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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 572 - (show annotations)
Wed Nov 17 17:55:57 2010 UTC (4 years, 7 months ago) by ph10
File size: 98300 byte(s)
Error occurred while calculating annotation data.
Documentation updates and tidies.
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 metacharacter in the pattern matches a character of
527 any value, including one that indicates a newline. However, it only ever
528 matches one character, even if newlines are coded as CRLF. Without this option,
529 a dot does not match when the current position is at a newline. This option is
530 equivalent to Perl's /s option, and it can be changed within a pattern by a
531 (?s) option setting. A negative class such as [^a] always matches newline
532 characters, independent of the setting of this option.
533 .sp
534 PCRE_DUPNAMES
535 .sp
536 If this bit is set, names used to identify capturing subpatterns need not be
537 unique. This can be helpful for certain types of pattern when it is known that
538 only one instance of the named subpattern can ever be matched. There are more
539 details of named subpatterns below; see also the
540 .\" HREF
541 \fBpcrepattern\fP
542 .\"
543 documentation.
544 .sp
545 PCRE_EXTENDED
546 .sp
547 If this bit is set, whitespace data characters in the pattern are totally
548 ignored except when escaped or inside a character class. Whitespace does not
549 include the VT character (code 11). In addition, characters between an
550 unescaped # outside a character class and the next newline, inclusive, are also
551 ignored. This is equivalent to Perl's /x option, and it can be changed within a
552 pattern by a (?x) option setting.
553 .P
554 Which characters are interpreted as newlines
555 is controlled by the options passed to \fBpcre_compile()\fP or by a special
556 sequence at the start of the pattern, as described in the section entitled
557 .\" HTML <a href="pcrepattern.html#newlines">
558 .\" </a>
559 "Newline conventions"
560 .\"
561 in the \fBpcrepattern\fP documentation. Note that the end of this type of
562 comment is a literal newline sequence in the pattern; escape sequences that
563 happen to represent a newline do not count.
564 .P
565 This option makes it possible to include comments inside complicated patterns.
566 Note, however, that this applies only to data characters. Whitespace characters
567 may never appear within special character sequences in a pattern, for example
568 within the sequence (?( that introduces a conditional subpattern.
569 .sp
570 PCRE_EXTRA
571 .sp
572 This option was invented in order to turn on additional functionality of PCRE
573 that is incompatible with Perl, but it is currently of very little use. When
574 set, any backslash in a pattern that is followed by a letter that has no
575 special meaning causes an error, thus reserving these combinations for future
576 expansion. By default, as in Perl, a backslash followed by a letter with no
577 special meaning is treated as a literal. (Perl can, however, be persuaded to
578 give an error for this, by running it with the -w option.) There are at present
579 no other features controlled by this option. It can also be set by a (?X)
580 option setting within a pattern.
581 .sp
582 PCRE_FIRSTLINE
583 .sp
584 If this option is set, an unanchored pattern is required to match before or at
585 the first newline in the subject string, though the matched text may continue
586 over the newline.
587 .sp
588 PCRE_JAVASCRIPT_COMPAT
589 .sp
590 If this option is set, PCRE's behaviour is changed in some ways so that it is
591 compatible with JavaScript rather than Perl. The changes are as follows:
592 .P
593 (1) A lone closing square bracket in a pattern causes a compile-time error,
594 because this is illegal in JavaScript (by default it is treated as a data
595 character). Thus, the pattern AB]CD becomes illegal when this option is set.
596 .P
597 (2) At run time, a back reference to an unset subpattern group matches an empty
598 string (by default this causes the current matching alternative to fail). A
599 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
600 an "a" in the subject), whereas it fails by default, for Perl compatibility.
601 .sp
602 PCRE_MULTILINE
603 .sp
604 By default, PCRE treats the subject string as consisting of a single line of
605 characters (even if it actually contains newlines). The "start of line"
606 metacharacter (^) matches only at the start of the string, while the "end of
607 line" metacharacter ($) matches only at the end of the string, or before a
608 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
609 Perl.
610 .P
611 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
612 match immediately following or immediately before internal newlines in the
613 subject string, respectively, as well as at the very start and end. This is
614 equivalent to Perl's /m option, and it can be changed within a pattern by a
615 (?m) option setting. If there are no newlines in a subject string, or no
616 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
617 .sp
618 PCRE_NEWLINE_CR
619 PCRE_NEWLINE_LF
620 PCRE_NEWLINE_CRLF
621 PCRE_NEWLINE_ANYCRLF
622 PCRE_NEWLINE_ANY
623 .sp
624 These options override the default newline definition that was chosen when PCRE
625 was built. Setting the first or the second specifies that a newline is
626 indicated by a single character (CR or LF, respectively). Setting
627 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
628 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
629 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
630 that any Unicode newline sequence should be recognized. The Unicode newline
631 sequences are the three just mentioned, plus the single characters VT (vertical
632 tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
633 separator, U+2028), and PS (paragraph separator, U+2029). The last two are
634 recognized only in UTF-8 mode.
635 .P
636 The newline setting in the options word uses three bits that are treated
637 as a number, giving eight possibilities. Currently only six are used (default
638 plus the five values above). This means that if you set more than one newline
639 option, the combination may or may not be sensible. For example,
640 PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
641 other combinations may yield unused numbers and cause an error.
642 .P
643 The only time that a line break in a pattern is specially recognized when
644 compiling is when PCRE_EXTENDED is set. CR and LF are whitespace characters,
645 and so are ignored in this mode. Also, an unescaped # outside a character class
646 indicates a comment that lasts until after the next line break sequence. In
647 other circumstances, line break sequences in patterns are treated as literal
648 data.
649 .P
650 The newline option that is set at compile time becomes the default that is used
651 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
652 .sp
653 PCRE_NO_AUTO_CAPTURE
654 .sp
655 If this option is set, it disables the use of numbered capturing parentheses in
656 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
657 were followed by ?: but named parentheses can still be used for capturing (and
658 they acquire numbers in the usual way). There is no equivalent of this option
659 in Perl.
660 .sp
661 PCRE_UCP
662 .sp
663 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
664 \ew, and some of the POSIX character classes. By default, only ASCII characters
665 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
666 classify characters. More details are given in the section on
667 .\" HTML <a href="pcre.html#genericchartypes">
668 .\" </a>
669 generic character types
670 .\"
671 in the
672 .\" HREF
673 \fBpcrepattern\fP
674 .\"
675 page. If you set PCRE_UCP, matching one of the items it affects takes much
676 longer. The option is available only if PCRE has been compiled with Unicode
677 property support.
678 .sp
679 PCRE_UNGREEDY
680 .sp
681 This option inverts the "greediness" of the quantifiers so that they are not
682 greedy by default, but become greedy if followed by "?". It is not compatible
683 with Perl. It can also be set by a (?U) option setting within the pattern.
684 .sp
685 PCRE_UTF8
686 .sp
687 This option causes PCRE to regard both the pattern and the subject as strings
688 of UTF-8 characters instead of single-byte character strings. However, it is
689 available only when PCRE is built to include UTF-8 support. If not, the use
690 of this option provokes an error. Details of how this option changes the
691 behaviour of PCRE are given in the
692 .\" HTML <a href="pcre.html#utf8support">
693 .\" </a>
694 section on UTF-8 support
695 .\"
696 in the main
697 .\" HREF
698 \fBpcre\fP
699 .\"
700 page.
701 .sp
702 PCRE_NO_UTF8_CHECK
703 .sp
704 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
705 automatically checked. There is a discussion about the
706 .\" HTML <a href="pcre.html#utf8strings">
707 .\" </a>
708 validity of UTF-8 strings
709 .\"
710 in the main
711 .\" HREF
712 \fBpcre\fP
713 .\"
714 page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_compile()\fP
715 returns an error. If you already know that your pattern is valid, and you want
716 to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
717 option. When it is set, the effect of passing an invalid UTF-8 string as a
718 pattern is undefined. It may cause your program to crash. Note that this option
719 can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
720 the UTF-8 validity checking of subject strings.
721 .
722 .
723 .SH "COMPILATION ERROR CODES"
724 .rs
725 .sp
726 The following table lists the error codes than may be returned by
727 \fBpcre_compile2()\fP, along with the error messages that may be returned by
728 both compiling functions. As PCRE has developed, some error codes have fallen
729 out of use. To avoid confusion, they have not been re-used.
730 .sp
731 0 no error
732 1 \e at end of pattern
733 2 \ec at end of pattern
734 3 unrecognized character follows \e
735 4 numbers out of order in {} quantifier
736 5 number too big in {} quantifier
737 6 missing terminating ] for character class
738 7 invalid escape sequence in character class
739 8 range out of order in character class
740 9 nothing to repeat
741 10 [this code is not in use]
742 11 internal error: unexpected repeat
743 12 unrecognized character after (? or (?-
744 13 POSIX named classes are supported only within a class
745 14 missing )
746 15 reference to non-existent subpattern
747 16 erroffset passed as NULL
748 17 unknown option bit(s) set
749 18 missing ) after comment
750 19 [this code is not in use]
751 20 regular expression is too large
752 21 failed to get memory
753 22 unmatched parentheses
754 23 internal error: code overflow
755 24 unrecognized character after (?<
756 25 lookbehind assertion is not fixed length
757 26 malformed number or name after (?(
758 27 conditional group contains more than two branches
759 28 assertion expected after (?(
760 29 (?R or (?[+-]digits must be followed by )
761 30 unknown POSIX class name
762 31 POSIX collating elements are not supported
763 32 this version of PCRE is not compiled with PCRE_UTF8 support
764 33 [this code is not in use]
765 34 character value in \ex{...} sequence is too large
766 35 invalid condition (?(0)
767 36 \eC not allowed in lookbehind assertion
768 37 PCRE does not support \eL, \el, \eN, \eU, or \eu
769 38 number after (?C is > 255
770 39 closing ) for (?C expected
771 40 recursive call could loop indefinitely
772 41 unrecognized character after (?P
773 42 syntax error in subpattern name (missing terminator)
774 43 two named subpatterns have the same name
775 44 invalid UTF-8 string
776 45 support for \eP, \ep, and \eX has not been compiled
777 46 malformed \eP or \ep sequence
778 47 unknown property name after \eP or \ep
779 48 subpattern name is too long (maximum 32 characters)
780 49 too many named subpatterns (maximum 10000)
781 50 [this code is not in use]
782 51 octal value is greater than \e377 (not in UTF-8 mode)
783 52 internal error: overran compiling workspace
784 53 internal error: previously-checked referenced subpattern
785 not found
786 54 DEFINE group contains more than one branch
787 55 repeating a DEFINE group is not allowed
788 56 inconsistent NEWLINE options
789 57 \eg is not followed by a braced, angle-bracketed, or quoted
790 name/number or by a plain number
791 58 a numbered reference must not be zero
792 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
793 60 (*VERB) not recognized
794 61 number is too big
795 62 subpattern name expected
796 63 digit expected after (?+
797 64 ] is an invalid data character in JavaScript compatibility mode
798 65 different names for subpatterns of the same number are
799 not allowed
800 66 (*MARK) must have an argument
801 67 this version of PCRE is not compiled with PCRE_UCP support
802 .sp
803 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
804 be used if the limits were changed when PCRE was built.
805 .
806 .
807 .SH "STUDYING A PATTERN"
808 .rs
809 .sp
810 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
811 .ti +5n
812 .B const char **\fIerrptr\fP);
813 .PP
814 If a compiled pattern is going to be used several times, it is worth spending
815 more time analyzing it in order to speed up the time taken for matching. The
816 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
817 argument. If studying the pattern produces additional information that will
818 help speed up matching, \fBpcre_study()\fP returns a pointer to a
819 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
820 results of the study.
821 .P
822 The returned value from \fBpcre_study()\fP can be passed directly to
823 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
824 also contains other fields that can be set by the caller before the block is
825 passed; these are described
826 .\" HTML <a href="#extradata">
827 .\" </a>
828 below
829 .\"
830 in the section on matching a pattern.
831 .P
832 If studying the pattern does not produce any useful information,
833 \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program
834 wants to pass any of the other fields to \fBpcre_exec()\fP or
835 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block.
836 .P
837 The second argument of \fBpcre_study()\fP contains option bits. At present, no
838 options are defined, and this argument should always be zero.
839 .P
840 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
841 studying succeeds (even if no data is returned), the variable it points to is
842 set to NULL. Otherwise it is set to point to a textual error message. This is a
843 static string that is part of the library. You must not try to free it. You
844 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
845 sure that it has run successfully.
846 .P
847 This is a typical call to \fBpcre_study\fP():
848 .sp
849 pcre_extra *pe;
850 pe = pcre_study(
851 re, /* result of pcre_compile() */
852 0, /* no options exist */
853 &error); /* set to NULL or points to a message */
854 .sp
855 Studying a pattern does two things: first, a lower bound for the length of
856 subject string that is needed to match the pattern is computed. This does not
857 mean that there are any strings of that length that match, but it does
858 guarantee that no shorter strings match. The value is used by
859 \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP to avoid wasting time by trying to
860 match strings that are shorter than the lower bound. You can find out the value
861 in a calling program via the \fBpcre_fullinfo()\fP function.
862 .P
863 Studying a pattern is also useful for non-anchored patterns that do not have a
864 single fixed starting character. A bitmap of possible starting bytes is
865 created. This speeds up finding a position in the subject at which to start
866 matching.
867 .P
868 The two optimizations just described can be disabled by setting the
869 PCRE_NO_START_OPTIMIZE option when calling \fBpcre_exec()\fP or
870 \fBpcre_dfa_exec()\fP. You might want to do this if your pattern contains
871 callouts or (*MARK), and you want to make use of these facilities in cases
872 where matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
873 .\" HTML <a href="#execoptions">
874 .\" </a>
875 below.
876 .\"
877 .
878 .
879 .\" HTML <a name="localesupport"></a>
880 .SH "LOCALE SUPPORT"
881 .rs
882 .sp
883 PCRE handles caseless matching, and determines whether characters are letters,
884 digits, or whatever, by reference to a set of tables, indexed by character
885 value. When running in UTF-8 mode, this applies only to characters with codes
886 less than 128. By default, higher-valued codes never match escapes such as \ew
887 or \ed, but they can be tested with \ep if PCRE is built with Unicode character
888 property support. Alternatively, the PCRE_UCP option can be set at compile
889 time; this causes \ew and friends to use Unicode property support instead of
890 built-in tables. The use of locales with Unicode is discouraged. If you are
891 handling characters with codes greater than 128, you should either use UTF-8
892 and Unicode, or use locales, but not try to mix the two.
893 .P
894 PCRE contains an internal set of tables that are used when the final argument
895 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
896 Normally, the internal tables recognize only ASCII characters. However, when
897 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
898 default "C" locale of the local system, which may cause them to be different.
899 .P
900 The internal tables can always be overridden by tables supplied by the
901 application that calls PCRE. These may be created in a different locale from
902 the default. As more and more applications change to using Unicode, the need
903 for this locale support is expected to die away.
904 .P
905 External tables are built by calling the \fBpcre_maketables()\fP function,
906 which has no arguments, in the relevant locale. The result can then be passed
907 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
908 example, to build and use tables that are appropriate for the French locale
909 (where accented characters with values greater than 128 are treated as letters),
910 the following code could be used:
911 .sp
912 setlocale(LC_CTYPE, "fr_FR");
913 tables = pcre_maketables();
914 re = pcre_compile(..., tables);
915 .sp
916 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
917 are using Windows, the name for the French locale is "french".
918 .P
919 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
920 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
921 that the memory containing the tables remains available for as long as it is
922 needed.
923 .P
924 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
925 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
926 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
927 pattern, compilation, studying and matching all happen in the same locale, but
928 different patterns can be compiled in different locales.
929 .P
930 It is possible to pass a table pointer or NULL (indicating the use of the
931 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
932 this facility could be used to match a pattern in a different locale from the
933 one in which it was compiled. Passing table pointers at run time is discussed
934 below in the section on matching a pattern.
935 .
936 .
937 .SH "INFORMATION ABOUT A PATTERN"
938 .rs
939 .sp
940 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
941 .ti +5n
942 .B int \fIwhat\fP, void *\fIwhere\fP);
943 .PP
944 The \fBpcre_fullinfo()\fP function returns information about a compiled
945 pattern. It replaces the obsolete \fBpcre_info()\fP function, which is
946 nevertheless retained for backwards compability (and is documented below).
947 .P
948 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
949 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
950 the pattern was not studied. The third argument specifies which piece of
951 information is required, and the fourth argument is a pointer to a variable
952 to receive the data. The yield of the function is zero for success, or one of
953 the following negative numbers:
954 .sp
955 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
956 the argument \fIwhere\fP was NULL
957 PCRE_ERROR_BADMAGIC the "magic number" was not found
958 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
959 .sp
960 The "magic number" is placed at the start of each compiled pattern as an simple
961 check against passing an arbitrary memory pointer. Here is a typical call of
962 \fBpcre_fullinfo()\fP, to obtain the length of the compiled pattern:
963 .sp
964 int rc;
965 size_t length;
966 rc = pcre_fullinfo(
967 re, /* result of pcre_compile() */
968 pe, /* result of pcre_study(), or NULL */
969 PCRE_INFO_SIZE, /* what is required */
970 &length); /* where to put the data */
971 .sp
972 The possible values for the third argument are defined in \fBpcre.h\fP, and are
973 as follows:
974 .sp
975 PCRE_INFO_BACKREFMAX
976 .sp
977 Return the number of the highest back reference in the pattern. The fourth
978 argument should point to an \fBint\fP variable. Zero is returned if there are
979 no back references.
980 .sp
981 PCRE_INFO_CAPTURECOUNT
982 .sp
983 Return the number of capturing subpatterns in the pattern. The fourth argument
984 should point to an \fBint\fP variable.
985 .sp
986 PCRE_INFO_DEFAULT_TABLES
987 .sp
988 Return a pointer to the internal default character tables within PCRE. The
989 fourth argument should point to an \fBunsigned char *\fP variable. This
990 information call is provided for internal use by the \fBpcre_study()\fP
991 function. External callers can cause PCRE to use its internal tables by passing
992 a NULL table pointer.
993 .sp
994 PCRE_INFO_FIRSTBYTE
995 .sp
996 Return information about the first byte of any matched string, for a
997 non-anchored pattern. The fourth argument should point to an \fBint\fP
998 variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
999 still recognized for backwards compatibility.)
1000 .P
1001 If there is a fixed first byte, for example, from a pattern such as
1002 (cat|cow|coyote), its value is returned. Otherwise, if either
1003 .sp
1004 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1005 starts with "^", or
1006 .sp
1007 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1008 (if it were set, the pattern would be anchored),
1009 .sp
1010 -1 is returned, indicating that the pattern matches only at the start of a
1011 subject string or after any newline within the string. Otherwise -2 is
1012 returned. For anchored patterns, -2 is returned.
1013 .sp
1014 PCRE_INFO_FIRSTTABLE
1015 .sp
1016 If the pattern was studied, and this resulted in the construction of a 256-bit
1017 table indicating a fixed set of bytes for the first byte in any matching
1018 string, a pointer to the table is returned. Otherwise NULL is returned. The
1019 fourth argument should point to an \fBunsigned char *\fP variable.
1020 .sp
1021 PCRE_INFO_HASCRORLF
1022 .sp
1023 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1024 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1025 explicit match is either a literal CR or LF character, or \er or \en.
1026 .sp
1027 PCRE_INFO_JCHANGED
1028 .sp
1029 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1030 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1031 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1032 .sp
1033 PCRE_INFO_LASTLITERAL
1034 .sp
1035 Return the value of the rightmost literal byte that must exist in any matched
1036 string, other than at its start, if such a byte has been recorded. The fourth
1037 argument should point to an \fBint\fP variable. If there is no such byte, -1 is
1038 returned. For anchored patterns, a last literal byte is recorded only if it
1039 follows something of variable length. For example, for the pattern
1040 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1041 is -1.
1042 .sp
1043 PCRE_INFO_MINLENGTH
1044 .sp
1045 If the pattern was studied and a minimum length for matching subject strings
1046 was computed, its value is returned. Otherwise the returned value is -1. The
1047 value is a number of characters, not bytes (this may be relevant in UTF-8
1048 mode). The fourth argument should point to an \fBint\fP variable. A
1049 non-negative value is a lower bound to the length of any matching string. There
1050 may not be any strings of that length that do actually match, but every string
1051 that does match is at least that long.
1052 .sp
1053 PCRE_INFO_NAMECOUNT
1054 PCRE_INFO_NAMEENTRYSIZE
1055 PCRE_INFO_NAMETABLE
1056 .sp
1057 PCRE supports the use of named as well as numbered capturing parentheses. The
1058 names are just an additional way of identifying the parentheses, which still
1059 acquire numbers. Several convenience functions such as
1060 \fBpcre_get_named_substring()\fP are provided for extracting captured
1061 substrings by name. It is also possible to extract the data directly, by first
1062 converting the name to a number in order to access the correct pointers in the
1063 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1064 you need to use the name-to-number map, which is described by these three
1065 values.
1066 .P
1067 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1068 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1069 entry; both of these return an \fBint\fP value. The entry size depends on the
1070 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1071 entry of the table (a pointer to \fBchar\fP). The first two bytes of each entry
1072 are the number of the capturing parenthesis, most significant byte first. The
1073 rest of the entry is the corresponding name, zero terminated.
1074 .P
1075 The names are in alphabetical order. Duplicate names may appear if (?| is used
1076 to create multiple groups with the same number, as described in the
1077 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1078 .\" </a>
1079 section on duplicate subpattern numbers
1080 .\"
1081 in the
1082 .\" HREF
1083 \fBpcrepattern\fP
1084 .\"
1085 page. Duplicate names for subpatterns with different numbers are permitted only
1086 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1087 table in the order in which they were found in the pattern. In the absence of
1088 (?| this is the order of increasing number; when (?| is used this is not
1089 necessarily the case because later subpatterns may have lower numbers.
1090 .P
1091 As a simple example of the name/number table, consider the following pattern
1092 (assume PCRE_EXTENDED is set, so white space - including newlines - is
1093 ignored):
1094 .sp
1095 .\" JOIN
1096 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1097 (?<month>\ed\ed) - (?<day>\ed\ed) )
1098 .sp
1099 There are four named subpatterns, so the table has four entries, and each entry
1100 in the table is eight bytes long. The table is as follows, with non-printing
1101 bytes shows in hexadecimal, and undefined bytes shown as ??:
1102 .sp
1103 00 01 d a t e 00 ??
1104 00 05 d a y 00 ?? ??
1105 00 04 m o n t h 00
1106 00 02 y e a r 00 ??
1107 .sp
1108 When writing code to extract data from named subpatterns using the
1109 name-to-number map, remember that the length of the entries is likely to be
1110 different for each compiled pattern.
1111 .sp
1112 PCRE_INFO_OKPARTIAL
1113 .sp
1114 Return 1 if the pattern can be used for partial matching with
1115 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1116 \fBint\fP variable. From release 8.00, this always returns 1, because the
1117 restrictions that previously applied to partial matching have been lifted. The
1118 .\" HREF
1119 \fBpcrepartial\fP
1120 .\"
1121 documentation gives details of partial matching.
1122 .sp
1123 PCRE_INFO_OPTIONS
1124 .sp
1125 Return a copy of the options with which the pattern was compiled. The fourth
1126 argument should point to an \fBunsigned long int\fP variable. These option bits
1127 are those specified in the call to \fBpcre_compile()\fP, modified by any
1128 top-level option settings at the start of the pattern itself. In other words,
1129 they are the options that will be in force when matching starts. For example,
1130 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1131 result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1132 .P
1133 A pattern is automatically anchored by PCRE if all of its top-level
1134 alternatives begin with one of the following:
1135 .sp
1136 ^ unless PCRE_MULTILINE is set
1137 \eA always
1138 \eG always
1139 .\" JOIN
1140 .* if PCRE_DOTALL is set and there are no back
1141 references to the subpattern in which .* appears
1142 .sp
1143 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1144 \fBpcre_fullinfo()\fP.
1145 .sp
1146 PCRE_INFO_SIZE
1147 .sp
1148 Return the size of the compiled pattern, that is, the value that was passed as
1149 the argument to \fBpcre_malloc()\fP when PCRE was getting memory in which to
1150 place the compiled data. The fourth argument should point to a \fBsize_t\fP
1151 variable.
1152 .sp
1153 PCRE_INFO_STUDYSIZE
1154 .sp
1155 Return the size of the data block pointed to by the \fIstudy_data\fP field in
1156 a \fBpcre_extra\fP block. That is, it is the value that was passed to
1157 \fBpcre_malloc()\fP when PCRE was getting memory into which to place the data
1158 created by \fBpcre_study()\fP. If \fBpcre_extra\fP is NULL, or there is no
1159 study data, zero is returned. The fourth argument should point to a
1160 \fBsize_t\fP variable.
1161 .
1162 .
1163 .SH "OBSOLETE INFO FUNCTION"
1164 .rs
1165 .sp
1166 .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
1167 .B *\fIfirstcharptr\fP);
1168 .PP
1169 The \fBpcre_info()\fP function is now obsolete because its interface is too
1170 restrictive to return all the available data about a compiled pattern. New
1171 programs should use \fBpcre_fullinfo()\fP instead. The yield of
1172 \fBpcre_info()\fP is the number of capturing subpatterns, or one of the
1173 following negative numbers:
1174 .sp
1175 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1176 PCRE_ERROR_BADMAGIC the "magic number" was not found
1177 .sp
1178 If the \fIoptptr\fP argument is not NULL, a copy of the options with which the
1179 pattern was compiled is placed in the integer it points to (see
1180 PCRE_INFO_OPTIONS above).
1181 .P
1182 If the pattern is not anchored and the \fIfirstcharptr\fP argument is not NULL,
1183 it is used to pass back information about the first character of any matched
1184 string (see PCRE_INFO_FIRSTBYTE above).
1185 .
1186 .
1187 .SH "REFERENCE COUNTS"
1188 .rs
1189 .sp
1190 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1191 .PP
1192 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1193 data block that contains a compiled pattern. It is provided for the benefit of
1194 applications that operate in an object-oriented manner, where different parts
1195 of the application may be using the same compiled pattern, but you want to free
1196 the block when they are all done.
1197 .P
1198 When a pattern is compiled, the reference count field is initialized to zero.
1199 It is changed only by calling this function, whose action is to add the
1200 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1201 function is the new value. However, the value of the count is constrained to
1202 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1203 it is forced to the appropriate limit value.
1204 .P
1205 Except when it is zero, the reference count is not correctly preserved if a
1206 pattern is compiled on one host and then transferred to a host whose byte-order
1207 is different. (This seems a highly unlikely scenario.)
1208 .
1209 .
1210 .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1211 .rs
1212 .sp
1213 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1214 .ti +5n
1215 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1216 .ti +5n
1217 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1218 .P
1219 The function \fBpcre_exec()\fP is called to match a subject string against a
1220 compiled pattern, which is passed in the \fIcode\fP argument. If the
1221 pattern was studied, the result of the study should be passed in the
1222 \fIextra\fP argument. This function is the main matching facility of the
1223 library, and it operates in a Perl-like manner. For specialist use there is
1224 also an alternative matching function, which is described
1225 .\" HTML <a href="#dfamatch">
1226 .\" </a>
1227 below
1228 .\"
1229 in the section about the \fBpcre_dfa_exec()\fP function.
1230 .P
1231 In most applications, the pattern will have been compiled (and optionally
1232 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1233 possible to save compiled patterns and study data, and then use them later
1234 in different processes, possibly even on different hosts. For a discussion
1235 about this, see the
1236 .\" HREF
1237 \fBpcreprecompile\fP
1238 .\"
1239 documentation.
1240 .P
1241 Here is an example of a simple call to \fBpcre_exec()\fP:
1242 .sp
1243 int rc;
1244 int ovector[30];
1245 rc = pcre_exec(
1246 re, /* result of pcre_compile() */
1247 NULL, /* we didn't study the pattern */
1248 "some string", /* the subject string */
1249 11, /* the length of the subject string */
1250 0, /* start at offset 0 in the subject */
1251 0, /* default options */
1252 ovector, /* vector of integers for substring information */
1253 30); /* number of elements (NOT size in bytes) */
1254 .
1255 .\" HTML <a name="extradata"></a>
1256 .SS "Extra data for \fBpcre_exec()\fR"
1257 .rs
1258 .sp
1259 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1260 data block. The \fBpcre_study()\fP function returns such a block (when it
1261 doesn't return NULL), but you can also create one for yourself, and pass
1262 additional information in it. The \fBpcre_extra\fP block contains the following
1263 fields (not necessarily in this order):
1264 .sp
1265 unsigned long int \fIflags\fP;
1266 void *\fIstudy_data\fP;
1267 unsigned long int \fImatch_limit\fP;
1268 unsigned long int \fImatch_limit_recursion\fP;
1269 void *\fIcallout_data\fP;
1270 const unsigned char *\fItables\fP;
1271 unsigned char **\fImark\fP;
1272 .sp
1273 The \fIflags\fP field is a bitmap that specifies which of the other fields
1274 are set. The flag bits are:
1275 .sp
1276 PCRE_EXTRA_STUDY_DATA
1277 PCRE_EXTRA_MATCH_LIMIT
1278 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1279 PCRE_EXTRA_CALLOUT_DATA
1280 PCRE_EXTRA_TABLES
1281 PCRE_EXTRA_MARK
1282 .sp
1283 Other flag bits should be set to zero. The \fIstudy_data\fP field is set in the
1284 \fBpcre_extra\fP block that is returned by \fBpcre_study()\fP, together with
1285 the appropriate flag bit. You should not set this yourself, but you may add to
1286 the block by setting the other fields and their corresponding flag bits.
1287 .P
1288 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1289 vast amount of resources when running patterns that are not going to match,
1290 but which have a very large number of possibilities in their search trees. The
1291 classic example is a pattern that uses nested unlimited repeats.
1292 .P
1293 Internally, PCRE uses a function called \fBmatch()\fP which it calls repeatedly
1294 (sometimes recursively). The limit set by \fImatch_limit\fP is imposed on the
1295 number of times this function is called during a match, which has the effect of
1296 limiting the amount of backtracking that can take place. For patterns that are
1297 not anchored, the count restarts from zero for each position in the subject
1298 string.
1299 .P
1300 The default value for the limit can be set when PCRE is built; the default
1301 default is 10 million, which handles all but the most extreme cases. You can
1302 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1303 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1304 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1305 PCRE_ERROR_MATCHLIMIT.
1306 .P
1307 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1308 instead of limiting the total number of times that \fBmatch()\fP is called, it
1309 limits the depth of recursion. The recursion depth is a smaller number than the
1310 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1311 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1312 .P
1313 Limiting the recursion depth limits the amount of stack that can be used, or,
1314 when PCRE has been compiled to use memory on the heap instead of the stack, the
1315 amount of heap memory that can be used.
1316 .P
1317 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1318 built; the default default is the same value as the default for
1319 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1320 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1321 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1322 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1323 .P
1324 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1325 and is described in the
1326 .\" HREF
1327 \fBpcrecallout\fP
1328 .\"
1329 documentation.
1330 .P
1331 The \fItables\fP field is used to pass a character tables pointer to
1332 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1333 pattern. A non-NULL value is stored with the compiled pattern only if custom
1334 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1335 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1336 internal tables to be used. This facility is helpful when re-using patterns
1337 that have been saved after compiling with an external set of tables, because
1338 the external tables might be at a different address when \fBpcre_exec()\fP is
1339 called. See the
1340 .\" HREF
1341 \fBpcreprecompile\fP
1342 .\"
1343 documentation for a discussion of saving compiled patterns for later use.
1344 .P
1345 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1346 be set to point to a \fBchar *\fP variable. If the pattern contains any
1347 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1348 a name to pass back, a pointer to the name string (zero terminated) is placed
1349 in the variable pointed to by the \fImark\fP field. The names are within the
1350 compiled pattern; if you wish to retain such a name you must copy it before
1351 freeing the memory of a compiled pattern. If there is no name to pass back, the
1352 variable pointed to by the \fImark\fP field set to NULL. For details of the
1353 backtracking control verbs, see the section entitled
1354 .\" HTML <a href="pcrepattern#backtrackcontrol">
1355 .\" </a>
1356 "Backtracking control"
1357 .\"
1358 in the
1359 .\" HREF
1360 \fBpcrepattern\fP
1361 .\"
1362 documentation.
1363 .
1364 .
1365 .\" HTML <a name="execoptions"></a>
1366 .SS "Option bits for \fBpcre_exec()\fP"
1367 .rs
1368 .sp
1369 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1370 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1371 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1372 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_SOFT, and
1373 PCRE_PARTIAL_HARD.
1374 .sp
1375 PCRE_ANCHORED
1376 .sp
1377 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1378 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1379 to be anchored by virtue of its contents, it cannot be made unachored at
1380 matching time.
1381 .sp
1382 PCRE_BSR_ANYCRLF
1383 PCRE_BSR_UNICODE
1384 .sp
1385 These options (which are mutually exclusive) control what the \eR escape
1386 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1387 match any Unicode newline sequence. These options override the choice that was
1388 made or defaulted when the pattern was compiled.
1389 .sp
1390 PCRE_NEWLINE_CR
1391 PCRE_NEWLINE_LF
1392 PCRE_NEWLINE_CRLF
1393 PCRE_NEWLINE_ANYCRLF
1394 PCRE_NEWLINE_ANY
1395 .sp
1396 These options override the newline definition that was chosen or defaulted when
1397 the pattern was compiled. For details, see the description of
1398 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1399 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1400 the way the match position is advanced after a match failure for an unanchored
1401 pattern.
1402 .P
1403 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1404 match attempt for an unanchored pattern fails when the current position is at a
1405 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1406 characters, the match position is advanced by two characters instead of one, in
1407 other words, to after the CRLF.
1408 .P
1409 The above rule is a compromise that makes the most common cases work as
1410 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1411 set), it does not match the string "\er\enA" because, after failing at the
1412 start, it skips both the CR and the LF before retrying. However, the pattern
1413 [\er\en]A does match that string, because it contains an explicit CR or LF
1414 reference, and so advances only by one character after the first failure.
1415 .P
1416 An explicit match for CR of LF is either a literal appearance of one of those
1417 characters, or one of the \er or \en escape sequences. Implicit matches such as
1418 [^X] do not count, nor does \es (which includes CR and LF in the characters
1419 that it matches).
1420 .P
1421 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1422 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1423 .sp
1424 PCRE_NOTBOL
1425 .sp
1426 This option specifies that first character of the subject string is not the
1427 beginning of a line, so the circumflex metacharacter should not match before
1428 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1429 never to match. This option affects only the behaviour of the circumflex
1430 metacharacter. It does not affect \eA.
1431 .sp
1432 PCRE_NOTEOL
1433 .sp
1434 This option specifies that the end of the subject string is not the end of a
1435 line, so the dollar metacharacter should not match it nor (except in multiline
1436 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1437 compile time) causes dollar never to match. This option affects only the
1438 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1439 .sp
1440 PCRE_NOTEMPTY
1441 .sp
1442 An empty string is not considered to be a valid match if this option is set. If
1443 there are alternatives in the pattern, they are tried. If all the alternatives
1444 match the empty string, the entire match fails. For example, if the pattern
1445 .sp
1446 a?b?
1447 .sp
1448 is applied to a string not beginning with "a" or "b", it matches an empty
1449 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1450 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1451 .sp
1452 PCRE_NOTEMPTY_ATSTART
1453 .sp
1454 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1455 the start of the subject is permitted. If the pattern is anchored, such a match
1456 can occur only if the pattern contains \eK.
1457 .P
1458 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1459 does make a special case of a pattern match of the empty string within its
1460 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1461 emulate Perl's behaviour after matching a null string by first trying the match
1462 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1463 if that fails, by advancing the starting offset (see below) and trying an
1464 ordinary match again. There is some code that demonstrates how to do this in
1465 the
1466 .\" HREF
1467 \fBpcredemo\fP
1468 .\"
1469 sample program. In the most general case, you have to check to see if the
1470 newline convention recognizes CRLF as a newline, and if so, and the current
1471 character is CR followed by LF, advance the starting offset by two characters
1472 instead of one.
1473 .sp
1474 PCRE_NO_START_OPTIMIZE
1475 .sp
1476 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1477 a match, in order to speed up the process. For example, if it is known that an
1478 unanchored match must start with a specific character, it searches the subject
1479 for that character, and fails immediately if it cannot find it, without
1480 actually running the main matching function. This means that a special item
1481 such as (*COMMIT) at the start of a pattern is not considered until after a
1482 suitable starting point for the match has been found. When callouts or (*MARK)
1483 items are in use, these "start-up" optimizations can cause them to be skipped
1484 if the pattern is never actually used. The start-up optimizations are in effect
1485 a pre-scan of the subject that takes place before the pattern is run.
1486 .P
1487 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1488 causing performance to suffer, but ensuring that in cases where the result is
1489 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1490 are considered at every possible starting position in the subject string.
1491 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1492 Consider the pattern
1493 .sp
1494 (*COMMIT)ABC
1495 .sp
1496 When this is compiled, PCRE records the fact that a match must start with the
1497 character "A". Suppose the subject string is "DEFABC". The start-up
1498 optimization scans along the subject, finds "A" and runs the first match
1499 attempt from there. The (*COMMIT) item means that the pattern must match the
1500 current starting position, which in this case, it does. However, if the same
1501 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1502 subject string does not happen. The first match attempt is run starting from
1503 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1504 the overall result is "no match". If the pattern is studied, more start-up
1505 optimizations may be used. For example, a minimum length for the subject may be
1506 recorded. Consider the pattern
1507 .sp
1508 (*MARK:A)(X|Y)
1509 .sp
1510 The minimum length for a match is one character. If the subject is "ABC", there
1511 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1512 If the pattern is studied, the final attempt does not take place, because PCRE
1513 knows that the subject is too short, and so the (*MARK) is never encountered.
1514 In this case, studying the pattern does not affect the overall match result,
1515 which is still "no match", but it does affect the auxiliary information that is
1516 returned.
1517 .sp
1518 PCRE_NO_UTF8_CHECK
1519 .sp
1520 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1521 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1522 The value of \fIstartoffset\fP is also checked to ensure that it points to the
1523 start of a UTF-8 character. There is a discussion about the validity of UTF-8
1524 strings in the
1525 .\" HTML <a href="pcre.html#utf8strings">
1526 .\" </a>
1527 section on UTF-8 support
1528 .\"
1529 in the main
1530 .\" HREF
1531 \fBpcre\fP
1532 .\"
1533 page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_exec()\fP returns
1534 the error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is
1535 a truncated UTF-8 character at the end of the subject, PCRE_ERROR_SHORTUTF8. If
1536 \fIstartoffset\fP contains a value that does not point to the start of a UTF-8
1537 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1538 returned.
1539 .P
1540 If you already know that your subject is valid, and you want to skip these
1541 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1542 calling \fBpcre_exec()\fP. You might want to do this for the second and
1543 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1544 all the matches in a single subject string. However, you should be sure that
1545 the value of \fIstartoffset\fP points to the start of a UTF-8 character (or the
1546 end of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1547 invalid UTF-8 string as a subject or an invalid value of \fIstartoffset\fP is
1548 undefined. Your program may crash.
1549 .sp
1550 PCRE_PARTIAL_HARD
1551 PCRE_PARTIAL_SOFT
1552 .sp
1553 These options turn on the partial matching feature. For backwards
1554 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1555 occurs if the end of the subject string is reached successfully, but there are
1556 not enough subject characters to complete the match. If this happens when
1557 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1558 testing any remaining alternatives. Only if no complete match can be found is
1559 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1560 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1561 but only if no complete match can be found.
1562 .P
1563 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1564 partial match is found, \fBpcre_exec()\fP immediately returns
1565 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1566 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1567 important that an alternative complete match.
1568 .P
1569 In both cases, the portion of the string that was inspected when the partial
1570 match was found is set as the first matching string. There is a more detailed
1571 discussion of partial and multi-segment matching, with examples, in the
1572 .\" HREF
1573 \fBpcrepartial\fP
1574 .\"
1575 documentation.
1576 .
1577 .
1578 .SS "The string to be matched by \fBpcre_exec()\fP"
1579 .rs
1580 .sp
1581 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1582 \fIsubject\fP, a length (in bytes) in \fIlength\fP, and a starting byte offset
1583 in \fIstartoffset\fP. If this is negative or greater than the length of the
1584 subject, \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting
1585 offset is zero, the search for a match starts at the beginning of the subject,
1586 and this is by far the most common case. In UTF-8 mode, the byte offset must
1587 point to the start of a UTF-8 character (or the end of the subject). Unlike the
1588 pattern string, the subject may contain binary zero bytes.
1589 .P
1590 A non-zero starting offset is useful when searching for another match in the
1591 same subject by calling \fBpcre_exec()\fP again after a previous success.
1592 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1593 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1594 lookbehind. For example, consider the pattern
1595 .sp
1596 \eBiss\eB
1597 .sp
1598 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1599 the current position in the subject is not a word boundary.) When applied to
1600 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1601 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1602 subject, namely "issipi", it does not match, because \eB is always false at the
1603 start of the subject, which is deemed to be a word boundary. However, if
1604 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1605 set to 4, it finds the second occurrence of "iss" because it is able to look
1606 behind the starting point to discover that it is preceded by a letter.
1607 .P
1608 Finding all the matches in a subject is tricky when the pattern can match an
1609 empty string. It is possible to emulate Perl's /g behaviour by first trying the
1610 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1611 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1612 and trying an ordinary match again. There is some code that demonstrates how to
1613 do this in the
1614 .\" HREF
1615 \fBpcredemo\fP
1616 .\"
1617 sample program. In the most general case, you have to check to see if the
1618 newline convention recognizes CRLF as a newline, and if so, and the current
1619 character is CR followed by LF, advance the starting offset by two characters
1620 instead of one.
1621 .P
1622 If a non-zero starting offset is passed when the pattern is anchored, one
1623 attempt to match at the given offset is made. This can only succeed if the
1624 pattern does not require the match to be at the start of the subject.
1625 .
1626 .
1627 .SS "How \fBpcre_exec()\fP returns captured substrings"
1628 .rs
1629 .sp
1630 In general, a pattern matches a certain portion of the subject, and in
1631 addition, further substrings from the subject may be picked out by parts of the
1632 pattern. Following the usage in Jeffrey Friedl's book, this is called
1633 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1634 a fragment of a pattern that picks out a substring. PCRE supports several other
1635 kinds of parenthesized subpattern that do not cause substrings to be captured.
1636 .P
1637 Captured substrings are returned to the caller via a vector of integers whose
1638 address is passed in \fIovector\fP. The number of elements in the vector is
1639 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
1640 argument is NOT the size of \fIovector\fP in bytes.
1641 .P
1642 The first two-thirds of the vector is used to pass back captured substrings,
1643 each substring using a pair of integers. The remaining third of the vector is
1644 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1645 and is not available for passing back information. The number passed in
1646 \fIovecsize\fP should always be a multiple of three. If it is not, it is
1647 rounded down.
1648 .P
1649 When a match is successful, information about captured substrings is returned
1650 in pairs of integers, starting at the beginning of \fIovector\fP, and
1651 continuing up to two-thirds of its length at the most. The first element of
1652 each pair is set to the byte offset of the first character in a substring, and
1653 the second is set to the byte offset of the first character after the end of a
1654 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
1655 mode. They are not character counts.
1656 .P
1657 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
1658 portion of the subject string matched by the entire pattern. The next pair is
1659 used for the first capturing subpattern, and so on. The value returned by
1660 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
1661 For example, if two substrings have been captured, the returned value is 3. If
1662 there are no capturing subpatterns, the return value from a successful match is
1663 1, indicating that just the first pair of offsets has been set.
1664 .P
1665 If a capturing subpattern is matched repeatedly, it is the last portion of the
1666 string that it matched that is returned.
1667 .P
1668 If the vector is too small to hold all the captured substring offsets, it is
1669 used as far as possible (up to two-thirds of its length), and the function
1670 returns a value of zero. If the substring offsets are not of interest,
1671 \fBpcre_exec()\fP may be called with \fIovector\fP passed as NULL and
1672 \fIovecsize\fP as zero. However, if the pattern contains back references and
1673 the \fIovector\fP is not big enough to remember the related substrings, PCRE
1674 has to get additional memory for use during matching. Thus it is usually
1675 advisable to supply an \fIovector\fP.
1676 .P
1677 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
1678 subpatterns there are in a compiled pattern. The smallest size for
1679 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
1680 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
1681 .P
1682 It is possible for capturing subpattern number \fIn+1\fP to match some part of
1683 the subject when subpattern \fIn\fP has not been used at all. For example, if
1684 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1685 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1686 happens, both values in the offset pairs corresponding to unused subpatterns
1687 are set to -1.
1688 .P
1689 Offset values that correspond to unused subpatterns at the end of the
1690 expression are also set to -1. For example, if the string "abc" is matched
1691 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1692 return from the function is 2, because the highest used capturing subpattern
1693 number is 1, and the offsets for for the second and third capturing subpatterns
1694 (assuming the vector is large enough, of course) are set to -1.
1695 .P
1696 \fBNote\fP: Elements of \fIovector\fP that do not correspond to capturing
1697 parentheses in the pattern are never changed. That is, if a pattern contains
1698 \fIn\fP capturing parentheses, no more than \fIovector[0]\fP to
1699 \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other elements retain
1700 whatever values they previously had.
1701 .P
1702 Some convenience functions are provided for extracting the captured substrings
1703 as separate strings. These are described below.
1704 .
1705 .\" HTML <a name="errorlist"></a>
1706 .SS "Error return values from \fBpcre_exec()\fP"
1707 .rs
1708 .sp
1709 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1710 defined in the header file:
1711 .sp
1712 PCRE_ERROR_NOMATCH (-1)
1713 .sp
1714 The subject string did not match the pattern.
1715 .sp
1716 PCRE_ERROR_NULL (-2)
1717 .sp
1718 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
1719 NULL and \fIovecsize\fP was not zero.
1720 .sp
1721 PCRE_ERROR_BADOPTION (-3)
1722 .sp
1723 An unrecognized bit was set in the \fIoptions\fP argument.
1724 .sp
1725 PCRE_ERROR_BADMAGIC (-4)
1726 .sp
1727 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1728 the case when it is passed a junk pointer and to detect when a pattern that was
1729 compiled in an environment of one endianness is run in an environment with the
1730 other endianness. This is the error that PCRE gives when the magic number is
1731 not present.
1732 .sp
1733 PCRE_ERROR_UNKNOWN_OPCODE (-5)
1734 .sp
1735 While running the pattern match, an unknown item was encountered in the
1736 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1737 of the compiled pattern.
1738 .sp
1739 PCRE_ERROR_NOMEMORY (-6)
1740 .sp
1741 If a pattern contains back references, but the \fIovector\fP that is passed to
1742 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
1743 gets a block of memory at the start of matching to use for this purpose. If the
1744 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
1745 automatically freed at the end of matching.
1746 .P
1747 This error is also given if \fBpcre_stack_malloc()\fP fails in
1748 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
1749 \fB--disable-stack-for-recursion\fP.
1750 .sp
1751 PCRE_ERROR_NOSUBSTRING (-7)
1752 .sp
1753 This error is used by the \fBpcre_copy_substring()\fP,
1754 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
1755 below). It is never returned by \fBpcre_exec()\fP.
1756 .sp
1757 PCRE_ERROR_MATCHLIMIT (-8)
1758 .sp
1759 The backtracking limit, as specified by the \fImatch_limit\fP field in a
1760 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
1761 above.
1762 .sp
1763 PCRE_ERROR_CALLOUT (-9)
1764 .sp
1765 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
1766 use by callout functions that want to yield a distinctive error code. See the
1767 .\" HREF
1768 \fBpcrecallout\fP
1769 .\"
1770 documentation for details.
1771 .sp
1772 PCRE_ERROR_BADUTF8 (-10)
1773 .sp
1774 A string that contains an invalid UTF-8 byte sequence was passed as a subject.
1775 However, if PCRE_PARTIAL_HARD is set and the problem is a truncated UTF-8
1776 character at the end of the subject, PCRE_ERROR_SHORTUTF8 is used instead.
1777 .sp
1778 PCRE_ERROR_BADUTF8_OFFSET (-11)
1779 .sp
1780 The UTF-8 byte sequence that was passed as a subject was valid, but the value
1781 of \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
1782 end of the subject.
1783 .sp
1784 PCRE_ERROR_PARTIAL (-12)
1785 .sp
1786 The subject string did not match, but it did match partially. See the
1787 .\" HREF
1788 \fBpcrepartial\fP
1789 .\"
1790 documentation for details of partial matching.
1791 .sp
1792 PCRE_ERROR_BADPARTIAL (-13)
1793 .sp
1794 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1795 option was used with a compiled pattern containing items that were not
1796 supported for partial matching. From release 8.00 onwards, there are no
1797 restrictions on partial matching.
1798 .sp
1799 PCRE_ERROR_INTERNAL (-14)
1800 .sp
1801 An unexpected internal error has occurred. This error could be caused by a bug
1802 in PCRE or by overwriting of the compiled pattern.
1803 .sp
1804 PCRE_ERROR_BADCOUNT (-15)
1805 .sp
1806 This error is given if the value of the \fIovecsize\fP argument is negative.
1807 .sp
1808 PCRE_ERROR_RECURSIONLIMIT (-21)
1809 .sp
1810 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
1811 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
1812 description above.
1813 .sp
1814 PCRE_ERROR_BADNEWLINE (-23)
1815 .sp
1816 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
1817 .sp
1818 PCRE_ERROR_BADOFFSET (-24)
1819 .sp
1820 The value of \fIstartoffset\fP was negative or greater than the length of the
1821 subject, that is, the value in \fIlength\fP.
1822 .sp
1823 PCRE_ERROR_SHORTUTF8 (-25)
1824 .sp
1825 The subject string ended with an incomplete (truncated) UTF-8 character, and
1826 the PCRE_PARTIAL_HARD option was set. Without this option, PCRE_ERROR_BADUTF8
1827 is returned in this situation.
1828 .P
1829 Error numbers -16 to -20 and -22 are not used by \fBpcre_exec()\fP.
1830 .
1831 .
1832 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
1833 .rs
1834 .sp
1835 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1836 .ti +5n
1837 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
1838 .ti +5n
1839 .B int \fIbuffersize\fP);
1840 .PP
1841 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1842 .ti +5n
1843 .B int \fIstringcount\fP, int \fIstringnumber\fP,
1844 .ti +5n
1845 .B const char **\fIstringptr\fP);
1846 .PP
1847 .B int pcre_get_substring_list(const char *\fIsubject\fP,
1848 .ti +5n
1849 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
1850 .PP
1851 Captured substrings can be accessed directly by using the offsets returned by
1852 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
1853 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
1854 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
1855 as new, separate, zero-terminated strings. These functions identify substrings
1856 by number. The next section describes functions for extracting named
1857 substrings.
1858 .P
1859 A substring that contains a binary zero is correctly extracted and has a
1860 further zero added on the end, but the result is not, of course, a C string.
1861 However, you can process such a string by referring to the length that is
1862 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
1863 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
1864 for handling strings containing binary zeros, because the end of the final
1865 string is not independently indicated.
1866 .P
1867 The first three arguments are the same for all three of these functions:
1868 \fIsubject\fP is the subject string that has just been successfully matched,
1869 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
1870 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
1871 captured by the match, including the substring that matched the entire regular
1872 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
1873 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
1874 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
1875 number of elements in the vector divided by three.
1876 .P
1877 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
1878 extract a single substring, whose number is given as \fIstringnumber\fP. A
1879 value of zero extracts the substring that matched the entire pattern, whereas
1880 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
1881 the string is placed in \fIbuffer\fP, whose length is given by
1882 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
1883 obtained via \fBpcre_malloc\fP, and its address is returned via
1884 \fIstringptr\fP. The yield of the function is the length of the string, not
1885 including the terminating zero, or one of these error codes:
1886 .sp
1887 PCRE_ERROR_NOMEMORY (-6)
1888 .sp
1889 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
1890 memory failed for \fBpcre_get_substring()\fP.
1891 .sp
1892 PCRE_ERROR_NOSUBSTRING (-7)
1893 .sp
1894 There is no substring whose number is \fIstringnumber\fP.
1895 .P
1896 The \fBpcre_get_substring_list()\fP function extracts all available substrings
1897 and builds a list of pointers to them. All this is done in a single block of
1898 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
1899 is returned via \fIlistptr\fP, which is also the start of the list of string
1900 pointers. The end of the list is marked by a NULL pointer. The yield of the
1901 function is zero if all went well, or the error code
1902 .sp
1903 PCRE_ERROR_NOMEMORY (-6)
1904 .sp
1905 if the attempt to get the memory block failed.
1906 .P
1907 When any of these functions encounter a substring that is unset, which can
1908 happen when capturing subpattern number \fIn+1\fP matches some part of the
1909 subject, but subpattern \fIn\fP has not been used at all, they return an empty
1910 string. This can be distinguished from a genuine zero-length substring by
1911 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
1912 substrings.
1913 .P
1914 The two convenience functions \fBpcre_free_substring()\fP and
1915 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
1916 a previous call of \fBpcre_get_substring()\fP or
1917 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
1918 the function pointed to by \fBpcre_free\fP, which of course could be called
1919 directly from a C program. However, PCRE is used in some situations where it is
1920 linked via a special interface to another programming language that cannot use
1921 \fBpcre_free\fP directly; it is for these cases that the functions are
1922 provided.
1923 .
1924 .
1925 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
1926 .rs
1927 .sp
1928 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
1929 .ti +5n
1930 .B const char *\fIname\fP);
1931 .PP
1932 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
1933 .ti +5n
1934 .B const char *\fIsubject\fP, int *\fIovector\fP,
1935 .ti +5n
1936 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1937 .ti +5n
1938 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
1939 .PP
1940 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
1941 .ti +5n
1942 .B const char *\fIsubject\fP, int *\fIovector\fP,
1943 .ti +5n
1944 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1945 .ti +5n
1946 .B const char **\fIstringptr\fP);
1947 .PP
1948 To extract a substring by name, you first have to find associated number.
1949 For example, for this pattern
1950 .sp
1951 (a+)b(?<xxx>\ed+)...
1952 .sp
1953 the number of the subpattern called "xxx" is 2. If the name is known to be
1954 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
1955 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
1956 pattern, and the second is the name. The yield of the function is the
1957 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
1958 that name.
1959 .P
1960 Given the number, you can extract the substring directly, or use one of the
1961 functions described in the previous section. For convenience, there are also
1962 two functions that do the whole job.
1963 .P
1964 Most of the arguments of \fBpcre_copy_named_substring()\fP and
1965 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
1966 functions that extract by number. As these are described in the previous
1967 section, they are not re-described here. There are just two differences:
1968 .P
1969 First, instead of a substring number, a substring name is given. Second, there
1970 is an extra argument, given at the start, which is a pointer to the compiled
1971 pattern. This is needed in order to gain access to the name-to-number
1972 translation table.
1973 .P
1974 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
1975 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
1976 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
1977 the behaviour may not be what you want (see the next section).
1978 .P
1979 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
1980 subpatterns with the same number, as described in the
1981 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1982 .\" </a>
1983 section on duplicate subpattern numbers
1984 .\"
1985 in the
1986 .\" HREF
1987 \fBpcrepattern\fP
1988 .\"
1989 page, you cannot use names to distinguish the different subpatterns, because
1990 names are not included in the compiled code. The matching process uses only
1991 numbers. For this reason, the use of different names for subpatterns of the
1992 same number causes an error at compile time.
1993 .
1994 .SH "DUPLICATE SUBPATTERN NAMES"
1995 .rs
1996 .sp
1997 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
1998 .ti +5n
1999 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
2000 .PP
2001 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2002 are not required to be unique. (Duplicate names are always allowed for
2003 subpatterns with the same number, created by using the (?| feature. Indeed, if
2004 such subpatterns are named, they are required to use the same names.)
2005 .P
2006 Normally, patterns with duplicate names are such that in any one match, only
2007 one of the named subpatterns participates. An example is shown in the
2008 .\" HREF
2009 \fBpcrepattern\fP
2010 .\"
2011 documentation.
2012 .P
2013 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2014 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2015 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2016 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2017 returns one of the numbers that are associated with the name, but it is not
2018 defined which it is.
2019 .P
2020 If you want to get full details of all captured substrings for a given name,
2021 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2022 argument is the compiled pattern, and the second is the name. The third and
2023 fourth are pointers to variables which are updated by the function. After it
2024 has run, they point to the first and last entries in the name-to-number table
2025 for the given name. The function itself returns the length of each entry, or
2026 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2027 described above in the section entitled \fIInformation about a pattern\fP.
2028 Given all the relevant entries for the name, you can extract each of their
2029 numbers, and hence the captured data, if any.
2030 .
2031 .
2032 .SH "FINDING ALL POSSIBLE MATCHES"
2033 .rs
2034 .sp
2035 The traditional matching function uses a similar algorithm to Perl, which stops
2036 when it finds the first match, starting at a given point in the subject. If you
2037 want to find all possible matches, or the longest possible match, consider
2038 using the alternative matching function (see below) instead. If you cannot use
2039 the alternative function, but still need to find all possible matches, you
2040 can kludge it up by making use of the callout facility, which is described in
2041 the
2042 .\" HREF
2043 \fBpcrecallout\fP
2044 .\"
2045 documentation.
2046 .P
2047 What you have to do is to insert a callout right at the end of the pattern.
2048 When your callout function is called, extract and save the current matched
2049 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2050 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2051 will yield PCRE_ERROR_NOMATCH.
2052 .
2053 .
2054 .\" HTML <a name="dfamatch"></a>
2055 .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2056 .rs
2057 .sp
2058 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2059 .ti +5n
2060 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2061 .ti +5n
2062 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2063 .ti +5n
2064 .B int *\fIworkspace\fP, int \fIwscount\fP);
2065 .P
2066 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2067 a compiled pattern, using a matching algorithm that scans the subject string
2068 just once, and does not backtrack. This has different characteristics to the
2069 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2070 patterns are not supported. Nevertheless, there are times when this kind of
2071 matching can be useful. For a discussion of the two matching algorithms, and a
2072 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2073 .\" HREF
2074 \fBpcrematching\fP
2075 .\"
2076 documentation.
2077 .P
2078 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2079 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2080 different way, and this is described below. The other common arguments are used
2081 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2082 here.
2083 .P
2084 The two additional arguments provide workspace for the function. The workspace
2085 vector should contain at least 20 elements. It is used for keeping track of
2086 multiple paths through the pattern tree. More workspace will be needed for
2087 patterns and subjects where there are a lot of potential matches.
2088 .P
2089 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2090 .sp
2091 int rc;
2092 int ovector[10];
2093 int wspace[20];
2094 rc = pcre_dfa_exec(
2095 re, /* result of pcre_compile() */
2096 NULL, /* we didn't study the pattern */
2097 "some string", /* the subject string */
2098 11, /* the length of the subject string */
2099 0, /* start at offset 0 in the subject */
2100 0, /* default options */
2101 ovector, /* vector of integers for substring information */
2102 10, /* number of elements (NOT size in bytes) */
2103 wspace, /* working space vector */
2104 20); /* number of elements (NOT size in bytes) */
2105 .
2106 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2107 .rs
2108 .sp
2109 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2110 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2111 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2112 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2113 PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2114 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2115 so their description is not repeated here.
2116 .sp
2117 PCRE_PARTIAL_HARD
2118 PCRE_PARTIAL_SOFT
2119 .sp
2120 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2121 details are slightly different. When PCRE_PARTIAL_HARD is set for
2122 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2123 is reached and there is still at least one matching possibility that requires
2124 additional characters. This happens even if some complete matches have also
2125 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2126 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2127 there have been no complete matches, but there is still at least one matching
2128 possibility. The portion of the string that was inspected when the longest
2129 partial match was found is set as the first matching string in both cases.
2130 There is a more detailed discussion of partial and multi-segment matching, with
2131 examples, in the
2132 .\" HREF
2133 \fBpcrepartial\fP
2134 .\"
2135 documentation.
2136 .sp
2137 PCRE_DFA_SHORTEST
2138 .sp
2139 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2140 soon as it has found one match. Because of the way the alternative algorithm
2141 works, this is necessarily the shortest possible match at the first possible
2142 matching point in the subject string.
2143 .sp
2144 PCRE_DFA_RESTART
2145 .sp
2146 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2147 again, with additional subject characters, and have it continue with the same
2148 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2149 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2150 before because data about the match so far is left in them after a partial
2151 match. There is more discussion of this facility in the
2152 .\" HREF
2153 \fBpcrepartial\fP
2154 .\"
2155 documentation.
2156 .
2157 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2158 .rs
2159 .sp
2160 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2161 substring in the subject. Note, however, that all the matches from one run of
2162 the function start at the same point in the subject. The shorter matches are
2163 all initial substrings of the longer matches. For example, if the pattern
2164 .sp
2165 <.*>
2166 .sp
2167 is matched against the string
2168 .sp
2169 This is <something> <something else> <something further> no more
2170 .sp
2171 the three matched strings are
2172 .sp
2173 <something>
2174 <something> <something else>
2175 <something> <something else> <something further>
2176 .sp
2177 On success, the yield of the function is a number greater than zero, which is
2178 the number of matched substrings. The substrings themselves are returned in
2179 \fIovector\fP. Each string uses two elements; the first is the offset to the
2180 start, and the second is the offset to the end. In fact, all the strings have
2181 the same start offset. (Space could have been saved by giving this only once,
2182 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2183 returns data, even though the meaning of the strings is different.)
2184 .P
2185 The strings are returned in reverse order of length; that is, the longest
2186 matching string is given first. If there were too many matches to fit into
2187 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2188 the longest matches.
2189 .
2190 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2191 .rs
2192 .sp
2193 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2194 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2195 described
2196 .\" HTML <a href="#errorlist">
2197 .\" </a>
2198 above.
2199 .\"
2200 There are in addition the following errors that are specific to
2201 \fBpcre_dfa_exec()\fP:
2202 .sp
2203 PCRE_ERROR_DFA_UITEM (-16)
2204 .sp
2205 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2206 that it does not support, for instance, the use of \eC or a back reference.
2207 .sp
2208 PCRE_ERROR_DFA_UCOND (-17)
2209 .sp
2210 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2211 uses a back reference for the condition, or a test for recursion in a specific
2212 group. These are not supported.
2213 .sp
2214 PCRE_ERROR_DFA_UMLIMIT (-18)
2215 .sp
2216 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2217 block that contains a setting of the \fImatch_limit\fP field. This is not
2218 supported (it is meaningless).
2219 .sp
2220 PCRE_ERROR_DFA_WSSIZE (-19)
2221 .sp
2222 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2223 \fIworkspace\fP vector.
2224 .sp
2225 PCRE_ERROR_DFA_RECURSE (-20)
2226 .sp
2227 When a recursive subpattern is processed, the matching function calls itself
2228 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2229 error is given if the output vector is not large enough. This should be
2230 extremely rare, as a vector of size 1000 is used.
2231 .
2232 .
2233 .SH "SEE ALSO"
2234 .rs
2235 .sp
2236 \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
2237 \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
2238 \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
2239 .
2240 .
2241 .SH AUTHOR
2242 .rs
2243 .sp
2244 .nf
2245 Philip Hazel
2246 University Computing Service
2247 Cambridge CB2 3QH, England.
2248 .fi
2249 .
2250 .
2251 .SH REVISION
2252 .rs
2253 .sp
2254 .nf
2255 Last updated: 13 November 2010
2256 Copyright (c) 1997-2010 University of Cambridge.
2257 .fi

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5