLITERAL
1022 .sp
1023 Return the value of the rightmost literal byte that must exist in any matched
1024 string, other than at its start, if such a byte has been recorded. The fourth
1025 argument should point to an \fBint\fP variable. If there is no such byte, -1 is
1026 returned. For anchored patterns, a last literal byte is recorded only if it
1027 follows something of variable length. For example, for the pattern
1028 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1029 is -1.
1030 .sp
1031 PCRE_INFO_MINLENGTH
1032 .sp
1033 If the pattern was studied and a minimum length for matching subject strings
1034 was computed, its value is returned. Otherwise the returned value is -1. The
1035 value is a number of characters, not bytes (this may be relevant in UTF-8
1036 mode). The fourth argument should point to an \fBint\fP variable. A
1037 non-negative value is a lower bound to the length of any matching string. There
1038 may not be any strings of that length that do actually match, but every string
1039 that does match is at least that long.
1040 .sp
1041 PCRE_INFO_NAMECOUNT
1042 PCRE_INFO_NAMEENTRYSIZE
1043 PCRE_INFO_NAMETABLE
1044 .sp
1045 PCRE supports the use of named as well as numbered capturing parentheses. The
1046 names are just an additional way of identifying the parentheses, which still
1047 acquire numbers. Several convenience functions such as
1048 \fBpcre_get_named_substring()\fP are provided for extracting captured
1049 substrings by name. It is also possible to extract the data directly, by first
1050 converting the name to a number in order to access the correct pointers in the
1051 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1052 you need to use the name-to-number map, which is described by these three
1053 values.
1054 .P
1055 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1056 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1057 entry; both of these return an \fBint\fP value. The entry size depends on the
1058 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1059 entry of the table (a pointer to \fBchar\fP). The first two bytes of each entry
1060 are the number of the capturing parenthesis, most significant byte first. The
1061 rest of the entry is the corresponding name, zero terminated.
1062 .P
1063 The names are in alphabetical order. Duplicate names may appear if (?| is used
1064 to create multiple groups with the same number, as described in the
1065 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1066 .\" </a>
1067 section on duplicate subpattern numbers
1068 .\"
1069 in the
1070 .\" HREF
1071 \fBpcrepattern\fP
1072 .\"
1073 page. Duplicate names for subpatterns with different numbers are permitted only
1074 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1075 table in the order in which they were found in the pattern. In the absence of
1076 (?| this is the order of increasing number; when (?| is used this is not
1077 necessarily the case because later subpatterns may have lower numbers.
1078 .P
1079 As a simple example of the name/number table, consider the following pattern
1080 (assume PCRE_EXTENDED is set, so white space - including newlines - is
1081 ignored):
1082 .sp
1083 .\" JOIN
1084 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1085 (?<month>\ed\ed) - (?<day>\ed\ed) )
1086 .sp
1087 There are four named subpatterns, so the table has four entries, and each entry
1088 in the table is eight bytes long. The table is as follows, with non-printing
1089 bytes shows in hexadecimal, and undefined bytes shown as ??:
1090 .sp
1091 00 01 d a t e 00 ??
1092 00 05 d a y 00 ?? ??
1093 00 04 m o n t h 00
1094 00 02 y e a r 00 ??
1095 .sp
1096 When writing code to extract data from named subpatterns using the
1097 name-to-number map, remember that the length of the entries is likely to be
1098 different for each compiled pattern.
1099 .sp
1100 PCRE_INFO_OKPARTIAL
1101 .sp
1102 Return 1 if the pattern can be used for partial matching with
1103 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1104 \fBint\fP variable. From release 8.00, this always returns 1, because the
1105 restrictions that previously applied to partial matching have been lifted. The
1106 .\" HREF
1107 \fBpcrepartial\fP
1108 .\"
1109 documentation gives details of partial matching.
1110 .sp
1111 PCRE_INFO_OPTIONS
1112 .sp
1113 Return a copy of the options with which the pattern was compiled. The fourth
1114 argument should point to an \fBunsigned long int\fP variable. These option bits
1115 are those specified in the call to \fBpcre_compile()\fP, modified by any
1116 top-level option settings at the start of the pattern itself. In other words,
1117 they are the options that will be in force when matching starts. For example,
1118 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1119 result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1120 .P
1121 A pattern is automatically anchored by PCRE if all of its top-level
1122 alternatives begin with one of the following:
1123 .sp
1124 ^ unless PCRE_MULTILINE is set
1125 \eA always
1126 \eG always
1127 .\" JOIN
1128 .* if PCRE_DOTALL is set and there are no back
1129 references to the subpattern in which .* appears
1130 .sp
1131 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1132 \fBpcre_fullinfo()\fP.
1133 .sp
1134 PCRE_INFO_SIZE
1135 .sp
1136 Return the size of the compiled pattern, that is, the value that was passed as
1137 the argument to \fBpcre_malloc()\fP when PCRE was getting memory in which to
1138 place the compiled data. The fourth argument should point to a \fBsize_t\fP
1139 variable.
1140 .sp
1141 PCRE_INFO_STUDYSIZE
1142 .sp
1143 Return the size of the data block pointed to by the \fIstudy_data\fP field in
1144 a \fBpcre_extra\fP block. That is, it is the value that was passed to
1145 \fBpcre_malloc()\fP when PCRE was getting memory into which to place the data
1146 created by \fBpcre_study()\fP. If \fBpcre_extra\fP is NULL, or there is no
1147 study data, zero is returned. The fourth argument should point to a
1148 \fBsize_t\fP variable.
1149 .
1150 .
1151 .SH "OBSOLETE INFO FUNCTION"
1152 .rs
1153 .sp
1154 .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
1155 .B *\fIfirstcharptr\fP);
1156 .PP
1157 The \fBpcre_info()\fP function is now obsolete because its interface is too
1158 restrictive to return all the available data about a compiled pattern. New
1159 programs should use \fBpcre_fullinfo()\fP instead. The yield of
1160 \fBpcre_info()\fP is the number of capturing subpatterns, or one of the
1161 following negative numbers:
1162 .sp
1163 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1164 PCRE_ERROR_BADMAGIC the "magic number" was not found
1165 .sp
1166 If the \fIoptptr\fP argument is not NULL, a copy of the options with which the
1167 pattern was compiled is placed in the integer it points to (see
1168 PCRE_INFO_OPTIONS above).
1169 .P
1170 If the pattern is not anchored and the \fIfirstcharptr\fP argument is not NULL,
1171 it is used to pass back information about the first character of any matched
1172 string (see PCRE_INFO_FIRSTBYTE above).
1173 .
1174 .
1175 .SH "REFERENCE COUNTS"
1176 .rs
1177 .sp
1178 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1179 .PP
1180 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1181 data block that contains a compiled pattern. It is provided for the benefit of
1182 applications that operate in an object-oriented manner, where different parts
1183 of the application may be using the same compiled pattern, but you want to free
1184 the block when they are all done.
1185 .P
1186 When a pattern is compiled, the reference count field is initialized to zero.
1187 It is changed only by calling this function, whose action is to add the
1188 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1189 function is the new value. However, the value of the count is constrained to
1190 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1191 it is forced to the appropriate limit value.
1192 .P
1193 Except when it is zero, the reference count is not correctly preserved if a
1194 pattern is compiled on one host and then transferred to a host whose byte-order
1195 is different. (This seems a highly unlikely scenario.)
1196 .
1197 .
1198 .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1199 .rs
1200 .sp
1201 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1202 .ti +5n
1203 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1204 .ti +5n
1205 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1206 .P
1207 The function \fBpcre_exec()\fP is called to match a subject string against a
1208 compiled pattern, which is passed in the \fIcode\fP argument. If the
1209 pattern was studied, the result of the study should be passed in the
1210 \fIextra\fP argument. This function is the main matching facility of the
1211 library, and it operates in a Perl-like manner. For specialist use there is
1212 also an alternative matching function, which is described
1213 .\" HTML <a href="#dfamatch">
1214 .\" </a>
1215 below
1216 .\"
1217 in the section about the \fBpcre_dfa_exec()\fP function.
1218 .P
1219 In most applications, the pattern will have been compiled (and optionally
1220 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1221 possible to save compiled patterns and study data, and then use them later
1222 in different processes, possibly even on different hosts. For a discussion
1223 about this, see the
1224 .\" HREF
1225 \fBpcreprecompile\fP
1226 .\"
1227 documentation.
1228 .P
1229 Here is an example of a simple call to \fBpcre_exec()\fP:
1230 .sp
1231 int rc;
1232 int ovector[30];
1233 rc = pcre_exec(
1234 re, /* result of pcre_compile() */
1235 NULL, /* we didn't study the pattern */
1236 "some string", /* the subject string */
1237 11, /* the length of the subject string */
1238 0, /* start at offset 0 in the subject */
1239 0, /* default options */
1240 ovector, /* vector of integers for substring information */
1241 30); /* number of elements (NOT size in bytes) */
1242 .
1243 .\" HTML <a name="extradata"></a>
1244 .SS "Extra data for \fBpcre_exec()\fR"
1245 .rs
1246 .sp
1247 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1248 data block. The \fBpcre_study()\fP function returns such a block (when it
1249 doesn't return NULL), but you can also create one for yourself, and pass
1250 additional information in it. The \fBpcre_extra\fP block contains the following
1251 fields (not necessarily in this order):
1252 .sp
1253 unsigned long int \fIflags\fP;
1254 void *\fIstudy_data\fP;
1255 unsigned long int \fImatch_limit\fP;
1256 unsigned long int \fImatch_limit_recursion\fP;
1257 void *\fIcallout_data\fP;
1258 const unsigned char *\fItables\fP;
1259 unsigned char **\fImark\fP;
1260 .sp
1261 The \fIflags\fP field is a bitmap that specifies which of the other fields
1262 are set. The flag bits are:
1263 .sp
1264 PCRE_EXTRA_STUDY_DATA
1265 PCRE_EXTRA_MATCH_LIMIT
1266 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1267 PCRE_EXTRA_CALLOUT_DATA
1268 PCRE_EXTRA_TABLES
1269 PCRE_EXTRA_MARK
1270 .sp
1271 Other flag bits should be set to zero. The \fIstudy_data\fP field is set in the
1272 \fBpcre_extra\fP block that is returned by \fBpcre_study()\fP, together with
1273 the appropriate flag bit. You should not set this yourself, but you may add to
1274 the block by setting the other fields and their corresponding flag bits.
1275 .P
1276 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1277 vast amount of resources when running patterns that are not going to match,
1278 but which have a very large number of possibilities in their search trees. The
1279 classic example is a pattern that uses nested unlimited repeats.
1280 .P
1281 Internally, PCRE uses a function called \fBmatch()\fP which it calls repeatedly
1282 (sometimes recursively). The limit set by \fImatch_limit\fP is imposed on the
1283 number of times this function is called during a match, which has the effect of
1284 limiting the amount of backtracking that can take place. For patterns that are
1285 not anchored, the count restarts from zero for each position in the subject
1286 string.
1287 .P
1288 The default value for the limit can be set when PCRE is built; the default
1289 default is 10 million, which handles all but the most extreme cases. You can
1290 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1291 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1292 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1293 PCRE_ERROR_MATCHLIMIT.
1294 .P
1295 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1296 instead of limiting the total number of times that \fBmatch()\fP is called, it
1297 limits the depth of recursion. The recursion depth is a smaller number than the
1298 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1299 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1300 .P
1301 Limiting the recursion depth limits the amount of stack that can be used, or,
1302 when PCRE has been compiled to use memory on the heap instead of the stack, the
1303 amount of heap memory that can be used.
1304 .P
1305 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1306 built; the default default is the same value as the default for
1307 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1308 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1309 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1310 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1311 .P
1312 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1313 and is described in the
1314 .\" HREF
1315 \fBpcrecallout\fP
1316 .\"
1317 documentation.
1318 .P
1319 The \fItables\fP field is used to pass a character tables pointer to
1320 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1321 pattern. A non-NULL value is stored with the compiled pattern only if custom
1322 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1323 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1324 internal tables to be used. This facility is helpful when re-using patterns
1325 that have been saved after compiling with an external set of tables, because
1326 the external tables might be at a different address when \fBpcre_exec()\fP is
1327 called. See the
1328 .\" HREF
1329 \fBpcreprecompile\fP
1330 .\"
1331 documentation for a discussion of saving compiled patterns for later use.
1332 .P
1333 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1334 be set to point to a \fBchar *\fP variable. If the pattern contains any
1335 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1336 a name to pass back, a pointer to the name string (zero terminated) is placed
1337 in the variable pointed to by the \fImark\fP field. The names are within the
1338 compiled pattern; if you wish to retain such a name you must copy it before
1339 freeing the memory of a compiled pattern. If there is no name to pass back, the
1340 variable pointed to by the \fImark\fP field set to NULL. For details of the
1341 backtracking control verbs, see the section entitled
1342 .\" HTML <a href="pcrepattern#backtrackcontrol">
1343 .\" </a>
1344 "Backtracking control"
1345 .\"
1346 in the
1347 .\" HREF
1348 \fBpcrepattern\fP
1349 .\"
1350 documentation.
1351 .
1352 .
1353 .\" HTML <a name="execoptions"></a>
1354 .SS "Option bits for \fBpcre_exec()\fP"
1355 .rs
1356 .sp
1357 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1358 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1359 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1360 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_SOFT, and
1361 PCRE_PARTIAL_HARD.
1362 .sp
1363 PCRE_ANCHORED
1364 .sp
1365 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1366 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1367 to be anchored by virtue of its contents, it cannot be made unachored at
1368 matching time.
1369 .sp
1370 PCRE_BSR_ANYCRLF
1371 PCRE_BSR_UNICODE
1372 .sp
1373 These options (which are mutually exclusive) control what the \eR escape
1374 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1375 match any Unicode newline sequence. These options override the choice that was
1376 made or defaulted when the pattern was compiled.
1377 .sp
1378 PCRE_NEWLINE_CR
1379 PCRE_NEWLINE_LF
1380 PCRE_NEWLINE_CRLF
1381 PCRE_NEWLINE_ANYCRLF
1382 PCRE_NEWLINE_ANY
1383 .sp
1384 These options override the newline definition that was chosen or defaulted when
1385 the pattern was compiled. For details, see the description of
1386 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1387 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1388 the way the match position is advanced after a match failure for an unanchored
1389 pattern.
1390 .P
1391 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1392 match attempt for an unanchored pattern fails when the current position is at a
1393 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1394 characters, the match position is advanced by two characters instead of one, in
1395 other words, to after the CRLF.
1396 .P
1397 The above rule is a compromise that makes the most common cases work as
1398 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1399 set), it does not match the string "\er\enA" because, after failing at the
1400 start, it skips both the CR and the LF before retrying. However, the pattern
1401 [\er\en]A does match that string, because it contains an explicit CR or LF
1402 reference, and so advances only by one character after the first failure.
1403 .P
1404 An explicit match for CR of LF is either a literal appearance of one of those
1405 characters, or one of the \er or \en escape sequences. Implicit matches such as
1406 [^X] do not count, nor does \es (which includes CR and LF in the characters
1407 that it matches).
1408 .P
1409 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1410 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1411 .sp
1412 PCRE_NOTBOL
1413 .sp
1414 This option specifies that first character of the subject string is not the
1415 beginning of a line, so the circumflex metacharacter should not match before
1416 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1417 never to match. This option affects only the behaviour of the circumflex
1418 metacharacter. It does not affect \eA.
1419 .sp
1420 PCRE_NOTEOL
1421 .sp
1422 This option specifies that the end of the subject string is not the end of a
1423 line, so the dollar metacharacter should not match it nor (except in multiline
1424 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1425 compile time) causes dollar never to match. This option affects only the
1426 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1427 .sp
1428 PCRE_NOTEMPTY
1429 .sp
1430 An empty string is not considered to be a valid match if this option is set. If
1431 there are alternatives in the pattern, they are tried. If all the alternatives
1432 match the empty string, the entire match fails. For example, if the pattern
1433 .sp
1434 a?b?
1435 .sp
1436 is applied to a string not beginning with "a" or "b", it matches an empty
1437 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1438 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1439 .sp
1440 PCRE_NOTEMPTY_ATSTART
1441 .sp
1442 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1443 the start of the subject is permitted. If the pattern is anchored, such a match
1444 can occur only if the pattern contains \eK.
1445 .P
1446 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1447 does make a special case of a pattern match of the empty string within its
1448 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1449 emulate Perl's behaviour after matching a null string by first trying the match
1450 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1451 if that fails, by advancing the starting offset (see below) and trying an
1452 ordinary match again. There is some code that demonstrates how to do this in
1453 the
1454 .\" HREF
1455 \fBpcredemo\fP
1456 .\"
1457 sample program. In the most general case, you have to check to see if the
1458 newline convention recognizes CRLF as a newline, and if so, and the current
1459 character is CR followed by LF, advance the starting offset by two characters
1460 instead of one.
1461 .sp
1462 PCRE_NO_START_OPTIMIZE
1463 .sp
1464 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1465 a match, in order to speed up the process. For example, if it is known that an
1466 unanchored match must start with a specific character, it searches the subject
1467 for that character, and fails immediately if it cannot find it, without
1468 actually running the main matching function. This means that a special item
1469 such as (*COMMIT) at the start of a pattern is not considered until after a
1470 suitable starting point for the match has been found. When callouts or (*MARK)
1471 items are in use, these "start-up" optimizations can cause them to be skipped
1472 if the pattern is never actually used. The start-up optimizations are in effect
1473 a pre-scan of the subject that takes place before the pattern is run.
1474 .P
1475 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1476 causing performance to suffer, but ensuring that in cases where the result is
1477 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1478 are considered at every possible starting position in the subject string.
1479 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1480 Consider the pattern
1481 .sp
1482 (*COMMIT)ABC
1483 .sp
1484 When this is compiled, PCRE records the fact that a match must start with the
1485 character "A". Suppose the subject string is "DEFABC". The start-up
1486 optimization scans along the subject, finds "A" and runs the first match
1487 attempt from there. The (*COMMIT) item means that the pattern must match the
1488 current starting position, which in this case, it does. However, if the same
1489 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1490 subject string does not happen. The first match attempt is run starting from
1491 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1492 the overall result is "no match". If the pattern is studied, more start-up
1493 optimizations may be used. For example, a minimum length for the subject may be
1494 recorded. Consider the pattern
1495 .sp
1496 (*MARK:A)(X|Y)
1497 .sp
1498 The minimum length for a match is one character. If the subject is "ABC", there
1499 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1500 If the pattern is studied, the final attempt does not take place, because PCRE
1501 knows that the subject is too short, and so the (*MARK) is never encountered.
1502 In this case, studying the pattern does not affect the overall match result,
1503 which is still "no match", but it does affect the auxiliary information that is
1504 returned.
1505 .sp
1506 PCRE_NO_UTF8_CHECK
1507 .sp
1508 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1509 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1510 The value of \fIstartoffset\fP is also checked to ensure that it points to the
1511 start of a UTF-8 character. There is a discussion about the validity of UTF-8
1512 strings in the
1513 .\" HTML <a href="pcre.html#utf8strings">
1514 .\" </a>
1515 section on UTF-8 support
1516 .\"
1517 in the main
1518 .\" HREF
1519 \fBpcre\fP
1520 .\"
1521 page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_exec()\fP returns
1522 the error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is
1523 a truncated UTF-8 character at the end of the subject, PCRE_ERROR_SHORTUTF8. If
1524 \fIstartoffset\fP contains a value that does not point to the start of a UTF-8
1525 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1526 returned.
1527 .P
1528 If you already know that your subject is valid, and you want to skip these
1529 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1530 calling \fBpcre_exec()\fP. You might want to do this for the second and
1531 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1532 all the matches in a single subject string. However, you should be sure that
1533 the value of \fIstartoffset\fP points to the start of a UTF-8 character (or the
1534 end of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1535 invalid UTF-8 string as a subject or an invalid value of \fIstartoffset\fP is
1536 undefined. Your program may crash.
1537 .sp
1538 PCRE_PARTIAL_HARD
1539 PCRE_PARTIAL_SOFT
1540 .sp
1541 These options turn on the partial matching feature. For backwards
1542 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1543 occurs if the end of the subject string is reached successfully, but there are
1544 not enough subject characters to complete the match. If this happens when
1545 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1546 testing any remaining alternatives. Only if no complete match can be found is
1547 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1548 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1549 but only if no complete match can be found.
1550 .P
1551 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1552 partial match is found, \fBpcre_exec()\fP immediately returns
1553 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1554 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1555 important that an alternative complete match.
1556 .P
1557 In both cases, the portion of the string that was inspected when the partial
1558 match was found is set as the first matching string. There is a more detailed
1559 discussion of partial and multi-segment matching, with examples, in the
1560 .\" HREF
1561 \fBpcrepartial\fP
1562 .\"
1563 documentation.
1564 .
1565 .
1566 .SS "The string to be matched by \fBpcre_exec()\fP"
1567 .rs
1568 .sp
1569 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1570 \fIsubject\fP, a length (in bytes) in \fIlength\fP, and a starting byte offset
1571 in \fIstartoffset\fP. If this is negative or greater than the length of the
1572 subject, \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET.
1573 .P
1574 In UTF-8 mode, the byte offset must point to the start of a UTF-8 character (or
1575 the end of the subject). Unlike the pattern string, the subject may contain
1576 binary zero bytes. When the starting offset is zero, the search for a match
1577 starts at the beginning of the subject, and this is by far the most common
1578 case.
1579 .P
1580 A non-zero starting offset is useful when searching for another match in the
1581 same subject by calling \fBpcre_exec()\fP again after a previous success.
1582 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1583 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1584 lookbehind. For example, consider the pattern
1585 .sp
1586 \eBiss\eB
1587 .sp
1588 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1589 the current position in the subject is not a word boundary.) When applied to
1590 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1591 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1592 subject, namely "issipi", it does not match, because \eB is always false at the
1593 start of the subject, which is deemed to be a word boundary. However, if
1594 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1595 set to 4, it finds the second occurrence of "iss" because it is able to look
1596 behind the starting point to discover that it is preceded by a letter.
1597 .P
1598 Finding all the matches in a subject is tricky when the pattern can match an
1599 empty string. It is possible to emulate Perl's /g behaviour by first trying the
1600 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1601 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1602 and trying an ordinary match again. There is some code that demonstrates how to
1603 do this in the
1604 .\" HREF
1605 \fBpcredemo\fP
1606 .\"
1607 sample program. In the most general case, you have to check to see if the
1608 newline convention recognizes CRLF as a newline, and if so, and the current
1609 character is CR followed by LF, advance the starting offset by two characters
1610 instead of one.
1611 .P
1612 If a non-zero starting offset is passed when the pattern is anchored, one
1613 attempt to match at the given offset is made. This can only succeed if the
1614 pattern does not require the match to be at the start of the subject.
1615 .
1616 .
1617 .SS "How \fBpcre_exec()\fP returns captured substrings"
1618 .rs
1619 .sp
1620 In general, a pattern matches a certain portion of the subject, and in
1621 addition, further substrings from the subject may be picked out by parts of the
1622 pattern. Following the usage in Jeffrey Friedl's book, this is called
1623 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1624 a fragment of a pattern that picks out a substring. PCRE supports several other
1625 kinds of parenthesized subpattern that do not cause substrings to be captured.
1626 .P
1627 Captured substrings are returned to the caller via a vector of integers whose
1628 address is passed in \fIovector\fP. The number of elements in the vector is
1629 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
1630 argument is NOT the size of \fIovector\fP in bytes.
1631 .P
1632 The first two-thirds of the vector is used to pass back captured substrings,
1633 each substring using a pair of integers. The remaining third of the vector is
1634 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1635 and is not available for passing back information. The number passed in
1636 \fIovecsize\fP should always be a multiple of three. If it is not, it is
1637 rounded down.
1638 .P
1639 When a match is successful, information about captured substrings is returned
1640 in pairs of integers, starting at the beginning of \fIovector\fP, and
1641 continuing up to two-thirds of its length at the most. The first element of
1642 each pair is set to the byte offset of the first character in a substring, and
1643 the second is set to the byte offset of the first character after the end of a
1644 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
1645 mode. They are not character counts.
1646 .P
1647 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
1648 portion of the subject string matched by the entire pattern. The next pair is
1649 used for the first capturing subpattern, and so on. The value returned by
1650 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
1651 For example, if two substrings have been captured, the returned value is 3. If
1652 there are no capturing subpatterns, the return value from a successful match is
1653 1, indicating that just the first pair of offsets has been set.
1654 .P
1655 If a capturing subpattern is matched repeatedly, it is the last portion of the
1656 string that it matched that is returned.
1657 .P
1658 If the vector is too small to hold all the captured substring offsets, it is
1659 used as far as possible (up to two-thirds of its length), and the function
1660 returns a value of zero. If the substring offsets are not of interest,
1661 \fBpcre_exec()\fP may be called with \fIovector\fP passed as NULL and
1662 \fIovecsize\fP as zero. However, if the pattern contains back references and
1663 the \fIovector\fP is not big enough to remember the related substrings, PCRE
1664 has to get additional memory for use during matching. Thus it is usually
1665 advisable to supply an \fIovector\fP.
1666 .P
1667 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
1668 subpatterns there are in a compiled pattern. The smallest size for
1669 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
1670 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
1671 .P
1672 It is possible for capturing subpattern number \fIn+1\fP to match some part of
1673 the subject when subpattern \fIn\fP has not been used at all. For example, if
1674 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1675 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1676 happens, both values in the offset pairs corresponding to unused subpatterns
1677 are set to -1.
1678 .P
1679 Offset values that correspond to unused subpatterns at the end of the
1680 expression are also set to -1. For example, if the string "abc" is matched
1681 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1682 return from the function is 2, because the highest used capturing subpattern
1683 number is 1, and the offsets for for the second and third capturing subpatterns
1684 (assuming the vector is large enough, of course) are set to -1.
1685 .P
1686 \fBNote\fP: Elements of \fIovector\fP that do not correspond to capturing
1687 parentheses in the pattern are never changed. That is, if a pattern contains
1688 \fIn\fP capturing parentheses, no more than \fIovector[0]\fP to
1689 \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other elements retain
1690 whatever values they previously had.
1691 .P
1692 Some convenience functions are provided for extracting the captured substrings
1693 as separate strings. These are described below.
1694 .
1695 .\" HTML <a name="errorlist"></a>
1696 .SS "Error return values from \fBpcre_exec()\fP"
1697 .rs
1698 .sp
1699 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1700 defined in the header file:
1701 .sp
1702 PCRE_ERROR_NOMATCH (-1)
1703 .sp
1704 The subject string did not match the pattern.
1705 .sp
1706 PCRE_ERROR_NULL (-2)
1707 .sp
1708 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
1709 NULL and \fIovecsize\fP was not zero.
1710 .sp
1711 PCRE_ERROR_BADOPTION (-3)
1712 .sp
1713 An unrecognized bit was set in the \fIoptions\fP argument.
1714 .sp
1715 PCRE_ERROR_BADMAGIC (-4)
1716 .sp
1717 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1718 the case when it is passed a junk pointer and to detect when a pattern that was
1719 compiled in an environment of one endianness is run in an environment with the
1720 other endianness. This is the error that PCRE gives when the magic number is
1721 not present.
1722 .sp
1723 PCRE_ERROR_UNKNOWN_OPCODE (-5)
1724 .sp
1725 While running the pattern match, an unknown item was encountered in the
1726 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1727 of the compiled pattern.
1728 .sp
1729 PCRE_ERROR_NOMEMORY (-6)
1730 .sp
1731 If a pattern contains back references, but the \fIovector\fP that is passed to
1732 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
1733 gets a block of memory at the start of matching to use for this purpose. If the
1734 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
1735 automatically freed at the end of matching.
1736 .P
1737 This error is also given if \fBpcre_stack_malloc()\fP fails in
1738 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
1739 \fB--disable-stack-for-recursion\fP.
1740 .sp
1741 PCRE_ERROR_NOSUBSTRING (-7)
1742 .sp
1743 This error is used by the \fBpcre_copy_substring()\fP,
1744 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
1745 below). It is never returned by \fBpcre_exec()\fP.
1746 .sp
1747 PCRE_ERROR_MATCHLIMIT (-8)
1748 .sp
1749 The backtracking limit, as specified by the \fImatch_limit\fP field in a
1750 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
1751 above.
1752 .sp
1753 PCRE_ERROR_CALLOUT (-9)
1754 .sp
1755 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
1756 use by callout functions that want to yield a distinctive error code. See the
1757 .\" HREF
1758 \fBpcrecallout\fP
1759 .\"
1760 documentation for details.
1761 .sp
1762 PCRE_ERROR_BADUTF8 (-10)
1763 .sp
1764 A string that contains an invalid UTF-8 byte sequence was passed as a subject.
1765 However, if PCRE_PARTIAL_HARD is set and the problem is a truncated UTF-8
1766 character at the end of the subject, PCRE_ERROR_SHORTUTF8 is used instead.
1767 .sp
1768 PCRE_ERROR_BADUTF8_OFFSET (-11)
1769 .sp
1770 The UTF-8 byte sequence that was passed as a subject was valid, but the value
1771 of \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
1772 end of the subject.
1773 .sp
1774 PCRE_ERROR_PARTIAL (-12)
1775 .sp
1776 The subject string did not match, but it did match partially. See the
1777 .\" HREF
1778 \fBpcrepartial\fP
1779 .\"
1780 documentation for details of partial matching.
1781 .sp
1782 PCRE_ERROR_BADPARTIAL (-13)
1783 .sp
1784 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1785 option was used with a compiled pattern containing items that were not
1786 supported for partial matching. From release 8.00 onwards, there are no
1787 restrictions on partial matching.
1788 .sp
1789 PCRE_ERROR_INTERNAL (-14)
1790 .sp
1791 An unexpected internal error has occurred. This error could be caused by a bug
1792 in PCRE or by overwriting of the compiled pattern.
1793 .sp
1794 PCRE_ERROR_BADCOUNT (-15)
1795 .sp
1796 This error is given if the value of the \fIovecsize\fP argument is negative.
1797 .sp
1798 PCRE_ERROR_RECURSIONLIMIT (-21)
1799 .sp
1800 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
1801 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
1802 description above.
1803 .sp
1804 PCRE_ERROR_BADNEWLINE (-23)
1805 .sp
1806 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
1807 .sp
1808 PCRE_ERROR_BADOFFSET (-24)
1809 .sp
1810 The value of \fIstartoffset\fP was negative or greater than the length of the
1811 subject, that is, the value in \fIlength\fP.
1812 .sp
1813 PCRE_ERROR_SHORTUTF8 (-25)
1814 .sp
1815 The subject string ended with an incomplete (truncated) UTF-8 character, and
1816 the PCRE_PARTIAL_HARD option was set. Without this option, PCRE_ERROR_BADUTF8
1817 is returned in this situation.
1818 .P
1819 Error numbers -16 to -20 and -22 are not used by \fBpcre_exec()\fP.
1820 .
1821 .
1822 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
1823 .rs
1824 .sp
1825 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1826 .ti +5n
1827 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
1828 .ti +5n
1829 .B int \fIbuffersize\fP);
1830 .PP
1831 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1832 .ti +5n
1833 .B int \fIstringcount\fP, int \fIstringnumber\fP,
1834 .ti +5n
1835 .B const char **\fIstringptr\fP);
1836 .PP
1837 .B int pcre_get_substring_list(const char *\fIsubject\fP,
1838 .ti +5n
1839 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
1840 .PP
1841 Captured substrings can be accessed directly by using the offsets returned by
1842 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
1843 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
1844 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
1845 as new, separate, zero-terminated strings. These functions identify substrings
1846 by number. The next section describes functions for extracting named
1847 substrings.
1848 .P
1849 A substring that contains a binary zero is correctly extracted and has a
1850 further zero added on the end, but the result is not, of course, a C string.
1851 However, you can process such a string by referring to the length that is
1852 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
1853 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
1854 for handling strings containing binary zeros, because the end of the final
1855 string is not independently indicated.
1856 .P
1857 The first three arguments are the same for all three of these functions:
1858 \fIsubject\fP is the subject string that has just been successfully matched,
1859 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
1860 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
1861 captured by the match, including the substring that matched the entire regular
1862 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
1863 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
1864 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
1865 number of elements in the vector divided by three.
1866 .P
1867 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
1868 extract a single substring, whose number is given as \fIstringnumber\fP. A
1869 value of zero extracts the substring that matched the entire pattern, whereas
1870 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
1871 the string is placed in \fIbuffer\fP, whose length is given by
1872 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
1873 obtained via \fBpcre_malloc\fP, and its address is returned via
1874 \fIstringptr\fP. The yield of the function is the length of the string, not
1875 including the terminating zero, or one of these error codes:
1876 .sp
1877 PCRE_ERROR_NOMEMORY (-6)
1878 .sp
1879 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
1880 memory failed for \fBpcre_get_substring()\fP.
1881 .sp
1882 PCRE_ERROR_NOSUBSTRING (-7)
1883 .sp
1884 There is no substring whose number is \fIstringnumber\fP.
1885 .P
1886 The \fBpcre_get_substring_list()\fP function extracts all available substrings
1887 and builds a list of pointers to them. All this is done in a single block of
1888 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
1889 is returned via \fIlistptr\fP, which is also the start of the list of string
1890 pointers. The end of the list is marked by a NULL pointer. The yield of the
1891 function is zero if all went well, or the error code
1892 .sp
1893 PCRE_ERROR_NOMEMORY (-6)
1894 .sp
1895 if the attempt to get the memory block failed.
1896 .P
1897 When any of these functions encounter a substring that is unset, which can
1898 happen when capturing subpattern number \fIn+1\fP matches some part of the
1899 subject, but subpattern \fIn\fP has not been used at all, they return an empty
1900 string. This can be distinguished from a genuine zero-length substring by
1901 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
1902 substrings.
1903 .P
1904 The two convenience functions \fBpcre_free_substring()\fP and
1905 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
1906 a previous call of \fBpcre_get_substring()\fP or
1907 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
1908 the function pointed to by \fBpcre_free\fP, which of course could be called
1909 directly from a C program. However, PCRE is used in some situations where it is
1910 linked via a special interface to another programming language that cannot use
1911 \fBpcre_free\fP directly; it is for these cases that the functions are
1912 provided.
1913 .
1914 .
1915 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
1916 .rs
1917 .sp
1918 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
1919 .ti +5n
1920 .B const char *\fIname\fP);
1921 .PP
1922 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
1923 .ti +5n
1924 .B const char *\fIsubject\fP, int *\fIovector\fP,
1925 .ti +5n
1926 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1927 .ti +5n
1928 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
1929 .PP
1930 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
1931 .ti +5n
1932 .B const char *\fIsubject\fP, int *\fIovector\fP,
1933 .ti +5n
1934 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1935 .ti +5n
1936 .B const char **\fIstringptr\fP);
1937 .PP
1938 To extract a substring by name, you first have to find associated number.
1939 For example, for this pattern
1940 .sp
1941 (a+)b(?<xxx>\ed+)...
1942 .sp
1943 the number of the subpattern called "xxx" is 2. If the name is known to be
1944 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
1945 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
1946 pattern, and the second is the name. The yield of the function is the
1947 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
1948 that name.
1949 .P
1950 Given the number, you can extract the substring directly, or use one of the
1951 functions described in the previous section. For convenience, there are also
1952 two functions that do the whole job.
1953 .P
1954 Most of the arguments of \fBpcre_copy_named_substring()\fP and
1955 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
1956 functions that extract by number. As these are described in the previous
1957 section, they are not re-described here. There are just two differences:
1958 .P
1959 First, instead of a substring number, a substring name is given. Second, there
1960 is an extra argument, given at the start, which is a pointer to the compiled
1961 pattern. This is needed in order to gain access to the name-to-number
1962 translation table.
1963 .P
1964 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
1965 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
1966 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
1967 the behaviour may not be what you want (see the next section).
1968 .P
1969 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
1970 subpatterns with the same number, as described in the
1971 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1972 .\" </a>
1973 section on duplicate subpattern numbers
1974 .\"
1975 in the
1976 .\" HREF
1977 \fBpcrepattern\fP
1978 .\"
1979 page, you cannot use names to distinguish the different subpatterns, because
1980 names are not included in the compiled code. The matching process uses only
1981 numbers. For this reason, the use of different names for subpatterns of the
1982 same number causes an error at compile time.
1983 .
1984 .SH "DUPLICATE SUBPATTERN NAMES"
1985 .rs
1986 .sp
1987 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
1988 .ti +5n
1989 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
1990 .PP
1991 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
1992 are not required to be unique. (Duplicate names are always allowed for
1993 subpatterns with the same number, created by using the (?| feature. Indeed, if
1994 such subpatterns are named, they are required to use the same names.)
1995 .P
1996 Normally, patterns with duplicate names are such that in any one match, only
1997 one of the named subpatterns participates. An example is shown in the
1998 .\" HREF
1999 \fBpcrepattern\fP
2000 .\"
2001 documentation.
2002 .P
2003 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2004 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2005 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2006 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2007 returns one of the numbers that are associated with the name, but it is not
2008 defined which it is.
2009 .P
2010 If you want to get full details of all captured substrings for a given name,
2011 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2012 argument is the compiled pattern, and the second is the name. The third and
2013 fourth are pointers to variables which are updated by the function. After it
2014 has run, they point to the first and last entries in the name-to-number table
2015 for the given name. The function itself returns the length of each entry, or
2016 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2017 described above in the section entitled \fIInformation about a pattern\fP.
2018 Given all the relevant entries for the name, you can extract each of their
2019 numbers, and hence the captured data, if any.
2020 .
2021 .
2022 .SH "FINDING ALL POSSIBLE MATCHES"
2023 .rs
2024 .sp
2025 The traditional matching function uses a similar algorithm to Perl, which stops
2026 when it finds the first match, starting at a given point in the subject. If you
2027 want to find all possible matches, or the longest possible match, consider
2028 using the alternative matching function (see below) instead. If you cannot use
2029 the alternative function, but still need to find all possible matches, you
2030 can kludge it up by making use of the callout facility, which is described in
2031 the
2032 .\" HREF
2033 \fBpcrecallout\fP
2034 .\"
2035 documentation.
2036 .P
2037 What you have to do is to insert a callout right at the end of the pattern.
2038 When your callout function is called, extract and save the current matched
2039 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2040 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2041 will yield PCRE_ERROR_NOMATCH.
2042 .
2043 .
2044 .\" HTML <a name="dfamatch"></a>
2045 .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2046 .rs
2047 .sp
2048 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2049 .ti +5n
2050 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2051 .ti +5n
2052 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2053 .ti +5n
2054 .B int *\fIworkspace\fP, int \fIwscount\fP);
2055 .P
2056 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2057 a compiled pattern, using a matching algorithm that scans the subject string
2058 just once, and does not backtrack. This has different characteristics to the
2059 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2060 patterns are not supported. Nevertheless, there are times when this kind of
2061 matching can be useful. For a discussion of the two matching algorithms, and a
2062 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2063 .\" HREF
2064 \fBpcrematching\fP
2065 .\"
2066 documentation.
2067 .P
2068 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2069 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2070 different way, and this is described below. The other common arguments are used
2071 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2072 here.
2073 .P
2074 The two additional arguments provide workspace for the function. The workspace
2075 vector should contain at least 20 elements. It is used for keeping track of
2076 multiple paths through the pattern tree. More workspace will be needed for
2077 patterns and subjects where there are a lot of potential matches.
2078 .P
2079 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2080 .sp
2081 int rc;
2082 int ovector[10];
2083 int wspace[20];
2084 rc = pcre_dfa_exec(
2085 re, /* result of pcre_compile() */
2086 NULL, /* we didn't study the pattern */
2087 "some string", /* the subject string */
2088 11, /* the length of the subject string */
2089 0, /* start at offset 0 in the subject */
2090 0, /* default options */
2091 ovector, /* vector of integers for substring information */
2092 10, /* number of elements (NOT size in bytes) */
2093 wspace, /* working space vector */
2094 20); /* number of elements (NOT size in bytes) */
2095 .
2096 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2097 .rs
2098 .sp
2099 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2100 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2101 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2102 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2103 PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2104 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2105 so their description is not repeated here.
2106 .sp
2107 PCRE_PARTIAL_HARD
2108 PCRE_PARTIAL_SOFT
2109 .sp
2110 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2111 details are slightly different. When PCRE_PARTIAL_HARD is set for
2112 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2113 is reached and there is still at least one matching possibility that requires
2114 additional characters. This happens even if some complete matches have also
2115 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2116 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2117 there have been no complete matches, but there is still at least one matching
2118 possibility. The portion of the string that was inspected when the longest
2119 partial match was found is set as the first matching string in both cases.
2120 There is a more detailed discussion of partial and multi-segment matching, with
2121 examples, in the
2122 .\" HREF
2123 \fBpcrepartial\fP
2124 .\"
2125 documentation.
2126 .sp
2127 PCRE_DFA_SHORTEST
2128 .sp
2129 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2130 soon as it has found one match. Because of the way the alternative algorithm
2131 works, this is necessarily the shortest possible match at the first possible
2132 matching point in the subject string.
2133 .sp
2134 PCRE_DFA_RESTART
2135 .sp
2136 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2137 again, with additional subject characters, and have it continue with the same
2138 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2139 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2140 before because data about the match so far is left in them after a partial
2141 match. There is more discussion of this facility in the
2142 .\" HREF
2143 \fBpcrepartial\fP
2144 .\"
2145 documentation.
2146 .
2147 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2148 .rs
2149 .sp
2150 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2151 substring in the subject. Note, however, that all the matches from one run of
2152 the function start at the same point in the subject. The shorter matches are
2153 all initial substrings of the longer matches. For example, if the pattern
2154 .sp
2155 <.*>
2156 .sp
2157 is matched against the string
2158 .sp
2159 This is <something> <something else> <something further> no more
2160 .sp
2161 the three matched strings are
2162 .sp
2163 <something>
2164 <something> <something else>
2165 <something> <something else> <something further>
2166 .sp
2167 On success, the yield of the function is a number greater than zero, which is
2168 the number of matched substrings. The substrings themselves are returned in
2169 \fIovector\fP. Each string uses two elements; the first is the offset to the
2170 start, and the second is the offset to the end. In fact, all the strings have
2171 the same start offset. (Space could have been saved by giving this only once,
2172 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2173 returns data, even though the meaning of the strings is different.)
2174 .P
2175 The strings are returned in reverse order of length; that is, the longest
2176 matching string is given first. If there were too many matches to fit into
2177 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2178 the longest matches.
2179 .
2180 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2181 .rs
2182 .sp
2183 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2184 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2185 described
2186 .\" HTML <a href="#errorlist">
2187 .\" </a>
2188 above.
2189 .\"
2190 There are in addition the following errors that are specific to
2191 \fBpcre_dfa_exec()\fP:
2192 .sp
2193 PCRE_ERROR_DFA_UITEM (-16)
2194 .sp
2195 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2196 that it does not support, for instance, the use of \eC or a back reference.
2197 .sp
2198 PCRE_ERROR_DFA_UCOND (-17)
2199 .sp
2200 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2201 uses a back reference for the condition, or a test for recursion in a specific
2202 group. These are not supported.
2203 .sp
2204 PCRE_ERROR_DFA_UMLIMIT (-18)
2205 .sp
2206 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2207 block that contains a setting of the \fImatch_limit\fP field. This is not
2208 supported (it is meaningless).
2209 .sp
2210 PCRE_ERROR_DFA_WSSIZE (-19)
2211 .sp
2212 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2213 \fIworkspace\fP vector.
2214 .sp
2215 PCRE_ERROR_DFA_RECURSE (-20)
2216 .sp
2217 When a recursive subpattern is processed, the matching function calls itself
2218 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2219 error is given if the output vector is not large enough. This should be
2220 extremely rare, as a vector of size 1000 is used.
2221 .
2222 .
2223 .SH "SEE ALSO"
2224 .rs
2225 .sp
2226 \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
2227 \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
2228 \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
2229 .
2230 .
2231 .SH AUTHOR
2232 .rs
2233 .sp
2234 .nf
2235 Philip Hazel
2236 University Computing Service
2237 Cambridge CB2 3QH, England.
2238 .fi
2239 .
2240 .
2241 .SH REVISION
2242 .rs
2243 .sp
2244 .nf
2245 Last updated: 06 November 2010
2246 Copyright (c) 1997-2010 University of Cambridge.
2247 .fi

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5