ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log

Revision 142 - (show annotations)
Fri Mar 30 15:55:18 2007 UTC (14 years, 6 months ago) by ph10
File size: 79611 byte(s)
Trailing spaces.
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 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
136 and \fBpcre_exec()\fP are used for compiling and matching regular expressions
137 in a Perl-compatible manner. A sample program that demonstrates the simplest
138 way of using them is provided in the file called \fIpcredemo.c\fP in the source
139 distribution. The
140 .\" HREF
141 \fBpcresample\fP
142 .\"
143 documentation describes how to run it.
144 .P
145 A second matching function, \fBpcre_dfa_exec()\fP, which is not
146 Perl-compatible, is also provided. This uses a different algorithm for the
147 matching. The alternative algorithm finds all possible matches (at a given
148 point in the subject), and scans the subject just once. However, this algorithm
149 does not return captured substrings. A description of the two matching
150 algorithms and their advantages and disadvantages is given in the
151 .\" HREF
152 \fBpcrematching\fP
153 .\"
154 documentation.
155 .P
156 In addition to the main compiling and matching functions, there are convenience
157 functions for extracting captured substrings from a subject string that is
158 matched by \fBpcre_exec()\fP. They are:
159 .sp
160 \fBpcre_copy_substring()\fP
161 \fBpcre_copy_named_substring()\fP
162 \fBpcre_get_substring()\fP
163 \fBpcre_get_named_substring()\fP
164 \fBpcre_get_substring_list()\fP
165 \fBpcre_get_stringnumber()\fP
166 \fBpcre_get_stringtable_entries()\fP
167 .sp
168 \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
169 provided, to free the memory used for extracted strings.
170 .P
171 The function \fBpcre_maketables()\fP is used to build a set of character tables
172 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
173 or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
174 specialist use. Most commonly, no special tables are passed, in which case
175 internal tables that are generated when PCRE is built are used.
176 .P
177 The function \fBpcre_fullinfo()\fP is used to find out information about a
178 compiled pattern; \fBpcre_info()\fP is an obsolete version that returns only
179 some of the available information, but is retained for backwards compatibility.
180 The function \fBpcre_version()\fP returns a pointer to a string containing the
181 version of PCRE and its date of release.
182 .P
183 The function \fBpcre_refcount()\fP maintains a reference count in a data block
184 containing a compiled pattern. This is provided for the benefit of
185 object-oriented applications.
186 .P
187 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
188 the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
189 respectively. PCRE calls the memory management functions via these variables,
190 so a calling program can replace them if it wishes to intercept the calls. This
191 should be done before calling any PCRE functions.
192 .P
193 The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
194 indirections to memory management functions. These special functions are used
195 only when PCRE is compiled to use the heap for remembering data, instead of
196 recursive function calls, when running the \fBpcre_exec()\fP function. See the
197 .\" HREF
198 \fBpcrebuild\fP
199 .\"
200 documentation for details of how to do this. It is a non-standard way of
201 building PCRE, for use in environments that have limited stacks. Because of the
202 greater use of memory management, it runs more slowly. Separate functions are
203 provided so that special-purpose external code can be used for this case. When
204 used, these functions are always called in a stack-like manner (last obtained,
205 first freed), and always for memory blocks of the same size. There is a
206 discussion about PCRE's stack usage in the
207 .\" HREF
208 \fBpcrestack\fP
209 .\"
210 documentation.
211 .P
212 The global variable \fBpcre_callout\fP initially contains NULL. It can be set
213 by the caller to a "callout" function, which PCRE will then call at specified
214 points during a matching operation. Details are given in the
215 .\" HREF
216 \fBpcrecallout\fP
217 .\"
218 documentation.
219 .
220 .
222 .rs
223 .sp
224 PCRE supports four different conventions for indicating line breaks in
225 strings: a single CR (carriage return) character, a single LF (linefeed)
226 character, the two-character sequence CRLF, or any Unicode newline sequence.
227 The Unicode newline sequences are the three just mentioned, plus the single
228 characters VT (vertical tab, U+000B), FF (formfeed, U+000C), NEL (next line,
229 U+0085), LS (line separator, U+2028), and PS (paragraph separator, U+2029).
230 .P
231 Each of the first three conventions is used by at least one operating system as
232 its standard newline sequence. When PCRE is built, a default can be specified.
233 The default default is LF, which is the Unix standard. When PCRE is run, the
234 default can be overridden, either when a pattern is compiled, or when it is
235 matched.
236 .P
237 In the PCRE documentation the word "newline" is used to mean "the character or
238 pair of characters that indicate a line break". The choice of newline
239 convention affects the handling of the dot, circumflex, and dollar
240 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
241 recognized line ending sequence, the match position advancement for a
242 non-anchored pattern. The choice of newline convention does not affect the
243 interpretation of the \en or \er escape sequences.
244 .
245 .
247 .rs
248 .sp
249 The PCRE functions can be used in multi-threading applications, with the
250 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
251 \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
252 callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
253 .P
254 The compiled form of a regular expression is not altered during matching, so
255 the same compiled pattern can safely be used by several threads at once.
256 .
257 .
259 .rs
260 .sp
261 The compiled form of a regular expression can be saved and re-used at a later
262 time, possibly by a different program, and even on a host other than the one on
263 which it was compiled. Details are given in the
264 .\" HREF
265 \fBpcreprecompile\fP
266 .\"
267 documentation.
268 .
269 .
271 .rs
272 .sp
273 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
274 .PP
275 The function \fBpcre_config()\fP makes it possible for a PCRE client to
276 discover which optional features have been compiled into the PCRE library. The
277 .\" HREF
278 \fBpcrebuild\fP
279 .\"
280 documentation has more details about these optional features.
281 .P
282 The first argument for \fBpcre_config()\fP is an integer, specifying which
283 information is required; the second argument is a pointer to a variable into
284 which the information is placed. The following information is available:
285 .sp
287 .sp
288 The output is an integer that is set to one if UTF-8 support is available;
289 otherwise it is set to zero.
290 .sp
292 .sp
293 The output is an integer that is set to one if support for Unicode character
294 properties is available; otherwise it is set to zero.
295 .sp
297 .sp
298 The output is an integer whose value specifies the default character sequence
299 that is recognized as meaning "newline". The four values that are supported
300 are: 10 for LF, 13 for CR, 3338 for CRLF, and -1 for ANY. The default should
301 normally be the standard sequence for your operating system.
302 .sp
304 .sp
305 The output is an integer that contains the number of bytes used for internal
306 linkage in compiled regular expressions. The value is 2, 3, or 4. Larger values
307 allow larger regular expressions to be compiled, at the expense of slower
308 matching. The default value of 2 is sufficient for all but the most massive
309 patterns, since it allows the compiled pattern to be up to 64K in size.
310 .sp
312 .sp
313 The output is an integer that contains the threshold above which the POSIX
314 interface uses \fBmalloc()\fP for output vectors. Further details are given in
315 the
316 .\" HREF
317 \fBpcreposix\fP
318 .\"
319 documentation.
320 .sp
322 .sp
323 The output is an integer that gives the default limit for the number of
324 internal matching function calls in a \fBpcre_exec()\fP execution. Further
325 details are given with \fBpcre_exec()\fP below.
326 .sp
328 .sp
329 The output is an integer that gives the default limit for the depth of
330 recursion when calling the internal matching function in a \fBpcre_exec()\fP
331 execution. Further details are given with \fBpcre_exec()\fP below.
332 .sp
334 .sp
335 The output is an integer that is set to one if internal recursion when running
336 \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
337 to remember their state. This is the usual way that PCRE is compiled. The
338 output is zero if PCRE was compiled to use blocks of data on the heap instead
339 of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
340 \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
341 avoiding the use of the stack.
342 .
343 .
345 .rs
346 .sp
347 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
348 .ti +5n
349 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
350 .ti +5n
351 .B const unsigned char *\fItableptr\fP);
352 .sp
353 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
354 .ti +5n
355 .B int *\fIerrorcodeptr\fP,
356 .ti +5n
357 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
358 .ti +5n
359 .B const unsigned char *\fItableptr\fP);
360 .P
361 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
362 called to compile a pattern into an internal form. The only difference between
363 the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
364 \fIerrorcodeptr\fP, via which a numerical error code can be returned.
365 .P
366 The pattern is a C string terminated by a binary zero, and is passed in the
367 \fIpattern\fP argument. A pointer to a single block of memory that is obtained
368 via \fBpcre_malloc\fP is returned. This contains the compiled code and related
369 data. The \fBpcre\fP type is defined for the returned block; this is a typedef
370 for a structure whose contents are not externally defined. It is up to the
371 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
372 .P
373 Although the compiled code of a PCRE regex is relocatable, that is, it does not
374 depend on memory location, the complete \fBpcre\fP data block is not
375 fully relocatable, because it may contain a copy of the \fItableptr\fP
376 argument, which is an address (see below).
377 .P
378 The \fIoptions\fP argument contains various bit settings that affect the
379 compilation. It should be zero if no options are required. The available
380 options are described below. Some of them, in particular, those that are
381 compatible with Perl, can also be set and unset from within the pattern (see
382 the detailed description in the
383 .\" HREF
384 \fBpcrepattern\fP
385 .\"
386 documentation). For these options, the contents of the \fIoptions\fP argument
387 specifies their initial settings at the start of compilation and execution. The
388 PCRE_ANCHORED and PCRE_NEWLINE_\fIxxx\fP options can be set at the time of
389 matching as well as at compile time.
390 .P
391 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
392 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
393 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
394 error message. This is a static string that is part of the library. You must
395 not try to free it. The offset from the start of the pattern to the character
396 where the error was discovered is placed in the variable pointed to by
397 \fIerroffset\fP, which must not be NULL. If it is, an immediate error is given.
398 .P
399 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
400 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
401 returned via this argument in the event of an error. This is in addition to the
402 textual error message. Error codes and messages are listed below.
403 .P
404 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
405 character tables that are built when PCRE is compiled, using the default C
406 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
407 call to \fBpcre_maketables()\fP. This value is stored with the compiled
408 pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
409 passed to it. For more discussion, see the section on locale support below.
410 .P
411 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
412 .sp
413 pcre *re;
414 const char *error;
415 int erroffset;
416 re = pcre_compile(
417 "^A.*Z", /* the pattern */
418 0, /* default options */
419 &error, /* for error message */
420 &erroffset, /* for error offset */
421 NULL); /* use default character tables */
422 .sp
423 The following names for option bits are defined in the \fBpcre.h\fP header
424 file:
425 .sp
427 .sp
428 If this bit is set, the pattern is forced to be "anchored", that is, it is
429 constrained to match only at the first matching point in the string that is
430 being searched (the "subject string"). This effect can also be achieved by
431 appropriate constructs in the pattern itself, which is the only way to do it in
432 Perl.
433 .sp
435 .sp
436 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
437 all with number 255, before each pattern item. For discussion of the callout
438 facility, see the
439 .\" HREF
440 \fBpcrecallout\fP
441 .\"
442 documentation.
443 .sp
445 .sp
446 If this bit is set, letters in the pattern match both upper and lower case
447 letters. It is equivalent to Perl's /i option, and it can be changed within a
448 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
449 concept of case for characters whose values are less than 128, so caseless
450 matching is always possible. For characters with higher values, the concept of
451 case is supported if PCRE is compiled with Unicode property support, but not
452 otherwise. If you want to use caseless matching for characters 128 and above,
453 you must ensure that PCRE is compiled with Unicode property support as well as
454 with UTF-8 support.
455 .sp
457 .sp
458 If this bit is set, a dollar metacharacter in the pattern matches only at the
459 end of the subject string. Without this option, a dollar also matches
460 immediately before a newline at the end of the string (but not before any other
461 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
462 There is no equivalent to this option in Perl, and no way to set it within a
463 pattern.
464 .sp
466 .sp
467 If this bit is set, a dot metacharater in the pattern matches all characters,
468 including those that indicate newline. Without it, a dot does not match when
469 the current position is at a newline. This option is equivalent to Perl's /s
470 option, and it can be changed within a pattern by a (?s) option setting. A
471 negative class such as [^a] always matches newline characters, independent of
472 the setting of this option.
473 .sp
475 .sp
476 If this bit is set, names used to identify capturing subpatterns need not be
477 unique. This can be helpful for certain types of pattern when it is known that
478 only one instance of the named subpattern can ever be matched. There are more
479 details of named subpatterns below; see also the
480 .\" HREF
481 \fBpcrepattern\fP
482 .\"
483 documentation.
484 .sp
486 .sp
487 If this bit is set, whitespace data characters in the pattern are totally
488 ignored except when escaped or inside a character class. Whitespace does not
489 include the VT character (code 11). In addition, characters between an
490 unescaped # outside a character class and the next newline, inclusive, are also
491 ignored. This is equivalent to Perl's /x option, and it can be changed within a
492 pattern by a (?x) option setting.
493 .P
494 This option makes it possible to include comments inside complicated patterns.
495 Note, however, that this applies only to data characters. Whitespace characters
496 may never appear within special character sequences in a pattern, for example
497 within the sequence (?( which introduces a conditional subpattern.
498 .sp
500 .sp
501 This option was invented in order to turn on additional functionality of PCRE
502 that is incompatible with Perl, but it is currently of very little use. When
503 set, any backslash in a pattern that is followed by a letter that has no
504 special meaning causes an error, thus reserving these combinations for future
505 expansion. By default, as in Perl, a backslash followed by a letter with no
506 special meaning is treated as a literal. (Perl can, however, be persuaded to
507 give a warning for this.) There are at present no other features controlled by
508 this option. It can also be set by a (?X) option setting within a pattern.
509 .sp
511 .sp
512 If this option is set, an unanchored pattern is required to match before or at
513 the first newline in the subject string, though the matched text may continue
514 over the newline.
515 .sp
517 .sp
518 By default, PCRE treats the subject string as consisting of a single line of
519 characters (even if it actually contains newlines). The "start of line"
520 metacharacter (^) matches only at the start of the string, while the "end of
521 line" metacharacter ($) matches only at the end of the string, or before a
522 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
523 Perl.
524 .P
525 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
526 match immediately following or immediately before internal newlines in the
527 subject string, respectively, as well as at the very start and end. This is
528 equivalent to Perl's /m option, and it can be changed within a pattern by a
529 (?m) option setting. If there are no newlines in a subject string, or no
530 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
531 .sp
536 .sp
537 These options override the default newline definition that was chosen when PCRE
538 was built. Setting the first or the second specifies that a newline is
539 indicated by a single character (CR or LF, respectively). Setting
540 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
541 CRLF sequence. Setting PCRE_NEWLINE_ANY specifies that any Unicode newline
542 sequence should be recognized. The Unicode newline sequences are the three just
543 mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
544 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
545 (paragraph separator, U+2029). The last two are recognized only in UTF-8 mode.
546 .P
547 The newline setting in the options word uses three bits that are treated
548 as a number, giving eight possibilities. Currently only five are used (default
549 plus the four values above). This means that if you set more than one newline
550 option, the combination may or may not be sensible. For example,
552 other combinations yield unused numbers and cause an error.
553 .P
554 The only time that a line break is specially recognized when compiling a
555 pattern is if PCRE_EXTENDED is set, and an unescaped # outside a character
556 class is encountered. This indicates a comment that lasts until after the next
557 line break sequence. In other circumstances, line break sequences are treated
558 as literal data, except that in PCRE_EXTENDED mode, both CR and LF are treated
559 as whitespace characters and are therefore ignored.
560 .P
561 The newline option that is set at compile time becomes the default that is used
562 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
563 .sp
565 .sp
566 If this option is set, it disables the use of numbered capturing parentheses in
567 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
568 were followed by ?: but named parentheses can still be used for capturing (and
569 they acquire numbers in the usual way). There is no equivalent of this option
570 in Perl.
571 .sp
573 .sp
574 This option inverts the "greediness" of the quantifiers so that they are not
575 greedy by default, but become greedy if followed by "?". It is not compatible
576 with Perl. It can also be set by a (?U) option setting within the pattern.
577 .sp
579 .sp
580 This option causes PCRE to regard both the pattern and the subject as strings
581 of UTF-8 characters instead of single-byte character strings. However, it is
582 available only when PCRE is built to include UTF-8 support. If not, the use
583 of this option provokes an error. Details of how this option changes the
584 behaviour of PCRE are given in the
585 .\" HTML <a href="pcre.html#utf8support">
586 .\" </a>
587 section on UTF-8 support
588 .\"
589 in the main
590 .\" HREF
591 \fBpcre\fP
592 .\"
593 page.
594 .sp
596 .sp
597 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
598 automatically checked. If an invalid UTF-8 sequence of bytes is found,
599 \fBpcre_compile()\fP returns an error. If you already know that your pattern is
600 valid, and you want to skip this check for performance reasons, you can set the
601 PCRE_NO_UTF8_CHECK option. When it is set, the effect of passing an invalid
602 UTF-8 string as a pattern is undefined. It may cause your program to crash.
603 Note that this option can also be passed to \fBpcre_exec()\fP and
604 \fBpcre_dfa_exec()\fP, to suppress the UTF-8 validity checking of subject
605 strings.
606 .
607 .
609 .rs
610 .sp
611 The following table lists the error codes than may be returned by
612 \fBpcre_compile2()\fP, along with the error messages that may be returned by
613 both compiling functions. As PCRE has developed, some error codes have fallen
614 out of use. To avoid confusion, they have not been re-used.
615 .sp
616 0 no error
617 1 \e at end of pattern
618 2 \ec at end of pattern
619 3 unrecognized character follows \e
620 4 numbers out of order in {} quantifier
621 5 number too big in {} quantifier
622 6 missing terminating ] for character class
623 7 invalid escape sequence in character class
624 8 range out of order in character class
625 9 nothing to repeat
626 10 [this code is not in use]
627 11 internal error: unexpected repeat
628 12 unrecognized character after (?
629 13 POSIX named classes are supported only within a class
630 14 missing )
631 15 reference to non-existent subpattern
632 16 erroffset passed as NULL
633 17 unknown option bit(s) set
634 18 missing ) after comment
635 19 [this code is not in use]
636 20 regular expression too large
637 21 failed to get memory
638 22 unmatched parentheses
639 23 internal error: code overflow
640 24 unrecognized character after (?<
641 25 lookbehind assertion is not fixed length
642 26 malformed number or name after (?(
643 27 conditional group contains more than two branches
644 28 assertion expected after (?(
645 29 (?R or (?digits must be followed by )
646 30 unknown POSIX class name
647 31 POSIX collating elements are not supported
648 32 this version of PCRE is not compiled with PCRE_UTF8 support
649 33 [this code is not in use]
650 34 character value in \ex{...} sequence is too large
651 35 invalid condition (?(0)
652 36 \eC not allowed in lookbehind assertion
653 37 PCRE does not support \eL, \el, \eN, \eU, or \eu
654 38 number after (?C is > 255
655 39 closing ) for (?C expected
656 40 recursive call could loop indefinitely
657 41 unrecognized character after (?P
658 42 syntax error in subpattern name (missing terminator)
659 43 two named subpatterns have the same name
660 44 invalid UTF-8 string
661 45 support for \eP, \ep, and \eX has not been compiled
662 46 malformed \eP or \ep sequence
663 47 unknown property name after \eP or \ep
664 48 subpattern name is too long (maximum 32 characters)
665 49 too many named subpatterns (maximum 10,000)
666 50 repeated subpattern is too long
667 51 octal value is greater than \e377 (not in UTF-8 mode)
668 52 internal error: overran compiling workspace
669 53 internal error: previously-checked referenced subpattern not found
670 54 DEFINE group contains more than one branch
671 55 repeating a DEFINE group is not allowed
672 56 inconsistent NEWLINE options"
673 .
674 .
676 .rs
677 .sp
678 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
679 .ti +5n
680 .B const char **\fIerrptr\fP);
681 .PP
682 If a compiled pattern is going to be used several times, it is worth spending
683 more time analyzing it in order to speed up the time taken for matching. The
684 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
685 argument. If studying the pattern produces additional information that will
686 help speed up matching, \fBpcre_study()\fP returns a pointer to a
687 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
688 results of the study.
689 .P
690 The returned value from \fBpcre_study()\fP can be passed directly to
691 \fBpcre_exec()\fP. However, a \fBpcre_extra\fP block also contains other
692 fields that can be set by the caller before the block is passed; these are
693 described
694 .\" HTML <a href="#extradata">
695 .\" </a>
696 below
697 .\"
698 in the section on matching a pattern.
699 .P
700 If studying the pattern does not produce any additional information
701 \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program
702 wants to pass any of the other fields to \fBpcre_exec()\fP, it must set up its
703 own \fBpcre_extra\fP block.
704 .P
705 The second argument of \fBpcre_study()\fP contains option bits. At present, no
706 options are defined, and this argument should always be zero.
707 .P
708 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
709 studying succeeds (even if no data is returned), the variable it points to is
710 set to NULL. Otherwise it is set to point to a textual error message. This is a
711 static string that is part of the library. You must not try to free it. You
712 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
713 sure that it has run successfully.
714 .P
715 This is a typical call to \fBpcre_study\fP():
716 .sp
717 pcre_extra *pe;
718 pe = pcre_study(
719 re, /* result of pcre_compile() */
720 0, /* no options exist */
721 &error); /* set to NULL or points to a message */
722 .sp
723 At present, studying a pattern is useful only for non-anchored patterns that do
724 not have a single fixed starting character. A bitmap of possible starting
725 bytes is created.
726 .
727 .
728 .\" HTML <a name="localesupport"></a>
730 .rs
731 .sp
732 PCRE handles caseless matching, and determines whether characters are letters,
733 digits, or whatever, by reference to a set of tables, indexed by character
734 value. When running in UTF-8 mode, this applies only to characters with codes
735 less than 128. Higher-valued codes never match escapes such as \ew or \ed, but
736 can be tested with \ep if PCRE is built with Unicode character property
737 support. The use of locales with Unicode is discouraged. If you are handling
738 characters with codes greater than 128, you should either use UTF-8 and
739 Unicode, or use locales, but not try to mix the two.
740 .P
741 PCRE contains an internal set of tables that are used when the final argument
742 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
743 Normally, the internal tables recognize only ASCII characters. However, when
744 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
745 default "C" locale of the local system, which may cause them to be different.
746 .P
747 The internal tables can always be overridden by tables supplied by the
748 application that calls PCRE. These may be created in a different locale from
749 the default. As more and more applications change to using Unicode, the need
750 for this locale support is expected to die away.
751 .P
752 External tables are built by calling the \fBpcre_maketables()\fP function,
753 which has no arguments, in the relevant locale. The result can then be passed
754 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
755 example, to build and use tables that are appropriate for the French locale
756 (where accented characters with values greater than 128 are treated as letters),
757 the following code could be used:
758 .sp
759 setlocale(LC_CTYPE, "fr_FR");
760 tables = pcre_maketables();
761 re = pcre_compile(..., tables);
762 .sp
763 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
764 are using Windows, the name for the French locale is "french".
765 .P
766 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
767 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
768 that the memory containing the tables remains available for as long as it is
769 needed.
770 .P
771 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
772 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
773 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
774 pattern, compilation, studying and matching all happen in the same locale, but
775 different patterns can be compiled in different locales.
776 .P
777 It is possible to pass a table pointer or NULL (indicating the use of the
778 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
779 this facility could be used to match a pattern in a different locale from the
780 one in which it was compiled. Passing table pointers at run time is discussed
781 below in the section on matching a pattern.
782 .
783 .
785 .rs
786 .sp
787 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
788 .ti +5n
789 .B int \fIwhat\fP, void *\fIwhere\fP);
790 .PP
791 The \fBpcre_fullinfo()\fP function returns information about a compiled
792 pattern. It replaces the obsolete \fBpcre_info()\fP function, which is
793 nevertheless retained for backwards compability (and is documented below).
794 .P
795 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
796 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
797 the pattern was not studied. The third argument specifies which piece of
798 information is required, and the fourth argument is a pointer to a variable
799 to receive the data. The yield of the function is zero for success, or one of
800 the following negative numbers:
801 .sp
802 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
803 the argument \fIwhere\fP was NULL
804 PCRE_ERROR_BADMAGIC the "magic number" was not found
805 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
806 .sp
807 The "magic number" is placed at the start of each compiled pattern as an simple
808 check against passing an arbitrary memory pointer. Here is a typical call of
809 \fBpcre_fullinfo()\fP, to obtain the length of the compiled pattern:
810 .sp
811 int rc;
812 size_t length;
813 rc = pcre_fullinfo(
814 re, /* result of pcre_compile() */
815 pe, /* result of pcre_study(), or NULL */
816 PCRE_INFO_SIZE, /* what is required */
817 &length); /* where to put the data */
818 .sp
819 The possible values for the third argument are defined in \fBpcre.h\fP, and are
820 as follows:
821 .sp
823 .sp
824 Return the number of the highest back reference in the pattern. The fourth
825 argument should point to an \fBint\fP variable. Zero is returned if there are
826 no back references.
827 .sp
829 .sp
830 Return the number of capturing subpatterns in the pattern. The fourth argument
831 should point to an \fBint\fP variable.
832 .sp
834 .sp
835 Return a pointer to the internal default character tables within PCRE. The
836 fourth argument should point to an \fBunsigned char *\fP variable. This
837 information call is provided for internal use by the \fBpcre_study()\fP
838 function. External callers can cause PCRE to use its internal tables by passing
839 a NULL table pointer.
840 .sp
842 .sp
843 Return information about the first byte of any matched string, for a
844 non-anchored pattern. The fourth argument should point to an \fBint\fP
845 variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
846 still recognized for backwards compatibility.)
847 .P
848 If there is a fixed first byte, for example, from a pattern such as
849 (cat|cow|coyote), its value is returned. Otherwise, if either
850 .sp
851 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
852 starts with "^", or
853 .sp
854 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
855 (if it were set, the pattern would be anchored),
856 .sp
857 -1 is returned, indicating that the pattern matches only at the start of a
858 subject string or after any newline within the string. Otherwise -2 is
859 returned. For anchored patterns, -2 is returned.
860 .sp
862 .sp
863 If the pattern was studied, and this resulted in the construction of a 256-bit
864 table indicating a fixed set of bytes for the first byte in any matching
865 string, a pointer to the table is returned. Otherwise NULL is returned. The
866 fourth argument should point to an \fBunsigned char *\fP variable.
867 .sp
869 .sp
870 Return the value of the rightmost literal byte that must exist in any matched
871 string, other than at its start, if such a byte has been recorded. The fourth
872 argument should point to an \fBint\fP variable. If there is no such byte, -1 is
873 returned. For anchored patterns, a last literal byte is recorded only if it
874 follows something of variable length. For example, for the pattern
875 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
876 is -1.
877 .sp
881 .sp
882 PCRE supports the use of named as well as numbered capturing parentheses. The
883 names are just an additional way of identifying the parentheses, which still
884 acquire numbers. Several convenience functions such as
885 \fBpcre_get_named_substring()\fP are provided for extracting captured
886 substrings by name. It is also possible to extract the data directly, by first
887 converting the name to a number in order to access the correct pointers in the
888 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
889 you need to use the name-to-number map, which is described by these three
890 values.
891 .P
892 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
893 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
894 entry; both of these return an \fBint\fP value. The entry size depends on the
895 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
896 entry of the table (a pointer to \fBchar\fP). The first two bytes of each entry
897 are the number of the capturing parenthesis, most significant byte first. The
898 rest of the entry is the corresponding name, zero terminated. The names are in
899 alphabetical order. When PCRE_DUPNAMES is set, duplicate names are in order of
900 their parentheses numbers. For example, consider the following pattern (assume
901 PCRE_EXTENDED is set, so white space - including newlines - is ignored):
902 .sp
903 .\" JOIN
904 (?<date> (?<year>(\ed\ed)?\ed\ed) -
905 (?<month>\ed\ed) - (?<day>\ed\ed) )
906 .sp
907 There are four named subpatterns, so the table has four entries, and each entry
908 in the table is eight bytes long. The table is as follows, with non-printing
909 bytes shows in hexadecimal, and undefined bytes shown as ??:
910 .sp
911 00 01 d a t e 00 ??
912 00 05 d a y 00 ?? ??
913 00 04 m o n t h 00
914 00 02 y e a r 00 ??
915 .sp
916 When writing code to extract data from named subpatterns using the
917 name-to-number map, remember that the length of the entries is likely to be
918 different for each compiled pattern.
919 .sp
921 .sp
922 Return a copy of the options with which the pattern was compiled. The fourth
923 argument should point to an \fBunsigned long int\fP variable. These option bits
924 are those specified in the call to \fBpcre_compile()\fP, modified by any
925 top-level option settings within the pattern itself.
926 .P
927 A pattern is automatically anchored by PCRE if all of its top-level
928 alternatives begin with one of the following:
929 .sp
930 ^ unless PCRE_MULTILINE is set
931 \eA always
932 \eG always
933 .\" JOIN
934 .* if PCRE_DOTALL is set and there are no back
935 references to the subpattern in which .* appears
936 .sp
937 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
938 \fBpcre_fullinfo()\fP.
939 .sp
941 .sp
942 Return the size of the compiled pattern, that is, the value that was passed as
943 the argument to \fBpcre_malloc()\fP when PCRE was getting memory in which to
944 place the compiled data. The fourth argument should point to a \fBsize_t\fP
945 variable.
946 .sp
948 .sp
949 Return the size of the data block pointed to by the \fIstudy_data\fP field in
950 a \fBpcre_extra\fP block. That is, it is the value that was passed to
951 \fBpcre_malloc()\fP when PCRE was getting memory into which to place the data
952 created by \fBpcre_study()\fP. The fourth argument should point to a
953 \fBsize_t\fP variable.
954 .
955 .
957 .rs
958 .sp
959 .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
960 .B *\fIfirstcharptr\fP);
961 .PP
962 The \fBpcre_info()\fP function is now obsolete because its interface is too
963 restrictive to return all the available data about a compiled pattern. New
964 programs should use \fBpcre_fullinfo()\fP instead. The yield of
965 \fBpcre_info()\fP is the number of capturing subpatterns, or one of the
966 following negative numbers:
967 .sp
968 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
969 PCRE_ERROR_BADMAGIC the "magic number" was not found
970 .sp
971 If the \fIoptptr\fP argument is not NULL, a copy of the options with which the
972 pattern was compiled is placed in the integer it points to (see
974 .P
975 If the pattern is not anchored and the \fIfirstcharptr\fP argument is not NULL,
976 it is used to pass back information about the first character of any matched
977 string (see PCRE_INFO_FIRSTBYTE above).
978 .
979 .
981 .rs
982 .sp
983 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
984 .PP
985 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
986 data block that contains a compiled pattern. It is provided for the benefit of
987 applications that operate in an object-oriented manner, where different parts
988 of the application may be using the same compiled pattern, but you want to free
989 the block when they are all done.
990 .P
991 When a pattern is compiled, the reference count field is initialized to zero.
992 It is changed only by calling this function, whose action is to add the
993 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
994 function is the new value. However, the value of the count is constrained to
995 lie between 0 and 65535, inclusive. If the new value is outside these limits,
996 it is forced to the appropriate limit value.
997 .P
998 Except when it is zero, the reference count is not correctly preserved if a
999 pattern is compiled on one host and then transferred to a host whose byte-order
1000 is different. (This seems a highly unlikely scenario.)
1001 .
1002 .
1004 .rs
1005 .sp
1006 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1007 .ti +5n
1008 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1009 .ti +5n
1010 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1011 .P
1012 The function \fBpcre_exec()\fP is called to match a subject string against a
1013 compiled pattern, which is passed in the \fIcode\fP argument. If the
1014 pattern has been studied, the result of the study should be passed in the
1015 \fIextra\fP argument. This function is the main matching facility of the
1016 library, and it operates in a Perl-like manner. For specialist use there is
1017 also an alternative matching function, which is described
1018 .\" HTML <a href="#dfamatch">
1019 .\" </a>
1020 below
1021 .\"
1022 in the section about the \fBpcre_dfa_exec()\fP function.
1023 .P
1024 In most applications, the pattern will have been compiled (and optionally
1025 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1026 possible to save compiled patterns and study data, and then use them later
1027 in different processes, possibly even on different hosts. For a discussion
1028 about this, see the
1029 .\" HREF
1030 \fBpcreprecompile\fP
1031 .\"
1032 documentation.
1033 .P
1034 Here is an example of a simple call to \fBpcre_exec()\fP:
1035 .sp
1036 int rc;
1037 int ovector[30];
1038 rc = pcre_exec(
1039 re, /* result of pcre_compile() */
1040 NULL, /* we didn't study the pattern */
1041 "some string", /* the subject string */
1042 11, /* the length of the subject string */
1043 0, /* start at offset 0 in the subject */
1044 0, /* default options */
1045 ovector, /* vector of integers for substring information */
1046 30); /* number of elements (NOT size in bytes) */
1047 .
1048 .\" HTML <a name="extradata"></a>
1049 .SS "Extra data for \fBpcre_exec()\fR"
1050 .rs
1051 .sp
1052 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1053 data block. The \fBpcre_study()\fP function returns such a block (when it
1054 doesn't return NULL), but you can also create one for yourself, and pass
1055 additional information in it. The \fBpcre_extra\fP block contains the following
1056 fields (not necessarily in this order):
1057 .sp
1058 unsigned long int \fIflags\fP;
1059 void *\fIstudy_data\fP;
1060 unsigned long int \fImatch_limit\fP;
1061 unsigned long int \fImatch_limit_recursion\fP;
1062 void *\fIcallout_data\fP;
1063 const unsigned char *\fItables\fP;
1064 .sp
1065 The \fIflags\fP field is a bitmap that specifies which of the other fields
1066 are set. The flag bits are:
1067 .sp
1073 .sp
1074 Other flag bits should be set to zero. The \fIstudy_data\fP field is set in the
1075 \fBpcre_extra\fP block that is returned by \fBpcre_study()\fP, together with
1076 the appropriate flag bit. You should not set this yourself, but you may add to
1077 the block by setting the other fields and their corresponding flag bits.
1078 .P
1079 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1080 vast amount of resources when running patterns that are not going to match,
1081 but which have a very large number of possibilities in their search trees. The
1082 classic example is the use of nested unlimited repeats.
1083 .P
1084 Internally, PCRE uses a function called \fBmatch()\fP which it calls repeatedly
1085 (sometimes recursively). The limit set by \fImatch_limit\fP is imposed on the
1086 number of times this function is called during a match, which has the effect of
1087 limiting the amount of backtracking that can take place. For patterns that are
1088 not anchored, the count restarts from zero for each position in the subject
1089 string.
1090 .P
1091 The default value for the limit can be set when PCRE is built; the default
1092 default is 10 million, which handles all but the most extreme cases. You can
1093 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1094 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1095 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1097 .P
1098 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1099 instead of limiting the total number of times that \fBmatch()\fP is called, it
1100 limits the depth of recursion. The recursion depth is a smaller number than the
1101 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1102 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1103 .P
1104 Limiting the recursion depth limits the amount of stack that can be used, or,
1105 when PCRE has been compiled to use memory on the heap instead of the stack, the
1106 amount of heap memory that can be used.
1107 .P
1108 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1109 built; the default default is the same value as the default for
1110 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1111 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1112 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1113 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1114 .P
1115 The \fIpcre_callout\fP field is used in conjunction with the "callout" feature,
1116 which is described in the
1117 .\" HREF
1118 \fBpcrecallout\fP
1119 .\"
1120 documentation.
1121 .P
1122 The \fItables\fP field is used to pass a character tables pointer to
1123 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1124 pattern. A non-NULL value is stored with the compiled pattern only if custom
1125 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1126 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1127 internal tables to be used. This facility is helpful when re-using patterns
1128 that have been saved after compiling with an external set of tables, because
1129 the external tables might be at a different address when \fBpcre_exec()\fP is
1130 called. See the
1131 .\" HREF
1132 \fBpcreprecompile\fP
1133 .\"
1134 documentation for a discussion of saving compiled patterns for later use.
1135 .
1136 .SS "Option bits for \fBpcre_exec()\fP"
1137 .rs
1138 .sp
1139 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1140 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1142 .sp
1144 .sp
1145 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1146 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1147 to be anchored by virtue of its contents, it cannot be made unachored at
1148 matching time.
1149 .sp
1154 .sp
1155 These options override the newline definition that was chosen or defaulted when
1156 the pattern was compiled. For details, see the description of
1157 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1158 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1159 the way the match position is advanced after a match failure for an unanchored
1160 pattern. When PCRE_NEWLINE_CRLF or PCRE_NEWLINE_ANY is set, and a match attempt
1161 fails when the current position is at a CRLF sequence, the match position is
1162 advanced by two characters instead of one, in other words, to after the CRLF.
1163 .sp
1165 .sp
1166 This option specifies that first character of the subject string is not the
1167 beginning of a line, so the circumflex metacharacter should not match before
1168 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1169 never to match. This option affects only the behaviour of the circumflex
1170 metacharacter. It does not affect \eA.
1171 .sp
1173 .sp
1174 This option specifies that the end of the subject string is not the end of a
1175 line, so the dollar metacharacter should not match it nor (except in multiline
1176 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1177 compile time) causes dollar never to match. This option affects only the
1178 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1179 .sp
1181 .sp
1182 An empty string is not considered to be a valid match if this option is set. If
1183 there are alternatives in the pattern, they are tried. If all the alternatives
1184 match the empty string, the entire match fails. For example, if the pattern
1185 .sp
1186 a?b?
1187 .sp
1188 is applied to a string not beginning with "a" or "b", it matches the empty
1189 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1190 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1191 .P
1192 Perl has no direct equivalent of PCRE_NOTEMPTY, but it does make a special case
1193 of a pattern match of the empty string within its \fBsplit()\fP function, and
1194 when using the /g modifier. It is possible to emulate Perl's behaviour after
1195 matching a null string by first trying the match again at the same offset with
1196 PCRE_NOTEMPTY and PCRE_ANCHORED, and then if that fails by advancing the
1197 starting offset (see below) and trying an ordinary match again. There is some
1198 code that demonstrates how to do this in the \fIpcredemo.c\fP sample program.
1199 .sp
1201 .sp
1202 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1203 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1204 The value of \fIstartoffset\fP is also checked to ensure that it points to the
1205 start of a UTF-8 character. If an invalid UTF-8 sequence of bytes is found,
1206 \fBpcre_exec()\fP returns the error PCRE_ERROR_BADUTF8. If \fIstartoffset\fP
1207 contains an invalid value, PCRE_ERROR_BADUTF8_OFFSET is returned.
1208 .P
1209 If you already know that your subject is valid, and you want to skip these
1210 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1211 calling \fBpcre_exec()\fP. You might want to do this for the second and
1212 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1213 all the matches in a single subject string. However, you should be sure that
1214 the value of \fIstartoffset\fP points to the start of a UTF-8 character. When
1215 PCRE_NO_UTF8_CHECK is set, the effect of passing an invalid UTF-8 string as a
1216 subject, or a value of \fIstartoffset\fP that does not point to the start of a
1217 UTF-8 character, is undefined. Your program may crash.
1218 .sp
1220 .sp
1221 This option turns on the partial matching feature. If the subject string fails
1222 to match the pattern, but at some point during the matching process the end of
1223 the subject was reached (that is, the subject partially matches the pattern and
1224 the failure to match occurred only because there were not enough subject
1225 characters), \fBpcre_exec()\fP returns PCRE_ERROR_PARTIAL instead of
1226 PCRE_ERROR_NOMATCH. When PCRE_PARTIAL is used, there are restrictions on what
1227 may appear in the pattern. These are discussed in the
1228 .\" HREF
1229 \fBpcrepartial\fP
1230 .\"
1231 documentation.
1232 .
1233 .SS "The string to be matched by \fBpcre_exec()\fP"
1234 .rs
1235 .sp
1236 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1237 \fIsubject\fP, a length in \fIlength\fP, and a starting byte offset in
1238 \fIstartoffset\fP. In UTF-8 mode, the byte offset must point to the start of a
1239 UTF-8 character. Unlike the pattern string, the subject may contain binary zero
1240 bytes. When the starting offset is zero, the search for a match starts at the
1241 beginning of the subject, and this is by far the most common case.
1242 .P
1243 A non-zero starting offset is useful when searching for another match in the
1244 same subject by calling \fBpcre_exec()\fP again after a previous success.
1245 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1246 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1247 lookbehind. For example, consider the pattern
1248 .sp
1249 \eBiss\eB
1250 .sp
1251 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1252 the current position in the subject is not a word boundary.) When applied to
1253 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1254 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1255 subject, namely "issipi", it does not match, because \eB is always false at the
1256 start of the subject, which is deemed to be a word boundary. However, if
1257 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1258 set to 4, it finds the second occurrence of "iss" because it is able to look
1259 behind the starting point to discover that it is preceded by a letter.
1260 .P
1261 If a non-zero starting offset is passed when the pattern is anchored, one
1262 attempt to match at the given offset is made. This can only succeed if the
1263 pattern does not require the match to be at the start of the subject.
1264 .
1265 .SS "How \fBpcre_exec()\fP returns captured substrings"
1266 .rs
1267 .sp
1268 In general, a pattern matches a certain portion of the subject, and in
1269 addition, further substrings from the subject may be picked out by parts of the
1270 pattern. Following the usage in Jeffrey Friedl's book, this is called
1271 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1272 a fragment of a pattern that picks out a substring. PCRE supports several other
1273 kinds of parenthesized subpattern that do not cause substrings to be captured.
1274 .P
1275 Captured substrings are returned to the caller via a vector of integer offsets
1276 whose address is passed in \fIovector\fP. The number of elements in the vector
1277 is passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP:
1278 this argument is NOT the size of \fIovector\fP in bytes.
1279 .P
1280 The first two-thirds of the vector is used to pass back captured substrings,
1281 each substring using a pair of integers. The remaining third of the vector is
1282 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1283 and is not available for passing back information. The length passed in
1284 \fIovecsize\fP should always be a multiple of three. If it is not, it is
1285 rounded down.
1286 .P
1287 When a match is successful, information about captured substrings is returned
1288 in pairs of integers, starting at the beginning of \fIovector\fP, and
1289 continuing up to two-thirds of its length at the most. The first element of a
1290 pair is set to the offset of the first character in a substring, and the second
1291 is set to the offset of the first character after the end of a substring. The
1292 first pair, \fIovector[0]\fP and \fIovector[1]\fP, identify the portion of the
1293 subject string matched by the entire pattern. The next pair is used for the
1294 first capturing subpattern, and so on. The value returned by \fBpcre_exec()\fP
1295 is one more than the highest numbered pair that has been set. For example, if
1296 two substrings have been captured, the returned value is 3. If there are no
1297 capturing subpatterns, the return value from a successful match is 1,
1298 indicating that just the first pair of offsets has been set.
1299 .P
1300 If a capturing subpattern is matched repeatedly, it is the last portion of the
1301 string that it matched that is returned.
1302 .P
1303 If the vector is too small to hold all the captured substring offsets, it is
1304 used as far as possible (up to two-thirds of its length), and the function
1305 returns a value of zero. In particular, if the substring offsets are not of
1306 interest, \fBpcre_exec()\fP may be called with \fIovector\fP passed as NULL and
1307 \fIovecsize\fP as zero. However, if the pattern contains back references and
1308 the \fIovector\fP is not big enough to remember the related substrings, PCRE
1309 has to get additional memory for use during matching. Thus it is usually
1310 advisable to supply an \fIovector\fP.
1311 .P
1312 The \fBpcre_info()\fP function can be used to find out how many capturing
1313 subpatterns there are in a compiled pattern. The smallest size for
1314 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
1315 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
1316 .P
1317 It is possible for capturing subpattern number \fIn+1\fP to match some part of
1318 the subject when subpattern \fIn\fP has not been used at all. For example, if
1319 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1320 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1321 happens, both values in the offset pairs corresponding to unused subpatterns
1322 are set to -1.
1323 .P
1324 Offset values that correspond to unused subpatterns at the end of the
1325 expression are also set to -1. For example, if the string "abc" is matched
1326 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1327 return from the function is 2, because the highest used capturing subpattern
1328 number is 1. However, you can refer to the offsets for the second and third
1329 capturing subpatterns if you wish (assuming the vector is large enough, of
1330 course).
1331 .P
1332 Some convenience functions are provided for extracting the captured substrings
1333 as separate strings. These are described below.
1334 .
1335 .\" HTML <a name="errorlist"></a>
1336 .SS "Error return values from \fBpcre_exec()\fP"
1337 .rs
1338 .sp
1339 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1340 defined in the header file:
1341 .sp
1343 .sp
1344 The subject string did not match the pattern.
1345 .sp
1347 .sp
1348 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
1349 NULL and \fIovecsize\fP was not zero.
1350 .sp
1352 .sp
1353 An unrecognized bit was set in the \fIoptions\fP argument.
1354 .sp
1356 .sp
1357 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1358 the case when it is passed a junk pointer and to detect when a pattern that was
1359 compiled in an environment of one endianness is run in an environment with the
1360 other endianness. This is the error that PCRE gives when the magic number is
1361 not present.
1362 .sp
1364 .sp
1365 While running the pattern match, an unknown item was encountered in the
1366 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1367 of the compiled pattern.
1368 .sp
1370 .sp
1371 If a pattern contains back references, but the \fIovector\fP that is passed to
1372 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
1373 gets a block of memory at the start of matching to use for this purpose. If the
1374 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
1375 automatically freed at the end of matching.
1376 .sp
1378 .sp
1379 This error is used by the \fBpcre_copy_substring()\fP,
1380 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
1381 below). It is never returned by \fBpcre_exec()\fP.
1382 .sp
1384 .sp
1385 The backtracking limit, as specified by the \fImatch_limit\fP field in a
1386 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
1387 above.
1388 .sp
1390 .sp
1391 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
1392 use by callout functions that want to yield a distinctive error code. See the
1393 .\" HREF
1394 \fBpcrecallout\fP
1395 .\"
1396 documentation for details.
1397 .sp
1399 .sp
1400 A string that contains an invalid UTF-8 byte sequence was passed as a subject.
1401 .sp
1403 .sp
1404 The UTF-8 byte sequence that was passed as a subject was valid, but the value
1405 of \fIstartoffset\fP did not point to the beginning of a UTF-8 character.
1406 .sp
1408 .sp
1409 The subject string did not match, but it did match partially. See the
1410 .\" HREF
1411 \fBpcrepartial\fP
1412 .\"
1413 documentation for details of partial matching.
1414 .sp
1416 .sp
1417 The PCRE_PARTIAL option was used with a compiled pattern containing items that
1418 are not supported for partial matching. See the
1419 .\" HREF
1420 \fBpcrepartial\fP
1421 .\"
1422 documentation for details of partial matching.
1423 .sp
1425 .sp
1426 An unexpected internal error has occurred. This error could be caused by a bug
1427 in PCRE or by overwriting of the compiled pattern.
1428 .sp
1430 .sp
1431 This error is given if the value of the \fIovecsize\fP argument is negative.
1432 .sp
1434 .sp
1435 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
1436 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
1437 description above.
1438 .sp
1440 .sp
1441 When a group that can match an empty substring is repeated with an unbounded
1442 upper limit, the subject position at the start of the group must be remembered,
1443 so that a test for an empty string can be made when the end of the group is
1444 reached. Some workspace is required for this; if it runs out, this error is
1445 given.
1446 .sp
1448 .sp
1449 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
1450 .P
1451 Error numbers -16 to -20 are not used by \fBpcre_exec()\fP.
1452 .
1453 .
1455 .rs
1456 .sp
1457 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1458 .ti +5n
1459 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
1460 .ti +5n
1461 .B int \fIbuffersize\fP);
1462 .PP
1463 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1464 .ti +5n
1465 .B int \fIstringcount\fP, int \fIstringnumber\fP,
1466 .ti +5n
1467 .B const char **\fIstringptr\fP);
1468 .PP
1469 .B int pcre_get_substring_list(const char *\fIsubject\fP,
1470 .ti +5n
1471 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
1472 .PP
1473 Captured substrings can be accessed directly by using the offsets returned by
1474 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
1475 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
1476 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
1477 as new, separate, zero-terminated strings. These functions identify substrings
1478 by number. The next section describes functions for extracting named
1479 substrings.
1480 .P
1481 A substring that contains a binary zero is correctly extracted and has a
1482 further zero added on the end, but the result is not, of course, a C string.
1483 However, you can process such a string by referring to the length that is
1484 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
1485 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
1486 for handling strings containing binary zeros, because the end of the final
1487 string is not independently indicated.
1488 .P
1489 The first three arguments are the same for all three of these functions:
1490 \fIsubject\fP is the subject string that has just been successfully matched,
1491 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
1492 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
1493 captured by the match, including the substring that matched the entire regular
1494 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
1495 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
1496 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
1497 number of elements in the vector divided by three.
1498 .P
1499 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
1500 extract a single substring, whose number is given as \fIstringnumber\fP. A
1501 value of zero extracts the substring that matched the entire pattern, whereas
1502 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
1503 the string is placed in \fIbuffer\fP, whose length is given by
1504 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
1505 obtained via \fBpcre_malloc\fP, and its address is returned via
1506 \fIstringptr\fP. The yield of the function is the length of the string, not
1507 including the terminating zero, or one of these error codes:
1508 .sp
1510 .sp
1511 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
1512 memory failed for \fBpcre_get_substring()\fP.
1513 .sp
1515 .sp
1516 There is no substring whose number is \fIstringnumber\fP.
1517 .P
1518 The \fBpcre_get_substring_list()\fP function extracts all available substrings
1519 and builds a list of pointers to them. All this is done in a single block of
1520 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
1521 is returned via \fIlistptr\fP, which is also the start of the list of string
1522 pointers. The end of the list is marked by a NULL pointer. The yield of the
1523 function is zero if all went well, or the error code
1524 .sp
1526 .sp
1527 if the attempt to get the memory block failed.
1528 .P
1529 When any of these functions encounter a substring that is unset, which can
1530 happen when capturing subpattern number \fIn+1\fP matches some part of the
1531 subject, but subpattern \fIn\fP has not been used at all, they return an empty
1532 string. This can be distinguished from a genuine zero-length substring by
1533 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
1534 substrings.
1535 .P
1536 The two convenience functions \fBpcre_free_substring()\fP and
1537 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
1538 a previous call of \fBpcre_get_substring()\fP or
1539 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
1540 the function pointed to by \fBpcre_free\fP, which of course could be called
1541 directly from a C program. However, PCRE is used in some situations where it is
1542 linked via a special interface to another programming language that cannot use
1543 \fBpcre_free\fP directly; it is for these cases that the functions are
1544 provided.
1545 .
1546 .
1548 .rs
1549 .sp
1550 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
1551 .ti +5n
1552 .B const char *\fIname\fP);
1553 .PP
1554 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
1555 .ti +5n
1556 .B const char *\fIsubject\fP, int *\fIovector\fP,
1557 .ti +5n
1558 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1559 .ti +5n
1560 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
1561 .PP
1562 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
1563 .ti +5n
1564 .B const char *\fIsubject\fP, int *\fIovector\fP,
1565 .ti +5n
1566 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1567 .ti +5n
1568 .B const char **\fIstringptr\fP);
1569 .PP
1570 To extract a substring by name, you first have to find associated number.
1571 For example, for this pattern
1572 .sp
1573 (a+)b(?<xxx>\ed+)...
1574 .sp
1575 the number of the subpattern called "xxx" is 2. If the name is known to be
1576 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
1577 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
1578 pattern, and the second is the name. The yield of the function is the
1579 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
1580 that name.
1581 .P
1582 Given the number, you can extract the substring directly, or use one of the
1583 functions described in the previous section. For convenience, there are also
1584 two functions that do the whole job.
1585 .P
1586 Most of the arguments of \fBpcre_copy_named_substring()\fP and
1587 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
1588 functions that extract by number. As these are described in the previous
1589 section, they are not re-described here. There are just two differences:
1590 .P
1591 First, instead of a substring number, a substring name is given. Second, there
1592 is an extra argument, given at the start, which is a pointer to the compiled
1593 pattern. This is needed in order to gain access to the name-to-number
1594 translation table.
1595 .P
1596 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
1597 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
1598 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
1599 the behaviour may not be what you want (see the next section).
1600 .
1601 .
1603 .rs
1604 .sp
1605 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
1606 .ti +5n
1607 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
1608 .PP
1609 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
1610 are not required to be unique. Normally, patterns with duplicate names are such
1611 that in any one match, only one of the named subpatterns participates. An
1612 example is shown in the
1613 .\" HREF
1614 \fBpcrepattern\fP
1615 .\"
1616 documentation. When duplicates are present, \fBpcre_copy_named_substring()\fP
1617 and \fBpcre_get_named_substring()\fP return the first substring corresponding
1618 to the given name that is set. If none are set, an empty string is returned.
1619 The \fBpcre_get_stringnumber()\fP function returns one of the numbers that are
1620 associated with the name, but it is not defined which it is.
1621 .sp
1622 If you want to get full details of all captured substrings for a given name,
1623 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
1624 argument is the compiled pattern, and the second is the name. The third and
1625 fourth are pointers to variables which are updated by the function. After it
1626 has run, they point to the first and last entries in the name-to-number table
1627 for the given name. The function itself returns the length of each entry, or
1628 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
1629 described above in the section entitled \fIInformation about a pattern\fP.
1630 Given all the relevant entries for the name, you can extract each of their
1631 numbers, and hence the captured data, if any.
1632 .
1633 .
1635 .rs
1636 .sp
1637 The traditional matching function uses a similar algorithm to Perl, which stops
1638 when it finds the first match, starting at a given point in the subject. If you
1639 want to find all possible matches, or the longest possible match, consider
1640 using the alternative matching function (see below) instead. If you cannot use
1641 the alternative function, but still need to find all possible matches, you
1642 can kludge it up by making use of the callout facility, which is described in
1643 the
1644 .\" HREF
1645 \fBpcrecallout\fP
1646 .\"
1647 documentation.
1648 .P
1649 What you have to do is to insert a callout right at the end of the pattern.
1650 When your callout function is called, extract and save the current matched
1651 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
1652 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
1653 will yield PCRE_ERROR_NOMATCH.
1654 .
1655 .
1656 .\" HTML <a name="dfamatch"></a>
1658 .rs
1659 .sp
1660 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1661 .ti +5n
1662 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1663 .ti +5n
1664 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
1665 .ti +5n
1666 .B int *\fIworkspace\fP, int \fIwscount\fP);
1667 .P
1668 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
1669 a compiled pattern, using a matching algorithm that scans the subject string
1670 just once, and does not backtrack. This has different characteristics to the
1671 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
1672 patterns are not supported. Nevertheless, there are times when this kind of
1673 matching can be useful. For a discussion of the two matching algorithms, see
1674 the
1675 .\" HREF
1676 \fBpcrematching\fP
1677 .\"
1678 documentation.
1679 .P
1680 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
1681 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
1682 different way, and this is described below. The other common arguments are used
1683 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
1684 here.
1685 .P
1686 The two additional arguments provide workspace for the function. The workspace
1687 vector should contain at least 20 elements. It is used for keeping track of
1688 multiple paths through the pattern tree. More workspace will be needed for
1689 patterns and subjects where there are a lot of potential matches.
1690 .P
1691 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
1692 .sp
1693 int rc;
1694 int ovector[10];
1695 int wspace[20];
1696 rc = pcre_dfa_exec(
1697 re, /* result of pcre_compile() */
1698 NULL, /* we didn't study the pattern */
1699 "some string", /* the subject string */
1700 11, /* the length of the subject string */
1701 0, /* start at offset 0 in the subject */
1702 0, /* default options */
1703 ovector, /* vector of integers for substring information */
1704 10, /* number of elements (NOT size in bytes) */
1705 wspace, /* working space vector */
1706 20); /* number of elements (NOT size in bytes) */
1707 .
1708 .SS "Option bits for \fBpcre_dfa_exec()\fP"
1709 .rs
1710 .sp
1711 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
1712 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1714 PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART. All but the last three of these are
1715 the same as for \fBpcre_exec()\fP, so their description is not repeated here.
1716 .sp
1718 .sp
1719 This has the same general effect as it does for \fBpcre_exec()\fP, but the
1720 details are slightly different. When PCRE_PARTIAL is set for
1721 \fBpcre_dfa_exec()\fP, the return code PCRE_ERROR_NOMATCH is converted into
1722 PCRE_ERROR_PARTIAL if the end of the subject is reached, there have been no
1723 complete matches, but there is still at least one matching possibility. The
1724 portion of the string that provided the partial match is set as the first
1725 matching string.
1726 .sp
1728 .sp
1729 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
1730 soon as it has found one match. Because of the way the alternative algorithm
1731 works, this is necessarily the shortest possible match at the first possible
1732 matching point in the subject string.
1733 .sp
1735 .sp
1736 When \fBpcre_dfa_exec()\fP is called with the PCRE_PARTIAL option, and returns
1737 a partial match, it is possible to call it again, with additional subject
1738 characters, and have it continue with the same match. The PCRE_DFA_RESTART
1739 option requests this action; when it is set, the \fIworkspace\fP and
1740 \fIwscount\fP options must reference the same vector as before because data
1741 about the match so far is left in them after a partial match. There is more
1742 discussion of this facility in the
1743 .\" HREF
1744 \fBpcrepartial\fP
1745 .\"
1746 documentation.
1747 .
1748 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
1749 .rs
1750 .sp
1751 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
1752 substring in the subject. Note, however, that all the matches from one run of
1753 the function start at the same point in the subject. The shorter matches are
1754 all initial substrings of the longer matches. For example, if the pattern
1755 .sp
1756 <.*>
1757 .sp
1758 is matched against the string
1759 .sp
1760 This is <something> <something else> <something further> no more
1761 .sp
1762 the three matched strings are
1763 .sp
1764 <something>
1765 <something> <something else>
1766 <something> <something else> <something further>
1767 .sp
1768 On success, the yield of the function is a number greater than zero, which is
1769 the number of matched substrings. The substrings themselves are returned in
1770 \fIovector\fP. Each string uses two elements; the first is the offset to the
1771 start, and the second is the offset to the end. In fact, all the strings have
1772 the same start offset. (Space could have been saved by giving this only once,
1773 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
1774 returns data, even though the meaning of the strings is different.)
1775 .P
1776 The strings are returned in reverse order of length; that is, the longest
1777 matching string is given first. If there were too many matches to fit into
1778 \fIovector\fP, the yield of the function is zero, and the vector is filled with
1779 the longest matches.
1780 .
1781 .SS "Error returns from \fBpcre_dfa_exec()\fP"
1782 .rs
1783 .sp
1784 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
1785 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
1786 described
1787 .\" HTML <a href="#errorlist">
1788 .\" </a>
1789 above.
1790 .\"
1791 There are in addition the following errors that are specific to
1792 \fBpcre_dfa_exec()\fP:
1793 .sp
1795 .sp
1796 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
1797 that it does not support, for instance, the use of \eC or a back reference.
1798 .sp
1800 .sp
1801 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
1802 uses a back reference for the condition, or a test for recursion in a specific
1803 group. These are not supported.
1804 .sp
1806 .sp
1807 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
1808 block that contains a setting of the \fImatch_limit\fP field. This is not
1809 supported (it is meaningless).
1810 .sp
1812 .sp
1813 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
1814 \fIworkspace\fP vector.
1815 .sp
1817 .sp
1818 When a recursive subpattern is processed, the matching function calls itself
1819 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
1820 error is given if the output vector is not large enough. This should be
1821 extremely rare, as a vector of size 1000 is used.
1822 .
1823 .
1824 .SH "SEE ALSO"
1825 .rs
1826 .sp
1827 \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
1828 \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
1829 \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
1830 .
1831 .
1833 .rs
1834 .sp
1835 .nf
1836 Philip Hazel
1837 University Computing Service
1838 Cambridge CB2 3QH, England.
1839 .fi
1840 .
1841 .
1843 .rs
1844 .sp
1845 .nf
1846 Last updated: 06 March 2007
1847 Copyright (c) 1997-2007 University of Cambridge.
1848 .fi


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

  ViewVC Help
Powered by ViewVC 1.1.5