ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log

Revision 659 - (show annotations)
Tue Aug 16 09:48:29 2011 UTC (9 years, 2 months ago) by ph10
File size: 103320 byte(s)
Formatting tidies for 8.13 release.
3 PCRE - Perl-compatible regular expressions
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 .
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>
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 .
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 .
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 .
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
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
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
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
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
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
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
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
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
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 .
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,
432 PCRE_NO_START_OPT options can be set at the time of matching as well as at
433 compile time.
434 .P
435 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
436 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
437 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
438 error message. This is a static string that is part of the library. You must
439 not try to free it. Normally, the offset from the start of the pattern to the
440 byte that was being processed when the error was discovered is placed in the
441 variable pointed to by \fIerroffset\fP, which must not be NULL (if it is, an
442 immediate error is given). However, for an invalid UTF-8 string, the offset is
443 that of the first byte of the failing character. Also, some errors are not
444 detected until checks are carried out when the whole pattern has been scanned;
445 in these cases the offset passed back is the length of the pattern.
446 .P
447 Note that the offset is in bytes, not characters, even in UTF-8 mode. It may
448 sometimes point into the middle of a UTF-8 character.
449 .P
450 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
451 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
452 returned via this argument in the event of an error. This is in addition to the
453 textual error message. Error codes and messages are listed below.
454 .P
455 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
456 character tables that are built when PCRE is compiled, using the default C
457 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
458 call to \fBpcre_maketables()\fP. This value is stored with the compiled
459 pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
460 passed to it. For more discussion, see the section on locale support below.
461 .P
462 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
463 .sp
464 pcre *re;
465 const char *error;
466 int erroffset;
467 re = pcre_compile(
468 "^A.*Z", /* the pattern */
469 0, /* default options */
470 &error, /* for error message */
471 &erroffset, /* for error offset */
472 NULL); /* use default character tables */
473 .sp
474 The following names for option bits are defined in the \fBpcre.h\fP header
475 file:
476 .sp
478 .sp
479 If this bit is set, the pattern is forced to be "anchored", that is, it is
480 constrained to match only at the first matching point in the string that is
481 being searched (the "subject string"). This effect can also be achieved by
482 appropriate constructs in the pattern itself, which is the only way to do it in
483 Perl.
484 .sp
486 .sp
487 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
488 all with number 255, before each pattern item. For discussion of the callout
489 facility, see the
490 .\" HREF
491 \fBpcrecallout\fP
492 .\"
493 documentation.
494 .sp
497 .sp
498 These options (which are mutually exclusive) control what the \eR escape
499 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
500 match any Unicode newline sequence. The default is specified when PCRE is
501 built. It can be overridden from within the pattern, or by setting an option
502 when a compiled pattern is matched.
503 .sp
505 .sp
506 If this bit is set, letters in the pattern match both upper and lower case
507 letters. It is equivalent to Perl's /i option, and it can be changed within a
508 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
509 concept of case for characters whose values are less than 128, so caseless
510 matching is always possible. For characters with higher values, the concept of
511 case is supported if PCRE is compiled with Unicode property support, but not
512 otherwise. If you want to use caseless matching for characters 128 and above,
513 you must ensure that PCRE is compiled with Unicode property support as well as
514 with UTF-8 support.
515 .sp
517 .sp
518 If this bit is set, a dollar metacharacter in the pattern matches only at the
519 end of the subject string. Without this option, a dollar also matches
520 immediately before a newline at the end of the string (but not before any other
521 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
522 There is no equivalent to this option in Perl, and no way to set it within a
523 pattern.
524 .sp
526 .sp
527 If this bit is set, a dot metacharacter in the pattern matches a character of
528 any value, including one that indicates a newline. However, it only ever
529 matches one character, even if newlines are coded as CRLF. Without this option,
530 a dot does not match when the current position is at a newline. This option is
531 equivalent to Perl's /s option, and it can be changed within a pattern by a
532 (?s) option setting. A negative class such as [^a] always matches newline
533 characters, independent of the setting of this option.
534 .sp
536 .sp
537 If this bit is set, names used to identify capturing subpatterns need not be
538 unique. This can be helpful for certain types of pattern when it is known that
539 only one instance of the named subpattern can ever be matched. There are more
540 details of named subpatterns below; see also the
541 .\" HREF
542 \fBpcrepattern\fP
543 .\"
544 documentation.
545 .sp
547 .sp
548 If this bit is set, whitespace data characters in the pattern are totally
549 ignored except when escaped or inside a character class. Whitespace does not
550 include the VT character (code 11). In addition, characters between an
551 unescaped # outside a character class and the next newline, inclusive, are also
552 ignored. This is equivalent to Perl's /x option, and it can be changed within a
553 pattern by a (?x) option setting.
554 .P
555 Which characters are interpreted as newlines is controlled by the options
556 passed to \fBpcre_compile()\fP or by a special sequence at the start of the
557 pattern, as described in the section entitled
558 .\" HTML <a href="pcrepattern.html#newlines">
559 .\" </a>
560 "Newline conventions"
561 .\"
562 in the \fBpcrepattern\fP documentation. Note that the end of this type of
563 comment is a literal newline sequence in the pattern; escape sequences that
564 happen to represent a newline do not count.
565 .P
566 This option makes it possible to include comments inside complicated patterns.
567 Note, however, that this applies only to data characters. Whitespace characters
568 may never appear within special character sequences in a pattern, for example
569 within the sequence (?( that introduces a conditional subpattern.
570 .sp
572 .sp
573 This option was invented in order to turn on additional functionality of PCRE
574 that is incompatible with Perl, but it is currently of very little use. When
575 set, any backslash in a pattern that is followed by a letter that has no
576 special meaning causes an error, thus reserving these combinations for future
577 expansion. By default, as in Perl, a backslash followed by a letter with no
578 special meaning is treated as a literal. (Perl can, however, be persuaded to
579 give an error for this, by running it with the -w option.) There are at present
580 no other features controlled by this option. It can also be set by a (?X)
581 option setting within a pattern.
582 .sp
584 .sp
585 If this option is set, an unanchored pattern is required to match before or at
586 the first newline in the subject string, though the matched text may continue
587 over the newline.
588 .sp
590 .sp
591 If this option is set, PCRE's behaviour is changed in some ways so that it is
592 compatible with JavaScript rather than Perl. The changes are as follows:
593 .P
594 (1) A lone closing square bracket in a pattern causes a compile-time error,
595 because this is illegal in JavaScript (by default it is treated as a data
596 character). Thus, the pattern AB]CD becomes illegal when this option is set.
597 .P
598 (2) At run time, a back reference to an unset subpattern group matches an empty
599 string (by default this causes the current matching alternative to fail). A
600 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
601 an "a" in the subject), whereas it fails by default, for Perl compatibility.
602 .sp
604 .sp
605 By default, PCRE treats the subject string as consisting of a single line of
606 characters (even if it actually contains newlines). The "start of line"
607 metacharacter (^) matches only at the start of the string, while the "end of
608 line" metacharacter ($) matches only at the end of the string, or before a
609 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
610 Perl.
611 .P
612 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
613 match immediately following or immediately before internal newlines in the
614 subject string, respectively, as well as at the very start and end. This is
615 equivalent to Perl's /m option, and it can be changed within a pattern by a
616 (?m) option setting. If there are no newlines in a subject string, or no
617 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
618 .sp
624 .sp
625 These options override the default newline definition that was chosen when PCRE
626 was built. Setting the first or the second specifies that a newline is
627 indicated by a single character (CR or LF, respectively). Setting
628 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
629 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
630 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
631 that any Unicode newline sequence should be recognized. The Unicode newline
632 sequences are the three just mentioned, plus the single characters VT (vertical
633 tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
634 separator, U+2028), and PS (paragraph separator, U+2029). The last two are
635 recognized only in UTF-8 mode.
636 .P
637 The newline setting in the options word uses three bits that are treated
638 as a number, giving eight possibilities. Currently only six are used (default
639 plus the five values above). This means that if you set more than one newline
640 option, the combination may or may not be sensible. For example,
642 other combinations may yield unused numbers and cause an error.
643 .P
644 The only time that a line break in a pattern is specially recognized when
645 compiling is when PCRE_EXTENDED is set. CR and LF are whitespace characters,
646 and so are ignored in this mode. Also, an unescaped # outside a character class
647 indicates a comment that lasts until after the next line break sequence. In
648 other circumstances, line break sequences in patterns are treated as literal
649 data.
650 .P
651 The newline option that is set at compile time becomes the default that is used
652 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
653 .sp
655 .sp
656 If this option is set, it disables the use of numbered capturing parentheses in
657 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
658 were followed by ?: but named parentheses can still be used for capturing (and
659 they acquire numbers in the usual way). There is no equivalent of this option
660 in Perl.
661 .sp
663 .sp
664 This is an option that acts at matching time; that is, it is really an option
665 for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
666 it is remembered with the compiled pattern and assumed at matching time. For
667 details see the discussion of PCRE_NO_START_OPTIMIZE
668 .\" HTML <a href="#execoptions">
669 .\" </a>
670 below.
671 .\"
672 .sp
674 .sp
675 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
676 \ew, and some of the POSIX character classes. By default, only ASCII characters
677 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
678 classify characters. More details are given in the section on
679 .\" HTML <a href="pcre.html#genericchartypes">
680 .\" </a>
681 generic character types
682 .\"
683 in the
684 .\" HREF
685 \fBpcrepattern\fP
686 .\"
687 page. If you set PCRE_UCP, matching one of the items it affects takes much
688 longer. The option is available only if PCRE has been compiled with Unicode
689 property support.
690 .sp
692 .sp
693 This option inverts the "greediness" of the quantifiers so that they are not
694 greedy by default, but become greedy if followed by "?". It is not compatible
695 with Perl. It can also be set by a (?U) option setting within the pattern.
696 .sp
698 .sp
699 This option causes PCRE to regard both the pattern and the subject as strings
700 of UTF-8 characters instead of single-byte character strings. However, it is
701 available only when PCRE is built to include UTF-8 support. If not, the use
702 of this option provokes an error. Details of how this option changes the
703 behaviour of PCRE are given in the
704 .\" HTML <a href="pcre.html#utf8support">
705 .\" </a>
706 section on UTF-8 support
707 .\"
708 in the main
709 .\" HREF
710 \fBpcre\fP
711 .\"
712 page.
713 .sp
715 .sp
716 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
717 automatically checked. There is a discussion about the
718 .\" HTML <a href="pcre.html#utf8strings">
719 .\" </a>
720 validity of UTF-8 strings
721 .\"
722 in the main
723 .\" HREF
724 \fBpcre\fP
725 .\"
726 page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_compile()\fP
727 returns an error. If you already know that your pattern is valid, and you want
728 to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
729 option. When it is set, the effect of passing an invalid UTF-8 string as a
730 pattern is undefined. It may cause your program to crash. Note that this option
731 can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
732 the UTF-8 validity checking of subject strings.
733 .
734 .
736 .rs
737 .sp
738 The following table lists the error codes than may be returned by
739 \fBpcre_compile2()\fP, along with the error messages that may be returned by
740 both compiling functions. As PCRE has developed, some error codes have fallen
741 out of use. To avoid confusion, they have not been re-used.
742 .sp
743 0 no error
744 1 \e at end of pattern
745 2 \ec at end of pattern
746 3 unrecognized character follows \e
747 4 numbers out of order in {} quantifier
748 5 number too big in {} quantifier
749 6 missing terminating ] for character class
750 7 invalid escape sequence in character class
751 8 range out of order in character class
752 9 nothing to repeat
753 10 [this code is not in use]
754 11 internal error: unexpected repeat
755 12 unrecognized character after (? or (?-
756 13 POSIX named classes are supported only within a class
757 14 missing )
758 15 reference to non-existent subpattern
759 16 erroffset passed as NULL
760 17 unknown option bit(s) set
761 18 missing ) after comment
762 19 [this code is not in use]
763 20 regular expression is too large
764 21 failed to get memory
765 22 unmatched parentheses
766 23 internal error: code overflow
767 24 unrecognized character after (?<
768 25 lookbehind assertion is not fixed length
769 26 malformed number or name after (?(
770 27 conditional group contains more than two branches
771 28 assertion expected after (?(
772 29 (?R or (?[+-]digits must be followed by )
773 30 unknown POSIX class name
774 31 POSIX collating elements are not supported
775 32 this version of PCRE is not compiled with PCRE_UTF8 support
776 33 [this code is not in use]
777 34 character value in \ex{...} sequence is too large
778 35 invalid condition (?(0)
779 36 \eC not allowed in lookbehind assertion
780 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
781 38 number after (?C is > 255
782 39 closing ) for (?C expected
783 40 recursive call could loop indefinitely
784 41 unrecognized character after (?P
785 42 syntax error in subpattern name (missing terminator)
786 43 two named subpatterns have the same name
787 44 invalid UTF-8 string
788 45 support for \eP, \ep, and \eX has not been compiled
789 46 malformed \eP or \ep sequence
790 47 unknown property name after \eP or \ep
791 48 subpattern name is too long (maximum 32 characters)
792 49 too many named subpatterns (maximum 10000)
793 50 [this code is not in use]
794 51 octal value is greater than \e377 (not in UTF-8 mode)
795 52 internal error: overran compiling workspace
796 53 internal error: previously-checked referenced subpattern
797 not found
798 54 DEFINE group contains more than one branch
799 55 repeating a DEFINE group is not allowed
800 56 inconsistent NEWLINE options
801 57 \eg is not followed by a braced, angle-bracketed, or quoted
802 name/number or by a plain number
803 58 a numbered reference must not be zero
804 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
805 60 (*VERB) not recognized
806 61 number is too big
807 62 subpattern name expected
808 63 digit expected after (?+
809 64 ] is an invalid data character in JavaScript compatibility mode
810 65 different names for subpatterns of the same number are
811 not allowed
812 66 (*MARK) must have an argument
813 67 this version of PCRE is not compiled with PCRE_UCP support
814 68 \ec must be followed by an ASCII character
815 69 \ek is not followed by a braced, angle-bracketed, or quoted name
816 .sp
817 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
818 be used if the limits were changed when PCRE was built.
819 .
820 .
821 .\" HTML <a name="studyingapattern"></a>
823 .rs
824 .sp
825 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
826 .ti +5n
827 .B const char **\fIerrptr\fP);
828 .PP
829 If a compiled pattern is going to be used several times, it is worth spending
830 more time analyzing it in order to speed up the time taken for matching. The
831 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
832 argument. If studying the pattern produces additional information that will
833 help speed up matching, \fBpcre_study()\fP returns a pointer to a
834 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
835 results of the study.
836 .P
837 The returned value from \fBpcre_study()\fP can be passed directly to
838 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
839 also contains other fields that can be set by the caller before the block is
840 passed; these are described
841 .\" HTML <a href="#extradata">
842 .\" </a>
843 below
844 .\"
845 in the section on matching a pattern.
846 .P
847 If studying the pattern does not produce any useful information,
848 \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program
849 wants to pass any of the other fields to \fBpcre_exec()\fP or
850 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block.
851 .P
852 The second argument of \fBpcre_study()\fP contains option bits. At present, no
853 options are defined, and this argument should always be zero.
854 .P
855 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
856 studying succeeds (even if no data is returned), the variable it points to is
857 set to NULL. Otherwise it is set to point to a textual error message. This is a
858 static string that is part of the library. You must not try to free it. You
859 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
860 sure that it has run successfully.
861 .P
862 This is a typical call to \fBpcre_study\fP():
863 .sp
864 pcre_extra *pe;
865 pe = pcre_study(
866 re, /* result of pcre_compile() */
867 0, /* no options exist */
868 &error); /* set to NULL or points to a message */
869 .sp
870 Studying a pattern does two things: first, a lower bound for the length of
871 subject string that is needed to match the pattern is computed. This does not
872 mean that there are any strings of that length that match, but it does
873 guarantee that no shorter strings match. The value is used by
874 \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP to avoid wasting time by trying to
875 match strings that are shorter than the lower bound. You can find out the value
876 in a calling program via the \fBpcre_fullinfo()\fP function.
877 .P
878 Studying a pattern is also useful for non-anchored patterns that do not have a
879 single fixed starting character. A bitmap of possible starting bytes is
880 created. This speeds up finding a position in the subject at which to start
881 matching.
882 .P
883 The two optimizations just described can be disabled by setting the
884 PCRE_NO_START_OPTIMIZE option when calling \fBpcre_exec()\fP or
885 \fBpcre_dfa_exec()\fP. You might want to do this if your pattern contains
886 callouts or (*MARK), and you want to make use of these facilities in cases
887 where matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
888 .\" HTML <a href="#execoptions">
889 .\" </a>
890 below.
891 .\"
892 .
893 .
894 .\" HTML <a name="localesupport"></a>
896 .rs
897 .sp
898 PCRE handles caseless matching, and determines whether characters are letters,
899 digits, or whatever, by reference to a set of tables, indexed by character
900 value. When running in UTF-8 mode, this applies only to characters with codes
901 less than 128. By default, higher-valued codes never match escapes such as \ew
902 or \ed, but they can be tested with \ep if PCRE is built with Unicode character
903 property support. Alternatively, the PCRE_UCP option can be set at compile
904 time; this causes \ew and friends to use Unicode property support instead of
905 built-in tables. The use of locales with Unicode is discouraged. If you are
906 handling characters with codes greater than 128, you should either use UTF-8
907 and Unicode, or use locales, but not try to mix the two.
908 .P
909 PCRE contains an internal set of tables that are used when the final argument
910 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
911 Normally, the internal tables recognize only ASCII characters. However, when
912 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
913 default "C" locale of the local system, which may cause them to be different.
914 .P
915 The internal tables can always be overridden by tables supplied by the
916 application that calls PCRE. These may be created in a different locale from
917 the default. As more and more applications change to using Unicode, the need
918 for this locale support is expected to die away.
919 .P
920 External tables are built by calling the \fBpcre_maketables()\fP function,
921 which has no arguments, in the relevant locale. The result can then be passed
922 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
923 example, to build and use tables that are appropriate for the French locale
924 (where accented characters with values greater than 128 are treated as letters),
925 the following code could be used:
926 .sp
927 setlocale(LC_CTYPE, "fr_FR");
928 tables = pcre_maketables();
929 re = pcre_compile(..., tables);
930 .sp
931 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
932 are using Windows, the name for the French locale is "french".
933 .P
934 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
935 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
936 that the memory containing the tables remains available for as long as it is
937 needed.
938 .P
939 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
940 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
941 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
942 pattern, compilation, studying and matching all happen in the same locale, but
943 different patterns can be compiled in different locales.
944 .P
945 It is possible to pass a table pointer or NULL (indicating the use of the
946 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
947 this facility could be used to match a pattern in a different locale from the
948 one in which it was compiled. Passing table pointers at run time is discussed
949 below in the section on matching a pattern.
950 .
951 .
952 .\" HTML <a name="infoaboutpattern"></a>
954 .rs
955 .sp
956 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
957 .ti +5n
958 .B int \fIwhat\fP, void *\fIwhere\fP);
959 .PP
960 The \fBpcre_fullinfo()\fP function returns information about a compiled
961 pattern. It replaces the obsolete \fBpcre_info()\fP function, which is
962 nevertheless retained for backwards compability (and is documented below).
963 .P
964 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
965 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
966 the pattern was not studied. The third argument specifies which piece of
967 information is required, and the fourth argument is a pointer to a variable
968 to receive the data. The yield of the function is zero for success, or one of
969 the following negative numbers:
970 .sp
971 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
972 the argument \fIwhere\fP was NULL
973 PCRE_ERROR_BADMAGIC the "magic number" was not found
974 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
975 .sp
976 The "magic number" is placed at the start of each compiled pattern as an simple
977 check against passing an arbitrary memory pointer. Here is a typical call of
978 \fBpcre_fullinfo()\fP, to obtain the length of the compiled pattern:
979 .sp
980 int rc;
981 size_t length;
982 rc = pcre_fullinfo(
983 re, /* result of pcre_compile() */
984 pe, /* result of pcre_study(), or NULL */
985 PCRE_INFO_SIZE, /* what is required */
986 &length); /* where to put the data */
987 .sp
988 The possible values for the third argument are defined in \fBpcre.h\fP, and are
989 as follows:
990 .sp
992 .sp
993 Return the number of the highest back reference in the pattern. The fourth
994 argument should point to an \fBint\fP variable. Zero is returned if there are
995 no back references.
996 .sp
998 .sp
999 Return the number of capturing subpatterns in the pattern. The fourth argument
1000 should point to an \fBint\fP variable.
1001 .sp
1003 .sp
1004 Return a pointer to the internal default character tables within PCRE. The
1005 fourth argument should point to an \fBunsigned char *\fP variable. This
1006 information call is provided for internal use by the \fBpcre_study()\fP
1007 function. External callers can cause PCRE to use its internal tables by passing
1008 a NULL table pointer.
1009 .sp
1011 .sp
1012 Return information about the first byte of any matched string, for a
1013 non-anchored pattern. The fourth argument should point to an \fBint\fP
1014 variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
1015 still recognized for backwards compatibility.)
1016 .P
1017 If there is a fixed first byte, for example, from a pattern such as
1018 (cat|cow|coyote), its value is returned. Otherwise, if either
1019 .sp
1020 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1021 starts with "^", or
1022 .sp
1023 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1024 (if it were set, the pattern would be anchored),
1025 .sp
1026 -1 is returned, indicating that the pattern matches only at the start of a
1027 subject string or after any newline within the string. Otherwise -2 is
1028 returned. For anchored patterns, -2 is returned.
1029 .sp
1031 .sp
1032 If the pattern was studied, and this resulted in the construction of a 256-bit
1033 table indicating a fixed set of bytes for the first byte in any matching
1034 string, a pointer to the table is returned. Otherwise NULL is returned. The
1035 fourth argument should point to an \fBunsigned char *\fP variable.
1036 .sp
1038 .sp
1039 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1040 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1041 explicit match is either a literal CR or LF character, or \er or \en.
1042 .sp
1044 .sp
1045 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1046 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1047 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1048 .sp
1050 .sp
1051 Return the value of the rightmost literal byte that must exist in any matched
1052 string, other than at its start, if such a byte has been recorded. The fourth
1053 argument should point to an \fBint\fP variable. If there is no such byte, -1 is
1054 returned. For anchored patterns, a last literal byte is recorded only if it
1055 follows something of variable length. For example, for the pattern
1056 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1057 is -1.
1058 .sp
1060 .sp
1061 If the pattern was studied and a minimum length for matching subject strings
1062 was computed, its value is returned. Otherwise the returned value is -1. The
1063 value is a number of characters, not bytes (this may be relevant in UTF-8
1064 mode). The fourth argument should point to an \fBint\fP variable. A
1065 non-negative value is a lower bound to the length of any matching string. There
1066 may not be any strings of that length that do actually match, but every string
1067 that does match is at least that long.
1068 .sp
1072 .sp
1073 PCRE supports the use of named as well as numbered capturing parentheses. The
1074 names are just an additional way of identifying the parentheses, which still
1075 acquire numbers. Several convenience functions such as
1076 \fBpcre_get_named_substring()\fP are provided for extracting captured
1077 substrings by name. It is also possible to extract the data directly, by first
1078 converting the name to a number in order to access the correct pointers in the
1079 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1080 you need to use the name-to-number map, which is described by these three
1081 values.
1082 .P
1083 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1084 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1085 entry; both of these return an \fBint\fP value. The entry size depends on the
1086 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1087 entry of the table (a pointer to \fBchar\fP). The first two bytes of each entry
1088 are the number of the capturing parenthesis, most significant byte first. The
1089 rest of the entry is the corresponding name, zero terminated.
1090 .P
1091 The names are in alphabetical order. Duplicate names may appear if (?| is used
1092 to create multiple groups with the same number, as described in the
1093 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1094 .\" </a>
1095 section on duplicate subpattern numbers
1096 .\"
1097 in the
1098 .\" HREF
1099 \fBpcrepattern\fP
1100 .\"
1101 page. Duplicate names for subpatterns with different numbers are permitted only
1102 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1103 table in the order in which they were found in the pattern. In the absence of
1104 (?| this is the order of increasing number; when (?| is used this is not
1105 necessarily the case because later subpatterns may have lower numbers.
1106 .P
1107 As a simple example of the name/number table, consider the following pattern
1108 (assume PCRE_EXTENDED is set, so white space - including newlines - is
1109 ignored):
1110 .sp
1111 .\" JOIN
1112 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1113 (?<month>\ed\ed) - (?<day>\ed\ed) )
1114 .sp
1115 There are four named subpatterns, so the table has four entries, and each entry
1116 in the table is eight bytes long. The table is as follows, with non-printing
1117 bytes shows in hexadecimal, and undefined bytes shown as ??:
1118 .sp
1119 00 01 d a t e 00 ??
1120 00 05 d a y 00 ?? ??
1121 00 04 m o n t h 00
1122 00 02 y e a r 00 ??
1123 .sp
1124 When writing code to extract data from named subpatterns using the
1125 name-to-number map, remember that the length of the entries is likely to be
1126 different for each compiled pattern.
1127 .sp
1129 .sp
1130 Return 1 if the pattern can be used for partial matching with
1131 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1132 \fBint\fP variable. From release 8.00, this always returns 1, because the
1133 restrictions that previously applied to partial matching have been lifted. The
1134 .\" HREF
1135 \fBpcrepartial\fP
1136 .\"
1137 documentation gives details of partial matching.
1138 .sp
1140 .sp
1141 Return a copy of the options with which the pattern was compiled. The fourth
1142 argument should point to an \fBunsigned long int\fP variable. These option bits
1143 are those specified in the call to \fBpcre_compile()\fP, modified by any
1144 top-level option settings at the start of the pattern itself. In other words,
1145 they are the options that will be in force when matching starts. For example,
1146 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1148 .P
1149 A pattern is automatically anchored by PCRE if all of its top-level
1150 alternatives begin with one of the following:
1151 .sp
1152 ^ unless PCRE_MULTILINE is set
1153 \eA always
1154 \eG always
1155 .\" JOIN
1156 .* if PCRE_DOTALL is set and there are no back
1157 references to the subpattern in which .* appears
1158 .sp
1159 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1160 \fBpcre_fullinfo()\fP.
1161 .sp
1163 .sp
1164 Return the size of the compiled pattern, that is, the value that was passed as
1165 the argument to \fBpcre_malloc()\fP when PCRE was getting memory in which to
1166 place the compiled data. The fourth argument should point to a \fBsize_t\fP
1167 variable.
1168 .sp
1170 .sp
1171 Return the size of the data block pointed to by the \fIstudy_data\fP field in a
1172 \fBpcre_extra\fP block. If \fBpcre_extra\fP is NULL, or there is no study data,
1173 zero is returned. The fourth argument should point to a \fBsize_t\fP variable.
1174 The \fIstudy_data\fP field is set by \fBpcre_study()\fP to record information
1175 that will speed up matching (see the section entitled
1176 .\" HTML <a href="#studyingapattern">
1177 .\" </a>
1178 "Studying a pattern"
1179 .\"
1180 above). The format of the \fIstudy_data\fP block is private, but its length
1181 is made available via this option so that it can be saved and restored (see the
1182 .\" HREF
1183 \fBpcreprecompile\fP
1184 .\"
1185 documentation for details).
1186 .
1187 .
1189 .rs
1190 .sp
1191 .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
1192 .B *\fIfirstcharptr\fP);
1193 .PP
1194 The \fBpcre_info()\fP function is now obsolete because its interface is too
1195 restrictive to return all the available data about a compiled pattern. New
1196 programs should use \fBpcre_fullinfo()\fP instead. The yield of
1197 \fBpcre_info()\fP is the number of capturing subpatterns, or one of the
1198 following negative numbers:
1199 .sp
1200 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1201 PCRE_ERROR_BADMAGIC the "magic number" was not found
1202 .sp
1203 If the \fIoptptr\fP argument is not NULL, a copy of the options with which the
1204 pattern was compiled is placed in the integer it points to (see
1205 PCRE_INFO_OPTIONS above).
1206 .P
1207 If the pattern is not anchored and the \fIfirstcharptr\fP argument is not NULL,
1208 it is used to pass back information about the first character of any matched
1209 string (see PCRE_INFO_FIRSTBYTE above).
1210 .
1211 .
1213 .rs
1214 .sp
1215 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1216 .PP
1217 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1218 data block that contains a compiled pattern. It is provided for the benefit of
1219 applications that operate in an object-oriented manner, where different parts
1220 of the application may be using the same compiled pattern, but you want to free
1221 the block when they are all done.
1222 .P
1223 When a pattern is compiled, the reference count field is initialized to zero.
1224 It is changed only by calling this function, whose action is to add the
1225 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1226 function is the new value. However, the value of the count is constrained to
1227 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1228 it is forced to the appropriate limit value.
1229 .P
1230 Except when it is zero, the reference count is not correctly preserved if a
1231 pattern is compiled on one host and then transferred to a host whose byte-order
1232 is different. (This seems a highly unlikely scenario.)
1233 .
1234 .
1236 .rs
1237 .sp
1238 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1239 .ti +5n
1240 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1241 .ti +5n
1242 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1243 .P
1244 The function \fBpcre_exec()\fP is called to match a subject string against a
1245 compiled pattern, which is passed in the \fIcode\fP argument. If the
1246 pattern was studied, the result of the study should be passed in the
1247 \fIextra\fP argument. This function is the main matching facility of the
1248 library, and it operates in a Perl-like manner. For specialist use there is
1249 also an alternative matching function, which is described
1250 .\" HTML <a href="#dfamatch">
1251 .\" </a>
1252 below
1253 .\"
1254 in the section about the \fBpcre_dfa_exec()\fP function.
1255 .P
1256 In most applications, the pattern will have been compiled (and optionally
1257 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1258 possible to save compiled patterns and study data, and then use them later
1259 in different processes, possibly even on different hosts. For a discussion
1260 about this, see the
1261 .\" HREF
1262 \fBpcreprecompile\fP
1263 .\"
1264 documentation.
1265 .P
1266 Here is an example of a simple call to \fBpcre_exec()\fP:
1267 .sp
1268 int rc;
1269 int ovector[30];
1270 rc = pcre_exec(
1271 re, /* result of pcre_compile() */
1272 NULL, /* we didn't study the pattern */
1273 "some string", /* the subject string */
1274 11, /* the length of the subject string */
1275 0, /* start at offset 0 in the subject */
1276 0, /* default options */
1277 ovector, /* vector of integers for substring information */
1278 30); /* number of elements (NOT size in bytes) */
1279 .
1280 .
1281 .\" HTML <a name="extradata"></a>
1282 .SS "Extra data for \fBpcre_exec()\fR"
1283 .rs
1284 .sp
1285 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1286 data block. The \fBpcre_study()\fP function returns such a block (when it
1287 doesn't return NULL), but you can also create one for yourself, and pass
1288 additional information in it. The \fBpcre_extra\fP block contains the following
1289 fields (not necessarily in this order):
1290 .sp
1291 unsigned long int \fIflags\fP;
1292 void *\fIstudy_data\fP;
1293 unsigned long int \fImatch_limit\fP;
1294 unsigned long int \fImatch_limit_recursion\fP;
1295 void *\fIcallout_data\fP;
1296 const unsigned char *\fItables\fP;
1297 unsigned char **\fImark\fP;
1298 .sp
1299 The \fIflags\fP field is a bitmap that specifies which of the other fields
1300 are set. The flag bits are:
1301 .sp
1308 .sp
1309 Other flag bits should be set to zero. The \fIstudy_data\fP field is set in the
1310 \fBpcre_extra\fP block that is returned by \fBpcre_study()\fP, together with
1311 the appropriate flag bit. You should not set this yourself, but you may add to
1312 the block by setting the other fields and their corresponding flag bits.
1313 .P
1314 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1315 vast amount of resources when running patterns that are not going to match,
1316 but which have a very large number of possibilities in their search trees. The
1317 classic example is a pattern that uses nested unlimited repeats.
1318 .P
1319 Internally, PCRE uses a function called \fBmatch()\fP which it calls repeatedly
1320 (sometimes recursively). The limit set by \fImatch_limit\fP is imposed on the
1321 number of times this function is called during a match, which has the effect of
1322 limiting the amount of backtracking that can take place. For patterns that are
1323 not anchored, the count restarts from zero for each position in the subject
1324 string.
1325 .P
1326 The default value for the limit can be set when PCRE is built; the default
1327 default is 10 million, which handles all but the most extreme cases. You can
1328 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1329 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1330 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1332 .P
1333 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1334 instead of limiting the total number of times that \fBmatch()\fP is called, it
1335 limits the depth of recursion. The recursion depth is a smaller number than the
1336 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1337 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1338 .P
1339 Limiting the recursion depth limits the amount of stack that can be used, or,
1340 when PCRE has been compiled to use memory on the heap instead of the stack, the
1341 amount of heap memory that can be used.
1342 .P
1343 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1344 built; the default default is the same value as the default for
1345 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1346 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1347 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1348 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1349 .P
1350 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1351 and is described in the
1352 .\" HREF
1353 \fBpcrecallout\fP
1354 .\"
1355 documentation.
1356 .P
1357 The \fItables\fP field is used to pass a character tables pointer to
1358 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1359 pattern. A non-NULL value is stored with the compiled pattern only if custom
1360 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1361 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1362 internal tables to be used. This facility is helpful when re-using patterns
1363 that have been saved after compiling with an external set of tables, because
1364 the external tables might be at a different address when \fBpcre_exec()\fP is
1365 called. See the
1366 .\" HREF
1367 \fBpcreprecompile\fP
1368 .\"
1369 documentation for a discussion of saving compiled patterns for later use.
1370 .P
1371 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1372 be set to point to a \fBchar *\fP variable. If the pattern contains any
1373 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1374 a name to pass back, a pointer to the name string (zero terminated) is placed
1375 in the variable pointed to by the \fImark\fP field. The names are within the
1376 compiled pattern; if you wish to retain such a name you must copy it before
1377 freeing the memory of a compiled pattern. If there is no name to pass back, the
1378 variable pointed to by the \fImark\fP field set to NULL. For details of the
1379 backtracking control verbs, see the section entitled
1380 .\" HTML <a href="pcrepattern#backtrackcontrol">
1381 .\" </a>
1382 "Backtracking control"
1383 .\"
1384 in the
1385 .\" HREF
1386 \fBpcrepattern\fP
1387 .\"
1388 documentation.
1389 .
1390 .
1391 .\" HTML <a name="execoptions"></a>
1392 .SS "Option bits for \fBpcre_exec()\fP"
1393 .rs
1394 .sp
1395 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1396 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1400 .sp
1402 .sp
1403 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1404 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1405 to be anchored by virtue of its contents, it cannot be made unachored at
1406 matching time.
1407 .sp
1410 .sp
1411 These options (which are mutually exclusive) control what the \eR escape
1412 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1413 match any Unicode newline sequence. These options override the choice that was
1414 made or defaulted when the pattern was compiled.
1415 .sp
1421 .sp
1422 These options override the newline definition that was chosen or defaulted when
1423 the pattern was compiled. For details, see the description of
1424 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1425 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1426 the way the match position is advanced after a match failure for an unanchored
1427 pattern.
1428 .P
1430 match attempt for an unanchored pattern fails when the current position is at a
1431 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1432 characters, the match position is advanced by two characters instead of one, in
1433 other words, to after the CRLF.
1434 .P
1435 The above rule is a compromise that makes the most common cases work as
1436 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1437 set), it does not match the string "\er\enA" because, after failing at the
1438 start, it skips both the CR and the LF before retrying. However, the pattern
1439 [\er\en]A does match that string, because it contains an explicit CR or LF
1440 reference, and so advances only by one character after the first failure.
1441 .P
1442 An explicit match for CR of LF is either a literal appearance of one of those
1443 characters, or one of the \er or \en escape sequences. Implicit matches such as
1444 [^X] do not count, nor does \es (which includes CR and LF in the characters
1445 that it matches).
1446 .P
1447 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1448 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1449 .sp
1451 .sp
1452 This option specifies that first character of the subject string is not the
1453 beginning of a line, so the circumflex metacharacter should not match before
1454 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1455 never to match. This option affects only the behaviour of the circumflex
1456 metacharacter. It does not affect \eA.
1457 .sp
1459 .sp
1460 This option specifies that the end of the subject string is not the end of a
1461 line, so the dollar metacharacter should not match it nor (except in multiline
1462 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1463 compile time) causes dollar never to match. This option affects only the
1464 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1465 .sp
1467 .sp
1468 An empty string is not considered to be a valid match if this option is set. If
1469 there are alternatives in the pattern, they are tried. If all the alternatives
1470 match the empty string, the entire match fails. For example, if the pattern
1471 .sp
1472 a?b?
1473 .sp
1474 is applied to a string not beginning with "a" or "b", it matches an empty
1475 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1476 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1477 .sp
1479 .sp
1480 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1481 the start of the subject is permitted. If the pattern is anchored, such a match
1482 can occur only if the pattern contains \eK.
1483 .P
1484 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1485 does make a special case of a pattern match of the empty string within its
1486 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1487 emulate Perl's behaviour after matching a null string by first trying the match
1488 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1489 if that fails, by advancing the starting offset (see below) and trying an
1490 ordinary match again. There is some code that demonstrates how to do this in
1491 the
1492 .\" HREF
1493 \fBpcredemo\fP
1494 .\"
1495 sample program. In the most general case, you have to check to see if the
1496 newline convention recognizes CRLF as a newline, and if so, and the current
1497 character is CR followed by LF, advance the starting offset by two characters
1498 instead of one.
1499 .sp
1501 .sp
1502 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1503 a match, in order to speed up the process. For example, if it is known that an
1504 unanchored match must start with a specific character, it searches the subject
1505 for that character, and fails immediately if it cannot find it, without
1506 actually running the main matching function. This means that a special item
1507 such as (*COMMIT) at the start of a pattern is not considered until after a
1508 suitable starting point for the match has been found. When callouts or (*MARK)
1509 items are in use, these "start-up" optimizations can cause them to be skipped
1510 if the pattern is never actually used. The start-up optimizations are in effect
1511 a pre-scan of the subject that takes place before the pattern is run.
1512 .P
1513 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1514 causing performance to suffer, but ensuring that in cases where the result is
1515 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1516 are considered at every possible starting position in the subject string. If
1517 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1518 time.
1519 .P
1520 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1521 Consider the pattern
1522 .sp
1524 .sp
1525 When this is compiled, PCRE records the fact that a match must start with the
1526 character "A". Suppose the subject string is "DEFABC". The start-up
1527 optimization scans along the subject, finds "A" and runs the first match
1528 attempt from there. The (*COMMIT) item means that the pattern must match the
1529 current starting position, which in this case, it does. However, if the same
1530 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1531 subject string does not happen. The first match attempt is run starting from
1532 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1533 the overall result is "no match". If the pattern is studied, more start-up
1534 optimizations may be used. For example, a minimum length for the subject may be
1535 recorded. Consider the pattern
1536 .sp
1537 (*MARK:A)(X|Y)
1538 .sp
1539 The minimum length for a match is one character. If the subject is "ABC", there
1540 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1541 If the pattern is studied, the final attempt does not take place, because PCRE
1542 knows that the subject is too short, and so the (*MARK) is never encountered.
1543 In this case, studying the pattern does not affect the overall match result,
1544 which is still "no match", but it does affect the auxiliary information that is
1545 returned.
1546 .sp
1548 .sp
1549 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1550 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1551 The value of \fIstartoffset\fP is also checked to ensure that it points to the
1552 start of a UTF-8 character. There is a discussion about the validity of UTF-8
1553 strings in the
1554 .\" HTML <a href="pcre.html#utf8strings">
1555 .\" </a>
1556 section on UTF-8 support
1557 .\"
1558 in the main
1559 .\" HREF
1560 \fBpcre\fP
1561 .\"
1562 page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_exec()\fP returns
1563 the error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is
1564 a truncated UTF-8 character at the end of the subject, PCRE_ERROR_SHORTUTF8. In
1565 both cases, information about the precise nature of the error may also be
1566 returned (see the descriptions of these errors in the section entitled \fIError
1567 return values from\fP \fBpcre_exec()\fP
1568 .\" HTML <a href="#errorlist">
1569 .\" </a>
1570 below).
1571 .\"
1572 If \fIstartoffset\fP contains a value that does not point to the start of a
1573 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1574 returned.
1575 .P
1576 If you already know that your subject is valid, and you want to skip these
1577 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1578 calling \fBpcre_exec()\fP. You might want to do this for the second and
1579 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1580 all the matches in a single subject string. However, you should be sure that
1581 the value of \fIstartoffset\fP points to the start of a UTF-8 character (or the
1582 end of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1583 invalid UTF-8 string as a subject or an invalid value of \fIstartoffset\fP is
1584 undefined. Your program may crash.
1585 .sp
1588 .sp
1589 These options turn on the partial matching feature. For backwards
1590 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1591 occurs if the end of the subject string is reached successfully, but there are
1592 not enough subject characters to complete the match. If this happens when
1593 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1594 testing any remaining alternatives. Only if no complete match can be found is
1595 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1596 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1597 but only if no complete match can be found.
1598 .P
1599 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1600 partial match is found, \fBpcre_exec()\fP immediately returns
1601 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1602 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1603 important that an alternative complete match.
1604 .P
1605 In both cases, the portion of the string that was inspected when the partial
1606 match was found is set as the first matching string. There is a more detailed
1607 discussion of partial and multi-segment matching, with examples, in the
1608 .\" HREF
1609 \fBpcrepartial\fP
1610 .\"
1611 documentation.
1612 .
1613 .
1614 .SS "The string to be matched by \fBpcre_exec()\fP"
1615 .rs
1616 .sp
1617 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1618 \fIsubject\fP, a length (in bytes) in \fIlength\fP, and a starting byte offset
1619 in \fIstartoffset\fP. If this is negative or greater than the length of the
1620 subject, \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting
1621 offset is zero, the search for a match starts at the beginning of the subject,
1622 and this is by far the most common case. In UTF-8 mode, the byte offset must
1623 point to the start of a UTF-8 character (or the end of the subject). Unlike the
1624 pattern string, the subject may contain binary zero bytes.
1625 .P
1626 A non-zero starting offset is useful when searching for another match in the
1627 same subject by calling \fBpcre_exec()\fP again after a previous success.
1628 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1629 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1630 lookbehind. For example, consider the pattern
1631 .sp
1632 \eBiss\eB
1633 .sp
1634 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1635 the current position in the subject is not a word boundary.) When applied to
1636 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1637 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1638 subject, namely "issipi", it does not match, because \eB is always false at the
1639 start of the subject, which is deemed to be a word boundary. However, if
1640 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1641 set to 4, it finds the second occurrence of "iss" because it is able to look
1642 behind the starting point to discover that it is preceded by a letter.
1643 .P
1644 Finding all the matches in a subject is tricky when the pattern can match an
1645 empty string. It is possible to emulate Perl's /g behaviour by first trying the
1646 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1647 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1648 and trying an ordinary match again. There is some code that demonstrates how to
1649 do this in the
1650 .\" HREF
1651 \fBpcredemo\fP
1652 .\"
1653 sample program. In the most general case, you have to check to see if the
1654 newline convention recognizes CRLF as a newline, and if so, and the current
1655 character is CR followed by LF, advance the starting offset by two characters
1656 instead of one.
1657 .P
1658 If a non-zero starting offset is passed when the pattern is anchored, one
1659 attempt to match at the given offset is made. This can only succeed if the
1660 pattern does not require the match to be at the start of the subject.
1661 .
1662 .
1663 .SS "How \fBpcre_exec()\fP returns captured substrings"
1664 .rs
1665 .sp
1666 In general, a pattern matches a certain portion of the subject, and in
1667 addition, further substrings from the subject may be picked out by parts of the
1668 pattern. Following the usage in Jeffrey Friedl's book, this is called
1669 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1670 a fragment of a pattern that picks out a substring. PCRE supports several other
1671 kinds of parenthesized subpattern that do not cause substrings to be captured.
1672 .P
1673 Captured substrings are returned to the caller via a vector of integers whose
1674 address is passed in \fIovector\fP. The number of elements in the vector is
1675 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
1676 argument is NOT the size of \fIovector\fP in bytes.
1677 .P
1678 The first two-thirds of the vector is used to pass back captured substrings,
1679 each substring using a pair of integers. The remaining third of the vector is
1680 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1681 and is not available for passing back information. The number passed in
1682 \fIovecsize\fP should always be a multiple of three. If it is not, it is
1683 rounded down.
1684 .P
1685 When a match is successful, information about captured substrings is returned
1686 in pairs of integers, starting at the beginning of \fIovector\fP, and
1687 continuing up to two-thirds of its length at the most. The first element of
1688 each pair is set to the byte offset of the first character in a substring, and
1689 the second is set to the byte offset of the first character after the end of a
1690 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
1691 mode. They are not character counts.
1692 .P
1693 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
1694 portion of the subject string matched by the entire pattern. The next pair is
1695 used for the first capturing subpattern, and so on. The value returned by
1696 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
1697 For example, if two substrings have been captured, the returned value is 3. If
1698 there are no capturing subpatterns, the return value from a successful match is
1699 1, indicating that just the first pair of offsets has been set.
1700 .P
1701 If a capturing subpattern is matched repeatedly, it is the last portion of the
1702 string that it matched that is returned.
1703 .P
1704 If the vector is too small to hold all the captured substring offsets, it is
1705 used as far as possible (up to two-thirds of its length), and the function
1706 returns a value of zero. If the substring offsets are not of interest,
1707 \fBpcre_exec()\fP may be called with \fIovector\fP passed as NULL and
1708 \fIovecsize\fP as zero. However, if the pattern contains back references and
1709 the \fIovector\fP is not big enough to remember the related substrings, PCRE
1710 has to get additional memory for use during matching. Thus it is usually
1711 advisable to supply an \fIovector\fP.
1712 .P
1713 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
1714 subpatterns there are in a compiled pattern. The smallest size for
1715 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
1716 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
1717 .P
1718 It is possible for capturing subpattern number \fIn+1\fP to match some part of
1719 the subject when subpattern \fIn\fP has not been used at all. For example, if
1720 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1721 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1722 happens, both values in the offset pairs corresponding to unused subpatterns
1723 are set to -1.
1724 .P
1725 Offset values that correspond to unused subpatterns at the end of the
1726 expression are also set to -1. For example, if the string "abc" is matched
1727 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1728 return from the function is 2, because the highest used capturing subpattern
1729 number is 1, and the offsets for for the second and third capturing subpatterns
1730 (assuming the vector is large enough, of course) are set to -1.
1731 .P
1732 \fBNote\fP: Elements of \fIovector\fP that do not correspond to capturing
1733 parentheses in the pattern are never changed. That is, if a pattern contains
1734 \fIn\fP capturing parentheses, no more than \fIovector[0]\fP to
1735 \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other elements retain
1736 whatever values they previously had.
1737 .P
1738 Some convenience functions are provided for extracting the captured substrings
1739 as separate strings. These are described below.
1740 .
1741 .
1742 .\" HTML <a name="errorlist"></a>
1743 .SS "Error return values from \fBpcre_exec()\fP"
1744 .rs
1745 .sp
1746 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1747 defined in the header file:
1748 .sp
1750 .sp
1751 The subject string did not match the pattern.
1752 .sp
1754 .sp
1755 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
1756 NULL and \fIovecsize\fP was not zero.
1757 .sp
1759 .sp
1760 An unrecognized bit was set in the \fIoptions\fP argument.
1761 .sp
1763 .sp
1764 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1765 the case when it is passed a junk pointer and to detect when a pattern that was
1766 compiled in an environment of one endianness is run in an environment with the
1767 other endianness. This is the error that PCRE gives when the magic number is
1768 not present.
1769 .sp
1771 .sp
1772 While running the pattern match, an unknown item was encountered in the
1773 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1774 of the compiled pattern.
1775 .sp
1777 .sp
1778 If a pattern contains back references, but the \fIovector\fP that is passed to
1779 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
1780 gets a block of memory at the start of matching to use for this purpose. If the
1781 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
1782 automatically freed at the end of matching.
1783 .P
1784 This error is also given if \fBpcre_stack_malloc()\fP fails in
1785 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
1786 \fB--disable-stack-for-recursion\fP.
1787 .sp
1789 .sp
1790 This error is used by the \fBpcre_copy_substring()\fP,
1791 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
1792 below). It is never returned by \fBpcre_exec()\fP.
1793 .sp
1795 .sp
1796 The backtracking limit, as specified by the \fImatch_limit\fP field in a
1797 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
1798 above.
1799 .sp
1801 .sp
1802 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
1803 use by callout functions that want to yield a distinctive error code. See the
1804 .\" HREF
1805 \fBpcrecallout\fP
1806 .\"
1807 documentation for details.
1808 .sp
1810 .sp
1811 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
1812 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
1813 (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
1814 UTF-8 character is placed in the first element, and a reason code is placed in
1815 the second element. The reason codes are listed in the
1816 .\" HTML <a href="#badutf8reasons">
1817 .\" </a>
1818 following section.
1819 .\"
1820 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
1821 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
1822 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
1823 .sp
1825 .sp
1826 The UTF-8 byte sequence that was passed as a subject was checked and found to
1827 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
1828 \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
1829 end of the subject.
1830 .sp
1832 .sp
1833 The subject string did not match, but it did match partially. See the
1834 .\" HREF
1835 \fBpcrepartial\fP
1836 .\"
1837 documentation for details of partial matching.
1838 .sp
1840 .sp
1841 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1842 option was used with a compiled pattern containing items that were not
1843 supported for partial matching. From release 8.00 onwards, there are no
1844 restrictions on partial matching.
1845 .sp
1847 .sp
1848 An unexpected internal error has occurred. This error could be caused by a bug
1849 in PCRE or by overwriting of the compiled pattern.
1850 .sp
1852 .sp
1853 This error is given if the value of the \fIovecsize\fP argument is negative.
1854 .sp
1856 .sp
1857 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
1858 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
1859 description above.
1860 .sp
1862 .sp
1863 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
1864 .sp
1866 .sp
1867 The value of \fIstartoffset\fP was negative or greater than the length of the
1868 subject, that is, the value in \fIlength\fP.
1869 .sp
1871 .sp
1872 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
1873 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
1874 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
1875 fact sufficient to detect this case, but this special error code for
1876 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
1877 retained for backwards compatibility.
1878 .sp
1880 .sp
1881 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
1882 the pattern. Specifically, it means that either the whole pattern or a
1883 subpattern has been called recursively for the second time at the same position
1884 in the subject string. Some simple patterns that might do this are detected and
1885 faulted at compile time, but more complicated cases, in particular mutual
1886 recursions between two different subpatterns, cannot be detected until run
1887 time.
1888 .P
1889 Error numbers -16 to -20 and -22 are not used by \fBpcre_exec()\fP.
1890 .
1891 .
1892 .\" HTML <a name="badutf8reasons"></a>
1893 .SS "Reason codes for invalid UTF-8 strings"
1894 .rs
1895 .sp
1896 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
1897 PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
1898 least 2, the offset of the start of the invalid UTF-8 character is placed in
1899 the first output vector element (\fIovector[0]\fP) and a reason code is placed
1900 in the second element (\fIovector[1]\fP). The reason codes are given names in
1901 the \fBpcre.h\fP header file:
1902 .sp
1908 .sp
1909 The string ends with a truncated UTF-8 character; the code specifies how many
1910 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
1911 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
1912 allows for up to 6 bytes, and this is checked first; hence the possibility of
1913 4 or 5 missing bytes.
1914 .sp
1919 PCRE_UTF8_ERR10
1920 .sp
1921 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
1922 character do not have the binary value 0b10 (that is, either the most
1923 significant bit is 0, or the next bit is 1).
1924 .sp
1925 PCRE_UTF8_ERR11
1926 PCRE_UTF8_ERR12
1927 .sp
1928 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
1929 these code points are excluded by RFC 3629.
1930 .sp
1931 PCRE_UTF8_ERR13
1932 .sp
1933 A 4-byte character has a value greater than 0x10fff; these code points are
1934 excluded by RFC 3629.
1935 .sp
1936 PCRE_UTF8_ERR14
1937 .sp
1938 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
1939 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
1940 from UTF-8.
1941 .sp
1942 PCRE_UTF8_ERR15
1943 PCRE_UTF8_ERR16
1944 PCRE_UTF8_ERR17
1945 PCRE_UTF8_ERR18
1946 PCRE_UTF8_ERR19
1947 .sp
1948 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
1949 value that can be represented by fewer bytes, which is invalid. For example,
1950 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
1951 one byte.
1952 .sp
1953 PCRE_UTF8_ERR20
1954 .sp
1955 The two most significant bits of the first byte of a character have the binary
1956 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
1957 byte can only validly occur as the second or subsequent byte of a multi-byte
1958 character.
1959 .sp
1960 PCRE_UTF8_ERR21
1961 .sp
1962 The first byte of a character has the value 0xfe or 0xff. These values can
1963 never occur in a valid UTF-8 string.
1964 .
1965 .
1967 .rs
1968 .sp
1969 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1970 .ti +5n
1971 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
1972 .ti +5n
1973 .B int \fIbuffersize\fP);
1974 .PP
1975 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1976 .ti +5n
1977 .B int \fIstringcount\fP, int \fIstringnumber\fP,
1978 .ti +5n
1979 .B const char **\fIstringptr\fP);
1980 .PP
1981 .B int pcre_get_substring_list(const char *\fIsubject\fP,
1982 .ti +5n
1983 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
1984 .PP
1985 Captured substrings can be accessed directly by using the offsets returned by
1986 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
1987 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
1988 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
1989 as new, separate, zero-terminated strings. These functions identify substrings
1990 by number. The next section describes functions for extracting named
1991 substrings.
1992 .P
1993 A substring that contains a binary zero is correctly extracted and has a
1994 further zero added on the end, but the result is not, of course, a C string.
1995 However, you can process such a string by referring to the length that is
1996 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
1997 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
1998 for handling strings containing binary zeros, because the end of the final
1999 string is not independently indicated.
2000 .P
2001 The first three arguments are the same for all three of these functions:
2002 \fIsubject\fP is the subject string that has just been successfully matched,
2003 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2004 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2005 captured by the match, including the substring that matched the entire regular
2006 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2007 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2008 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2009 number of elements in the vector divided by three.
2010 .P
2011 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2012 extract a single substring, whose number is given as \fIstringnumber\fP. A
2013 value of zero extracts the substring that matched the entire pattern, whereas
2014 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2015 the string is placed in \fIbuffer\fP, whose length is given by
2016 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2017 obtained via \fBpcre_malloc\fP, and its address is returned via
2018 \fIstringptr\fP. The yield of the function is the length of the string, not
2019 including the terminating zero, or one of these error codes:
2020 .sp
2022 .sp
2023 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2024 memory failed for \fBpcre_get_substring()\fP.
2025 .sp
2027 .sp
2028 There is no substring whose number is \fIstringnumber\fP.
2029 .P
2030 The \fBpcre_get_substring_list()\fP function extracts all available substrings
2031 and builds a list of pointers to them. All this is done in a single block of
2032 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2033 is returned via \fIlistptr\fP, which is also the start of the list of string
2034 pointers. The end of the list is marked by a NULL pointer. The yield of the
2035 function is zero if all went well, or the error code
2036 .sp
2038 .sp
2039 if the attempt to get the memory block failed.
2040 .P
2041 When any of these functions encounter a substring that is unset, which can
2042 happen when capturing subpattern number \fIn+1\fP matches some part of the
2043 subject, but subpattern \fIn\fP has not been used at all, they return an empty
2044 string. This can be distinguished from a genuine zero-length substring by
2045 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2046 substrings.
2047 .P
2048 The two convenience functions \fBpcre_free_substring()\fP and
2049 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2050 a previous call of \fBpcre_get_substring()\fP or
2051 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2052 the function pointed to by \fBpcre_free\fP, which of course could be called
2053 directly from a C program. However, PCRE is used in some situations where it is
2054 linked via a special interface to another programming language that cannot use
2055 \fBpcre_free\fP directly; it is for these cases that the functions are
2056 provided.
2057 .
2058 .
2060 .rs
2061 .sp
2062 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2063 .ti +5n
2064 .B const char *\fIname\fP);
2065 .PP
2066 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2067 .ti +5n
2068 .B const char *\fIsubject\fP, int *\fIovector\fP,
2069 .ti +5n
2070 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2071 .ti +5n
2072 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
2073 .PP
2074 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2075 .ti +5n
2076 .B const char *\fIsubject\fP, int *\fIovector\fP,
2077 .ti +5n
2078 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2079 .ti +5n
2080 .B const char **\fIstringptr\fP);
2081 .PP
2082 To extract a substring by name, you first have to find associated number.
2083 For example, for this pattern
2084 .sp
2085 (a+)b(?<xxx>\ed+)...
2086 .sp
2087 the number of the subpattern called "xxx" is 2. If the name is known to be
2088 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2089 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2090 pattern, and the second is the name. The yield of the function is the
2091 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2092 that name.
2093 .P
2094 Given the number, you can extract the substring directly, or use one of the
2095 functions described in the previous section. For convenience, there are also
2096 two functions that do the whole job.
2097 .P
2098 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2099 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2100 functions that extract by number. As these are described in the previous
2101 section, they are not re-described here. There are just two differences:
2102 .P
2103 First, instead of a substring number, a substring name is given. Second, there
2104 is an extra argument, given at the start, which is a pointer to the compiled
2105 pattern. This is needed in order to gain access to the name-to-number
2106 translation table.
2107 .P
2108 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2109 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2110 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2111 the behaviour may not be what you want (see the next section).
2112 .P
2113 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2114 subpatterns with the same number, as described in the
2115 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2116 .\" </a>
2117 section on duplicate subpattern numbers
2118 .\"
2119 in the
2120 .\" HREF
2121 \fBpcrepattern\fP
2122 .\"
2123 page, you cannot use names to distinguish the different subpatterns, because
2124 names are not included in the compiled code. The matching process uses only
2125 numbers. For this reason, the use of different names for subpatterns of the
2126 same number causes an error at compile time.
2127 .
2129 .rs
2130 .sp
2131 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2132 .ti +5n
2133 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
2134 .PP
2135 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2136 are not required to be unique. (Duplicate names are always allowed for
2137 subpatterns with the same number, created by using the (?| feature. Indeed, if
2138 such subpatterns are named, they are required to use the same names.)
2139 .P
2140 Normally, patterns with duplicate names are such that in any one match, only
2141 one of the named subpatterns participates. An example is shown in the
2142 .\" HREF
2143 \fBpcrepattern\fP
2144 .\"
2145 documentation.
2146 .P
2147 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2148 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2149 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2150 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2151 returns one of the numbers that are associated with the name, but it is not
2152 defined which it is.
2153 .P
2154 If you want to get full details of all captured substrings for a given name,
2155 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2156 argument is the compiled pattern, and the second is the name. The third and
2157 fourth are pointers to variables which are updated by the function. After it
2158 has run, they point to the first and last entries in the name-to-number table
2159 for the given name. The function itself returns the length of each entry, or
2160 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2161 described above in the section entitled \fIInformation about a pattern\fP
2162 .\" HTML <a href="#infoaboutpattern">
2163 .\" </a>
2164 above.
2165 .\"
2166 Given all the relevant entries for the name, you can extract each of their
2167 numbers, and hence the captured data, if any.
2168 .
2169 .
2171 .rs
2172 .sp
2173 The traditional matching function uses a similar algorithm to Perl, which stops
2174 when it finds the first match, starting at a given point in the subject. If you
2175 want to find all possible matches, or the longest possible match, consider
2176 using the alternative matching function (see below) instead. If you cannot use
2177 the alternative function, but still need to find all possible matches, you
2178 can kludge it up by making use of the callout facility, which is described in
2179 the
2180 .\" HREF
2181 \fBpcrecallout\fP
2182 .\"
2183 documentation.
2184 .P
2185 What you have to do is to insert a callout right at the end of the pattern.
2186 When your callout function is called, extract and save the current matched
2187 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2188 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2189 will yield PCRE_ERROR_NOMATCH.
2190 .
2191 .
2192 .\" HTML <a name="dfamatch"></a>
2194 .rs
2195 .sp
2196 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2197 .ti +5n
2198 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2199 .ti +5n
2200 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2201 .ti +5n
2202 .B int *\fIworkspace\fP, int \fIwscount\fP);
2203 .P
2204 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2205 a compiled pattern, using a matching algorithm that scans the subject string
2206 just once, and does not backtrack. This has different characteristics to the
2207 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2208 patterns are not supported. Nevertheless, there are times when this kind of
2209 matching can be useful. For a discussion of the two matching algorithms, and a
2210 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2211 .\" HREF
2212 \fBpcrematching\fP
2213 .\"
2214 documentation.
2215 .P
2216 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2217 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2218 different way, and this is described below. The other common arguments are used
2219 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2220 here.
2221 .P
2222 The two additional arguments provide workspace for the function. The workspace
2223 vector should contain at least 20 elements. It is used for keeping track of
2224 multiple paths through the pattern tree. More workspace will be needed for
2225 patterns and subjects where there are a lot of potential matches.
2226 .P
2227 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2228 .sp
2229 int rc;
2230 int ovector[10];
2231 int wspace[20];
2232 rc = pcre_dfa_exec(
2233 re, /* result of pcre_compile() */
2234 NULL, /* we didn't study the pattern */
2235 "some string", /* the subject string */
2236 11, /* the length of the subject string */
2237 0, /* start at offset 0 in the subject */
2238 0, /* default options */
2239 ovector, /* vector of integers for substring information */
2240 10, /* number of elements (NOT size in bytes) */
2241 wspace, /* working space vector */
2242 20); /* number of elements (NOT size in bytes) */
2243 .
2244 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2245 .rs
2246 .sp
2247 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2248 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2252 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2253 so their description is not repeated here.
2254 .sp
2257 .sp
2258 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2259 details are slightly different. When PCRE_PARTIAL_HARD is set for
2260 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2261 is reached and there is still at least one matching possibility that requires
2262 additional characters. This happens even if some complete matches have also
2263 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2264 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2265 there have been no complete matches, but there is still at least one matching
2266 possibility. The portion of the string that was inspected when the longest
2267 partial match was found is set as the first matching string in both cases.
2268 There is a more detailed discussion of partial and multi-segment matching, with
2269 examples, in the
2270 .\" HREF
2271 \fBpcrepartial\fP
2272 .\"
2273 documentation.
2274 .sp
2276 .sp
2277 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2278 soon as it has found one match. Because of the way the alternative algorithm
2279 works, this is necessarily the shortest possible match at the first possible
2280 matching point in the subject string.
2281 .sp
2283 .sp
2284 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2285 again, with additional subject characters, and have it continue with the same
2286 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2287 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2288 before because data about the match so far is left in them after a partial
2289 match. There is more discussion of this facility in the
2290 .\" HREF
2291 \fBpcrepartial\fP
2292 .\"
2293 documentation.
2294 .
2295 .
2296 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2297 .rs
2298 .sp
2299 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2300 substring in the subject. Note, however, that all the matches from one run of
2301 the function start at the same point in the subject. The shorter matches are
2302 all initial substrings of the longer matches. For example, if the pattern
2303 .sp
2304 <.*>
2305 .sp
2306 is matched against the string
2307 .sp
2308 This is <something> <something else> <something further> no more
2309 .sp
2310 the three matched strings are
2311 .sp
2312 <something>
2313 <something> <something else>
2314 <something> <something else> <something further>
2315 .sp
2316 On success, the yield of the function is a number greater than zero, which is
2317 the number of matched substrings. The substrings themselves are returned in
2318 \fIovector\fP. Each string uses two elements; the first is the offset to the
2319 start, and the second is the offset to the end. In fact, all the strings have
2320 the same start offset. (Space could have been saved by giving this only once,
2321 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2322 returns data, even though the meaning of the strings is different.)
2323 .P
2324 The strings are returned in reverse order of length; that is, the longest
2325 matching string is given first. If there were too many matches to fit into
2326 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2327 the longest matches.
2328 .
2329 .
2330 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2331 .rs
2332 .sp
2333 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2334 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2335 described
2336 .\" HTML <a href="#errorlist">
2337 .\" </a>
2338 above.
2339 .\"
2340 There are in addition the following errors that are specific to
2341 \fBpcre_dfa_exec()\fP:
2342 .sp
2344 .sp
2345 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2346 that it does not support, for instance, the use of \eC or a back reference.
2347 .sp
2349 .sp
2350 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2351 uses a back reference for the condition, or a test for recursion in a specific
2352 group. These are not supported.
2353 .sp
2355 .sp
2356 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2357 block that contains a setting of the \fImatch_limit\fP field. This is not
2358 supported (it is meaningless).
2359 .sp
2361 .sp
2362 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2363 \fIworkspace\fP vector.
2364 .sp
2366 .sp
2367 When a recursive subpattern is processed, the matching function calls itself
2368 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2369 error is given if the output vector is not large enough. This should be
2370 extremely rare, as a vector of size 1000 is used.
2371 .
2372 .
2373 .SH "SEE ALSO"
2374 .rs
2375 .sp
2376 \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
2377 \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
2378 \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
2379 .
2380 .
2382 .rs
2383 .sp
2384 .nf
2385 Philip Hazel
2386 University Computing Service
2387 Cambridge CB2 3QH, England.
2388 .fi
2389 .
2390 .
2392 .rs
2393 .sp
2394 .nf
2395 Last updated: 13 August 2011
2396 Copyright (c) 1997-2011 University of Cambridge.
2397 .fi


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

  ViewVC Help
Powered by ViewVC 1.1.5