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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1370 - (show annotations)
Wed Oct 9 10:18:26 2013 UTC (6 years, 4 months ago) by ph10
File size: 126712 byte(s)
Error occurred while calculating annotation data.
Add \o{} and tidy up \x{} handling. Minor update to RunTest.
1 .TH PCREAPI 3 "08 October 2013" "PCRE 8.34"
2 .SH NAME
3 PCRE - Perl-compatible regular expressions
4 .sp
5 .B #include <pcre.h>
6 .
7 .
8 .SH "PCRE NATIVE API BASIC FUNCTIONS"
9 .rs
10 .sp
11 .nf
12 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
13 .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
14 .B " const unsigned char *\fItableptr\fP);"
15 .sp
16 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
17 .B " int *\fIerrorcodeptr\fP,"
18 .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
19 .B " const unsigned char *\fItableptr\fP);"
20 .sp
21 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
22 .B " const char **\fIerrptr\fP);"
23 .sp
24 .B void pcre_free_study(pcre_extra *\fIextra\fP);
25 .sp
26 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
27 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
28 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);"
29 .sp
30 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
31 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
32 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
33 .B " int *\fIworkspace\fP, int \fIwscount\fP);"
34 .fi
35 .
36 .
37 .SH "PCRE NATIVE API STRING EXTRACTION FUNCTIONS"
38 .rs
39 .sp
40 .nf
41 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
42 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
43 .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
44 .B " char *\fIbuffer\fP, int \fIbuffersize\fP);"
45 .sp
46 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
47 .B " int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,"
48 .B " int \fIbuffersize\fP);"
49 .sp
50 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
51 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
52 .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
53 .B " const char **\fIstringptr\fP);"
54 .sp
55 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
56 .B " const char *\fIname\fP);"
57 .sp
58 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
59 .B " const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);"
60 .sp
61 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
62 .B " int \fIstringcount\fP, int \fIstringnumber\fP,"
63 .B " const char **\fIstringptr\fP);"
64 .sp
65 .B int pcre_get_substring_list(const char *\fIsubject\fP,
66 .B " int *\fIovector\fP, int \fIstringcount\fP, const char ***\fIlistptr\fP);"
67 .sp
68 .B void pcre_free_substring(const char *\fIstringptr\fP);
69 .sp
70 .B void pcre_free_substring_list(const char **\fIstringptr\fP);
71 .fi
72 .
73 .
74 .SH "PCRE NATIVE API AUXILIARY FUNCTIONS"
75 .rs
76 .sp
77 .nf
78 .B int pcre_jit_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
79 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
80 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
81 .B " pcre_jit_stack *\fIjstack\fP);"
82 .sp
83 .B pcre_jit_stack *pcre_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
84 .sp
85 .B void pcre_jit_stack_free(pcre_jit_stack *\fIstack\fP);
86 .sp
87 .B void pcre_assign_jit_stack(pcre_extra *\fIextra\fP,
88 .B " pcre_jit_callback \fIcallback\fP, void *\fIdata\fP);"
89 .sp
90 .B const unsigned char *pcre_maketables(void);
91 .sp
92 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
93 .B " int \fIwhat\fP, void *\fIwhere\fP);"
94 .sp
95 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
96 .sp
97 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
98 .sp
99 .B const char *pcre_version(void);
100 .sp
101 .B int pcre_pattern_to_host_byte_order(pcre *\fIcode\fP,
102 .B " pcre_extra *\fIextra\fP, const unsigned char *\fItables\fP);"
103 .fi
104 .
105 .
106 .SH "PCRE NATIVE API INDIRECTED FUNCTIONS"
107 .rs
108 .sp
109 .nf
110 .B void *(*pcre_malloc)(size_t);
111 .sp
112 .B void (*pcre_free)(void *);
113 .sp
114 .B void *(*pcre_stack_malloc)(size_t);
115 .sp
116 .B void (*pcre_stack_free)(void *);
117 .sp
118 .B int (*pcre_callout)(pcre_callout_block *);
119 .fi
120 .
121 .
122 .SH "PCRE 8-BIT, 16-BIT, AND 32-BIT LIBRARIES"
123 .rs
124 .sp
125 As well as support for 8-bit character strings, PCRE also supports 16-bit
126 strings (from release 8.30) and 32-bit strings (from release 8.32), by means of
127 two additional libraries. They can be built as well as, or instead of, the
128 8-bit library. To avoid too much complication, this document describes the
129 8-bit versions of the functions, with only occasional references to the 16-bit
130 and 32-bit libraries.
131 .P
132 The 16-bit and 32-bit functions operate in the same way as their 8-bit
133 counterparts; they just use different data types for their arguments and
134 results, and their names start with \fBpcre16_\fP or \fBpcre32_\fP instead of
135 \fBpcre_\fP. For every option that has UTF8 in its name (for example,
136 PCRE_UTF8), there are corresponding 16-bit and 32-bit names with UTF8 replaced
137 by UTF16 or UTF32, respectively. This facility is in fact just cosmetic; the
138 16-bit and 32-bit option names define the same bit values.
139 .P
140 References to bytes and UTF-8 in this document should be read as references to
141 16-bit data units and UTF-16 when using the 16-bit library, or 32-bit data
142 units and UTF-32 when using the 32-bit library, unless specified otherwise.
143 More details of the specific differences for the 16-bit and 32-bit libraries
144 are given in the
145 .\" HREF
146 \fBpcre16\fP
147 .\"
148 and
149 .\" HREF
150 \fBpcre32\fP
151 .\"
152 pages.
153 .
154 .
155 .SH "PCRE API OVERVIEW"
156 .rs
157 .sp
158 PCRE has its own native API, which is described in this document. There are
159 also some wrapper functions (for the 8-bit library only) that correspond to the
160 POSIX regular expression API, but they do not give access to all the
161 functionality. They are described in the
162 .\" HREF
163 \fBpcreposix\fP
164 .\"
165 documentation. Both of these APIs define a set of C function calls. A C++
166 wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
167 documented in the
168 .\" HREF
169 \fBpcrecpp\fP
170 .\"
171 page.
172 .P
173 The native API C function prototypes are defined in the header file
174 \fBpcre.h\fP, and on Unix-like systems the (8-bit) library itself is called
175 \fBlibpcre\fP. It can normally be accessed by adding \fB-lpcre\fP to the
176 command for linking an application that uses PCRE. The header file defines the
177 macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
178 for the library. Applications can use these to include support for different
179 releases of PCRE.
180 .P
181 In a Windows environment, if you want to statically link an application program
182 against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
183 including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
184 \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
185 \fB__declspec(dllimport)\fP, with unwanted results.
186 .P
187 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
188 and \fBpcre_exec()\fP are used for compiling and matching regular expressions
189 in a Perl-compatible manner. A sample program that demonstrates the simplest
190 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
191 source distribution. A listing of this program is given in the
192 .\" HREF
193 \fBpcredemo\fP
194 .\"
195 documentation, and the
196 .\" HREF
197 \fBpcresample\fP
198 .\"
199 documentation describes how to compile and run it.
200 .P
201 Just-in-time compiler support is an optional feature of PCRE that can be built
202 in appropriate hardware environments. It greatly speeds up the matching
203 performance of many patterns. Simple programs can easily request that it be
204 used if available, by setting an option that is ignored when it is not
205 relevant. More complicated programs might need to make use of the functions
206 \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
207 \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
208 .P
209 From release 8.32 there is also a direct interface for JIT execution, which
210 gives improved performance. The JIT-specific functions are discussed in the
211 .\" HREF
212 \fBpcrejit\fP
213 .\"
214 documentation.
215 .P
216 A second matching function, \fBpcre_dfa_exec()\fP, which is not
217 Perl-compatible, is also provided. This uses a different algorithm for the
218 matching. The alternative algorithm finds all possible matches (at a given
219 point in the subject), and scans the subject just once (unless there are
220 lookbehind assertions). However, this algorithm does not return captured
221 substrings. A description of the two matching algorithms and their advantages
222 and disadvantages is given in the
223 .\" HREF
224 \fBpcrematching\fP
225 .\"
226 documentation.
227 .P
228 In addition to the main compiling and matching functions, there are convenience
229 functions for extracting captured substrings from a subject string that is
230 matched by \fBpcre_exec()\fP. They are:
231 .sp
232 \fBpcre_copy_substring()\fP
233 \fBpcre_copy_named_substring()\fP
234 \fBpcre_get_substring()\fP
235 \fBpcre_get_named_substring()\fP
236 \fBpcre_get_substring_list()\fP
237 \fBpcre_get_stringnumber()\fP
238 \fBpcre_get_stringtable_entries()\fP
239 .sp
240 \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
241 provided, to free the memory used for extracted strings.
242 .P
243 The function \fBpcre_maketables()\fP is used to build a set of character tables
244 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
245 or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
246 specialist use. Most commonly, no special tables are passed, in which case
247 internal tables that are generated when PCRE is built are used.
248 .P
249 The function \fBpcre_fullinfo()\fP is used to find out information about a
250 compiled pattern. The function \fBpcre_version()\fP returns a pointer to a
251 string containing the version of PCRE and its date of release.
252 .P
253 The function \fBpcre_refcount()\fP maintains a reference count in a data block
254 containing a compiled pattern. This is provided for the benefit of
255 object-oriented applications.
256 .P
257 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
258 the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
259 respectively. PCRE calls the memory management functions via these variables,
260 so a calling program can replace them if it wishes to intercept the calls. This
261 should be done before calling any PCRE functions.
262 .P
263 The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
264 indirections to memory management functions. These special functions are used
265 only when PCRE is compiled to use the heap for remembering data, instead of
266 recursive function calls, when running the \fBpcre_exec()\fP function. See the
267 .\" HREF
268 \fBpcrebuild\fP
269 .\"
270 documentation for details of how to do this. It is a non-standard way of
271 building PCRE, for use in environments that have limited stacks. Because of the
272 greater use of memory management, it runs more slowly. Separate functions are
273 provided so that special-purpose external code can be used for this case. When
274 used, these functions are always called in a stack-like manner (last obtained,
275 first freed), and always for memory blocks of the same size. There is a
276 discussion about PCRE's stack usage in the
277 .\" HREF
278 \fBpcrestack\fP
279 .\"
280 documentation.
281 .P
282 The global variable \fBpcre_callout\fP initially contains NULL. It can be set
283 by the caller to a "callout" function, which PCRE will then call at specified
284 points during a matching operation. Details are given in the
285 .\" HREF
286 \fBpcrecallout\fP
287 .\"
288 documentation.
289 .
290 .
291 .\" HTML <a name="newlines"></a>
292 .SH NEWLINES
293 .rs
294 .sp
295 PCRE supports five different conventions for indicating line breaks in
296 strings: a single CR (carriage return) character, a single LF (linefeed)
297 character, the two-character sequence CRLF, any of the three preceding, or any
298 Unicode newline sequence. The Unicode newline sequences are the three just
299 mentioned, plus the single characters VT (vertical tab, U+000B), FF (form feed,
300 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
301 (paragraph separator, U+2029).
302 .P
303 Each of the first three conventions is used by at least one operating system as
304 its standard newline sequence. When PCRE is built, a default can be specified.
305 The default default is LF, which is the Unix standard. When PCRE is run, the
306 default can be overridden, either when a pattern is compiled, or when it is
307 matched.
308 .P
309 At compile time, the newline convention can be specified by the \fIoptions\fP
310 argument of \fBpcre_compile()\fP, or it can be specified by special text at the
311 start of the pattern itself; this overrides any other settings. See the
312 .\" HREF
313 \fBpcrepattern\fP
314 .\"
315 page for details of the special character sequences.
316 .P
317 In the PCRE documentation the word "newline" is used to mean "the character or
318 pair of characters that indicate a line break". The choice of newline
319 convention affects the handling of the dot, circumflex, and dollar
320 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
321 recognized line ending sequence, the match position advancement for a
322 non-anchored pattern. There is more detail about this in the
323 .\" HTML <a href="#execoptions">
324 .\" </a>
325 section on \fBpcre_exec()\fP options
326 .\"
327 below.
328 .P
329 The choice of newline convention does not affect the interpretation of
330 the \en or \er escape sequences, nor does it affect what \eR matches, which is
331 controlled in a similar way, but by separate options.
332 .
333 .
334 .SH MULTITHREADING
335 .rs
336 .sp
337 The PCRE functions can be used in multi-threading applications, with the
338 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
339 \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
340 callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
341 .P
342 The compiled form of a regular expression is not altered during matching, so
343 the same compiled pattern can safely be used by several threads at once.
344 .P
345 If the just-in-time optimization feature is being used, it needs separate
346 memory stack areas for each thread. See the
347 .\" HREF
348 \fBpcrejit\fP
349 .\"
350 documentation for more details.
351 .
352 .
353 .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
354 .rs
355 .sp
356 The compiled form of a regular expression can be saved and re-used at a later
357 time, possibly by a different program, and even on a host other than the one on
358 which it was compiled. Details are given in the
359 .\" HREF
360 \fBpcreprecompile\fP
361 .\"
362 documentation, which includes a description of the
363 \fBpcre_pattern_to_host_byte_order()\fP function. However, compiling a regular
364 expression with one version of PCRE for use with a different version is not
365 guaranteed to work and may cause crashes.
366 .
367 .
368 .SH "CHECKING BUILD-TIME OPTIONS"
369 .rs
370 .sp
371 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
372 .PP
373 The function \fBpcre_config()\fP makes it possible for a PCRE client to
374 discover which optional features have been compiled into the PCRE library. The
375 .\" HREF
376 \fBpcrebuild\fP
377 .\"
378 documentation has more details about these optional features.
379 .P
380 The first argument for \fBpcre_config()\fP is an integer, specifying which
381 information is required; the second argument is a pointer to a variable into
382 which the information is placed. The returned value is zero on success, or the
383 negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
384 not recognized. The following information is available:
385 .sp
386 PCRE_CONFIG_UTF8
387 .sp
388 The output is an integer that is set to one if UTF-8 support is available;
389 otherwise it is set to zero. This value should normally be given to the 8-bit
390 version of this function, \fBpcre_config()\fP. If it is given to the 16-bit
391 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
392 .sp
393 PCRE_CONFIG_UTF16
394 .sp
395 The output is an integer that is set to one if UTF-16 support is available;
396 otherwise it is set to zero. This value should normally be given to the 16-bit
397 version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
398 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
399 .sp
400 PCRE_CONFIG_UTF32
401 .sp
402 The output is an integer that is set to one if UTF-32 support is available;
403 otherwise it is set to zero. This value should normally be given to the 32-bit
404 version of this function, \fBpcre32_config()\fP. If it is given to the 8-bit
405 or 16-bit version of this function, the result is PCRE_ERROR_BADOPTION.
406 .sp
407 PCRE_CONFIG_UNICODE_PROPERTIES
408 .sp
409 The output is an integer that is set to one if support for Unicode character
410 properties is available; otherwise it is set to zero.
411 .sp
412 PCRE_CONFIG_JIT
413 .sp
414 The output is an integer that is set to one if support for just-in-time
415 compiling is available; otherwise it is set to zero.
416 .sp
417 PCRE_CONFIG_JITTARGET
418 .sp
419 The output is a pointer to a zero-terminated "const char *" string. If JIT
420 support is available, the string contains the name of the architecture for
421 which the JIT compiler is configured, for example "x86 32bit (little endian +
422 unaligned)". If JIT support is not available, the result is NULL.
423 .sp
424 PCRE_CONFIG_NEWLINE
425 .sp
426 The output is an integer whose value specifies the default character sequence
427 that is recognized as meaning "newline". The values that are supported in
428 ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for
429 ANYCRLF, and -1 for ANY. In EBCDIC environments, CR, ANYCRLF, and ANY yield the
430 same values. However, the value for LF is normally 21, though some EBCDIC
431 environments use 37. The corresponding values for CRLF are 3349 and 3365. The
432 default should normally correspond to the standard sequence for your operating
433 system.
434 .sp
435 PCRE_CONFIG_BSR
436 .sp
437 The output is an integer whose value indicates what character sequences the \eR
438 escape sequence matches by default. A value of 0 means that \eR matches any
439 Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
440 or CRLF. The default can be overridden when a pattern is compiled or matched.
441 .sp
442 PCRE_CONFIG_LINK_SIZE
443 .sp
444 The output is an integer that contains the number of bytes used for internal
445 linkage in compiled regular expressions. For the 8-bit library, the value can
446 be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
447 a number of bytes. For the 32-bit library, the value is either 2 or 4 and is
448 still a number of bytes. The default value of 2 is sufficient for all but the
449 most massive patterns, since it allows the compiled pattern to be up to 64K in
450 size. Larger values allow larger regular expressions to be compiled, at the
451 expense of slower matching.
452 .sp
453 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
454 .sp
455 The output is an integer that contains the threshold above which the POSIX
456 interface uses \fBmalloc()\fP for output vectors. Further details are given in
457 the
458 .\" HREF
459 \fBpcreposix\fP
460 .\"
461 documentation.
462 .sp
463 PCRE_CONFIG_MATCH_LIMIT
464 .sp
465 The output is a long integer that gives the default limit for the number of
466 internal matching function calls in a \fBpcre_exec()\fP execution. Further
467 details are given with \fBpcre_exec()\fP below.
468 .sp
469 PCRE_CONFIG_MATCH_LIMIT_RECURSION
470 .sp
471 The output is a long integer that gives the default limit for the depth of
472 recursion when calling the internal matching function in a \fBpcre_exec()\fP
473 execution. Further details are given with \fBpcre_exec()\fP below.
474 .sp
475 PCRE_CONFIG_STACKRECURSE
476 .sp
477 The output is an integer that is set to one if internal recursion when running
478 \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
479 to remember their state. This is the usual way that PCRE is compiled. The
480 output is zero if PCRE was compiled to use blocks of data on the heap instead
481 of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
482 \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
483 avoiding the use of the stack.
484 .
485 .
486 .SH "COMPILING A PATTERN"
487 .rs
488 .sp
489 .nf
490 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
491 .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
492 .B " const unsigned char *\fItableptr\fP);"
493 .sp
494 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
495 .B " int *\fIerrorcodeptr\fP,"
496 .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
497 .B " const unsigned char *\fItableptr\fP);"
498 .fi
499 .P
500 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
501 called to compile a pattern into an internal form. The only difference between
502 the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
503 \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
504 too much repetition, we refer just to \fBpcre_compile()\fP below, but the
505 information applies equally to \fBpcre_compile2()\fP.
506 .P
507 The pattern is a C string terminated by a binary zero, and is passed in the
508 \fIpattern\fP argument. A pointer to a single block of memory that is obtained
509 via \fBpcre_malloc\fP is returned. This contains the compiled code and related
510 data. The \fBpcre\fP type is defined for the returned block; this is a typedef
511 for a structure whose contents are not externally defined. It is up to the
512 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
513 .P
514 Although the compiled code of a PCRE regex is relocatable, that is, it does not
515 depend on memory location, the complete \fBpcre\fP data block is not
516 fully relocatable, because it may contain a copy of the \fItableptr\fP
517 argument, which is an address (see below).
518 .P
519 The \fIoptions\fP argument contains various bit settings that affect the
520 compilation. It should be zero if no options are required. The available
521 options are described below. Some of them (in particular, those that are
522 compatible with Perl, but some others as well) can also be set and unset from
523 within the pattern (see the detailed description in the
524 .\" HREF
525 \fBpcrepattern\fP
526 .\"
527 documentation). For those options that can be different in different parts of
528 the pattern, the contents of the \fIoptions\fP argument specifies their
529 settings at the start of compilation and execution. The PCRE_ANCHORED,
530 PCRE_BSR_\fIxxx\fP, PCRE_NEWLINE_\fIxxx\fP, PCRE_NO_UTF8_CHECK, and
531 PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
532 compile time.
533 .P
534 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
535 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
536 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
537 error message. This is a static string that is part of the library. You must
538 not try to free it. Normally, the offset from the start of the pattern to the
539 data unit that was being processed when the error was discovered is placed in
540 the variable pointed to by \fIerroffset\fP, which must not be NULL (if it is,
541 an immediate error is given). However, for an invalid UTF-8 or UTF-16 string,
542 the offset is that of the first data unit of the failing character.
543 .P
544 Some errors are not detected until the whole pattern has been scanned; in these
545 cases, the offset passed back is the length of the pattern. Note that the
546 offset is in data units, not characters, even in a UTF mode. It may sometimes
547 point into the middle of a UTF-8 or UTF-16 character.
548 .P
549 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
550 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
551 returned via this argument in the event of an error. This is in addition to the
552 textual error message. Error codes and messages are listed below.
553 .P
554 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
555 character tables that are built when PCRE is compiled, using the default C
556 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
557 call to \fBpcre_maketables()\fP. This value is stored with the compiled
558 pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
559 passed to it. For more discussion, see the section on locale support below.
560 .P
561 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
562 .sp
563 pcre *re;
564 const char *error;
565 int erroffset;
566 re = pcre_compile(
567 "^A.*Z", /* the pattern */
568 0, /* default options */
569 &error, /* for error message */
570 &erroffset, /* for error offset */
571 NULL); /* use default character tables */
572 .sp
573 The following names for option bits are defined in the \fBpcre.h\fP header
574 file:
575 .sp
576 PCRE_ANCHORED
577 .sp
578 If this bit is set, the pattern is forced to be "anchored", that is, it is
579 constrained to match only at the first matching point in the string that is
580 being searched (the "subject string"). This effect can also be achieved by
581 appropriate constructs in the pattern itself, which is the only way to do it in
582 Perl.
583 .sp
584 PCRE_AUTO_CALLOUT
585 .sp
586 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
587 all with number 255, before each pattern item. For discussion of the callout
588 facility, see the
589 .\" HREF
590 \fBpcrecallout\fP
591 .\"
592 documentation.
593 .sp
594 PCRE_BSR_ANYCRLF
595 PCRE_BSR_UNICODE
596 .sp
597 These options (which are mutually exclusive) control what the \eR escape
598 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
599 match any Unicode newline sequence. The default is specified when PCRE is
600 built. It can be overridden from within the pattern, or by setting an option
601 when a compiled pattern is matched.
602 .sp
603 PCRE_CASELESS
604 .sp
605 If this bit is set, letters in the pattern match both upper and lower case
606 letters. It is equivalent to Perl's /i option, and it can be changed within a
607 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
608 concept of case for characters whose values are less than 128, so caseless
609 matching is always possible. For characters with higher values, the concept of
610 case is supported if PCRE is compiled with Unicode property support, but not
611 otherwise. If you want to use caseless matching for characters 128 and above,
612 you must ensure that PCRE is compiled with Unicode property support as well as
613 with UTF-8 support.
614 .sp
615 PCRE_DOLLAR_ENDONLY
616 .sp
617 If this bit is set, a dollar metacharacter in the pattern matches only at the
618 end of the subject string. Without this option, a dollar also matches
619 immediately before a newline at the end of the string (but not before any other
620 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
621 There is no equivalent to this option in Perl, and no way to set it within a
622 pattern.
623 .sp
624 PCRE_DOTALL
625 .sp
626 If this bit is set, a dot metacharacter in the pattern matches a character of
627 any value, including one that indicates a newline. However, it only ever
628 matches one character, even if newlines are coded as CRLF. Without this option,
629 a dot does not match when the current position is at a newline. This option is
630 equivalent to Perl's /s option, and it can be changed within a pattern by a
631 (?s) option setting. A negative class such as [^a] always matches newline
632 characters, independent of the setting of this option.
633 .sp
634 PCRE_DUPNAMES
635 .sp
636 If this bit is set, names used to identify capturing subpatterns need not be
637 unique. This can be helpful for certain types of pattern when it is known that
638 only one instance of the named subpattern can ever be matched. There are more
639 details of named subpatterns below; see also the
640 .\" HREF
641 \fBpcrepattern\fP
642 .\"
643 documentation.
644 .sp
645 PCRE_EXTENDED
646 .sp
647 If this bit is set, white space data characters in the pattern are totally
648 ignored except when escaped or inside a character class. White space did not
649 used to include the VT character (code 11), because Perl did not treat this
650 character as white space. However, Perl changed at release 5.18, so PCRE
651 followed at release 8.34, and VT is now treated as white space. PCRE_EXTENDED
652 also causes characters between an unescaped # outside a character class and the
653 next newline, inclusive, to be ignored. PCRE_EXTENDED is equivalent to
654 Perl's /x option, and it can be changed within a pattern by a (?x) option
655 setting.
656 .P
657 Which characters are interpreted as newlines is controlled by the options
658 passed to \fBpcre_compile()\fP or by a special sequence at the start of the
659 pattern, as described in the section entitled
660 .\" HTML <a href="pcrepattern.html#newlines">
661 .\" </a>
662 "Newline conventions"
663 .\"
664 in the \fBpcrepattern\fP documentation. Note that the end of this type of
665 comment is a literal newline sequence in the pattern; escape sequences that
666 happen to represent a newline do not count.
667 .P
668 This option makes it possible to include comments inside complicated patterns.
669 Note, however, that this applies only to data characters. White space characters
670 may never appear within special character sequences in a pattern, for example
671 within the sequence (?( that introduces a conditional subpattern.
672 .sp
673 PCRE_EXTRA
674 .sp
675 This option was invented in order to turn on additional functionality of PCRE
676 that is incompatible with Perl, but it is currently of very little use. When
677 set, any backslash in a pattern that is followed by a letter that has no
678 special meaning causes an error, thus reserving these combinations for future
679 expansion. By default, as in Perl, a backslash followed by a letter with no
680 special meaning is treated as a literal. (Perl can, however, be persuaded to
681 give an error for this, by running it with the -w option.) There are at present
682 no other features controlled by this option. It can also be set by a (?X)
683 option setting within a pattern.
684 .sp
685 PCRE_FIRSTLINE
686 .sp
687 If this option is set, an unanchored pattern is required to match before or at
688 the first newline in the subject string, though the matched text may continue
689 over the newline.
690 .sp
691 PCRE_JAVASCRIPT_COMPAT
692 .sp
693 If this option is set, PCRE's behaviour is changed in some ways so that it is
694 compatible with JavaScript rather than Perl. The changes are as follows:
695 .P
696 (1) A lone closing square bracket in a pattern causes a compile-time error,
697 because this is illegal in JavaScript (by default it is treated as a data
698 character). Thus, the pattern AB]CD becomes illegal when this option is set.
699 .P
700 (2) At run time, a back reference to an unset subpattern group matches an empty
701 string (by default this causes the current matching alternative to fail). A
702 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
703 an "a" in the subject), whereas it fails by default, for Perl compatibility.
704 .P
705 (3) \eU matches an upper case "U" character; by default \eU causes a compile
706 time error (Perl uses \eU to upper case subsequent characters).
707 .P
708 (4) \eu matches a lower case "u" character unless it is followed by four
709 hexadecimal digits, in which case the hexadecimal number defines the code point
710 to match. By default, \eu causes a compile time error (Perl uses it to upper
711 case the following character).
712 .P
713 (5) \ex matches a lower case "x" character unless it is followed by two
714 hexadecimal digits, in which case the hexadecimal number defines the code point
715 to match. By default, as in Perl, a hexadecimal number is always expected after
716 \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
717 binary zero character followed by z).
718 .sp
719 PCRE_MULTILINE
720 .sp
721 By default, for the purposes of matching "start of line" and "end of line",
722 PCRE treats the subject string as consisting of a single line of characters,
723 even if it actually contains newlines. The "start of line" metacharacter (^)
724 matches only at the start of the string, and the "end of line" metacharacter
725 ($) matches only at the end of the string, or before a terminating newline
726 (except when PCRE_DOLLAR_ENDONLY is set). Note, however, that unless
727 PCRE_DOTALL is set, the "any character" metacharacter (.) does not match at a
728 newline. This behaviour (for ^, $, and dot) is the same as Perl.
729 .P
730 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
731 match immediately following or immediately before internal newlines in the
732 subject string, respectively, as well as at the very start and end. This is
733 equivalent to Perl's /m option, and it can be changed within a pattern by a
734 (?m) option setting. If there are no newlines in a subject string, or no
735 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
736 .sp
737 PCRE_NEVER_UTF
738 .sp
739 This option locks out interpretation of the pattern as UTF-8 (or UTF-16 or
740 UTF-32 in the 16-bit and 32-bit libraries). In particular, it prevents the
741 creator of the pattern from switching to UTF interpretation by starting the
742 pattern with (*UTF). This may be useful in applications that process patterns
743 from external sources. The combination of PCRE_UTF8 and PCRE_NEVER_UTF also
744 causes an error.
745 .sp
746 PCRE_NEWLINE_CR
747 PCRE_NEWLINE_LF
748 PCRE_NEWLINE_CRLF
749 PCRE_NEWLINE_ANYCRLF
750 PCRE_NEWLINE_ANY
751 .sp
752 These options override the default newline definition that was chosen when PCRE
753 was built. Setting the first or the second specifies that a newline is
754 indicated by a single character (CR or LF, respectively). Setting
755 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
756 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
757 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
758 that any Unicode newline sequence should be recognized.
759 .P
760 In an ASCII/Unicode environment, the Unicode newline sequences are the three
761 just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
762 feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
763 (paragraph separator, U+2029). For the 8-bit library, the last two are
764 recognized only in UTF-8 mode.
765 .P
766 When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
767 CR is 0x0d, the same as ASCII. However, the character code for LF is normally
768 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
769 not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
770 less than 256. For more details, see the
771 .\" HREF
772 \fBpcrebuild\fP
773 .\"
774 documentation.
775 .P
776 The newline setting in the options word uses three bits that are treated
777 as a number, giving eight possibilities. Currently only six are used (default
778 plus the five values above). This means that if you set more than one newline
779 option, the combination may or may not be sensible. For example,
780 PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
781 other combinations may yield unused numbers and cause an error.
782 .P
783 The only time that a line break in a pattern is specially recognized when
784 compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
785 and so are ignored in this mode. Also, an unescaped # outside a character class
786 indicates a comment that lasts until after the next line break sequence. In
787 other circumstances, line break sequences in patterns are treated as literal
788 data.
789 .P
790 The newline option that is set at compile time becomes the default that is used
791 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
792 .sp
793 PCRE_NO_AUTO_CAPTURE
794 .sp
795 If this option is set, it disables the use of numbered capturing parentheses in
796 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
797 were followed by ?: but named parentheses can still be used for capturing (and
798 they acquire numbers in the usual way). There is no equivalent of this option
799 in Perl.
800 .sp
801 PCRE_NO_AUTO_POSSESSIFY
802 .sp
803 If this option is set, it disables "auto-possessification". This is an
804 optimization that, for example, turns a+b into a++b in order to avoid
805 backtracks into a+ that can never be successful.
806 .sp
807 PCRE_NO_START_OPTIMIZE
808 .sp
809 This is an option that acts at matching time; that is, it is really an option
810 for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
811 it is remembered with the compiled pattern and assumed at matching time. This
812 is necessary if you want to use JIT execution, because the JIT compiler needs
813 to know whether or not this option is set. For details see the discussion of
814 PCRE_NO_START_OPTIMIZE
815 .\" HTML <a href="#execoptions">
816 .\" </a>
817 below.
818 .\"
819 .sp
820 PCRE_UCP
821 .sp
822 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
823 \ew, and some of the POSIX character classes. By default, only ASCII characters
824 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
825 classify characters. More details are given in the section on
826 .\" HTML <a href="pcre.html#genericchartypes">
827 .\" </a>
828 generic character types
829 .\"
830 in the
831 .\" HREF
832 \fBpcrepattern\fP
833 .\"
834 page. If you set PCRE_UCP, matching one of the items it affects takes much
835 longer. The option is available only if PCRE has been compiled with Unicode
836 property support.
837 .sp
838 PCRE_UNGREEDY
839 .sp
840 This option inverts the "greediness" of the quantifiers so that they are not
841 greedy by default, but become greedy if followed by "?". It is not compatible
842 with Perl. It can also be set by a (?U) option setting within the pattern.
843 .sp
844 PCRE_UTF8
845 .sp
846 This option causes PCRE to regard both the pattern and the subject as strings
847 of UTF-8 characters instead of single-byte strings. However, it is available
848 only when PCRE is built to include UTF support. If not, the use of this option
849 provokes an error. Details of how this option changes the behaviour of PCRE are
850 given in the
851 .\" HREF
852 \fBpcreunicode\fP
853 .\"
854 page.
855 .sp
856 PCRE_NO_UTF8_CHECK
857 .sp
858 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
859 automatically checked. There is a discussion about the
860 .\" HTML <a href="pcreunicode.html#utf8strings">
861 .\" </a>
862 validity of UTF-8 strings
863 .\"
864 in the
865 .\" HREF
866 \fBpcreunicode\fP
867 .\"
868 page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
869 error. If you already know that your pattern is valid, and you want to skip
870 this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
871 When it is set, the effect of passing an invalid UTF-8 string as a pattern is
872 undefined. It may cause your program to crash or loop. Note that this option
873 can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
874 the validity checking of subject strings only. If the same string is being
875 matched many times, the option can be safely set for the second and subsequent
876 matchings to improve performance.
877 .
878 .
879 .SH "COMPILATION ERROR CODES"
880 .rs
881 .sp
882 The following table lists the error codes than may be returned by
883 \fBpcre_compile2()\fP, along with the error messages that may be returned by
884 both compiling functions. Note that error messages are always 8-bit ASCII
885 strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
886 have fallen out of use. To avoid confusion, they have not been re-used.
887 .sp
888 0 no error
889 1 \e at end of pattern
890 2 \ec at end of pattern
891 3 unrecognized character follows \e
892 4 numbers out of order in {} quantifier
893 5 number too big in {} quantifier
894 6 missing terminating ] for character class
895 7 invalid escape sequence in character class
896 8 range out of order in character class
897 9 nothing to repeat
898 10 [this code is not in use]
899 11 internal error: unexpected repeat
900 12 unrecognized character after (? or (?-
901 13 POSIX named classes are supported only within a class
902 14 missing )
903 15 reference to non-existent subpattern
904 16 erroffset passed as NULL
905 17 unknown option bit(s) set
906 18 missing ) after comment
907 19 [this code is not in use]
908 20 regular expression is too large
909 21 failed to get memory
910 22 unmatched parentheses
911 23 internal error: code overflow
912 24 unrecognized character after (?<
913 25 lookbehind assertion is not fixed length
914 26 malformed number or name after (?(
915 27 conditional group contains more than two branches
916 28 assertion expected after (?(
917 29 (?R or (?[+-]digits must be followed by )
918 30 unknown POSIX class name
919 31 POSIX collating elements are not supported
920 32 this version of PCRE is compiled without UTF support
921 33 [this code is not in use]
922 34 character value in \ex{} or \eo{} is too large
923 35 invalid condition (?(0)
924 36 \eC not allowed in lookbehind assertion
925 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
926 38 number after (?C is > 255
927 39 closing ) for (?C expected
928 40 recursive call could loop indefinitely
929 41 unrecognized character after (?P
930 42 syntax error in subpattern name (missing terminator)
931 43 two named subpatterns have the same name
932 44 invalid UTF-8 string (specifically UTF-8)
933 45 support for \eP, \ep, and \eX has not been compiled
934 46 malformed \eP or \ep sequence
935 47 unknown property name after \eP or \ep
936 48 subpattern name is too long (maximum 32 characters)
937 49 too many named subpatterns (maximum 10000)
938 50 [this code is not in use]
939 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
940 52 internal error: overran compiling workspace
941 53 internal error: previously-checked referenced subpattern
942 not found
943 54 DEFINE group contains more than one branch
944 55 repeating a DEFINE group is not allowed
945 56 inconsistent NEWLINE options
946 57 \eg is not followed by a braced, angle-bracketed, or quoted
947 name/number or by a plain number
948 58 a numbered reference must not be zero
949 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
950 60 (*VERB) not recognized or malformed
951 61 number is too big
952 62 subpattern name expected
953 63 digit expected after (?+
954 64 ] is an invalid data character in JavaScript compatibility mode
955 65 different names for subpatterns of the same number are
956 not allowed
957 66 (*MARK) must have an argument
958 67 this version of PCRE is not compiled with Unicode property
959 support
960 68 \ec must be followed by an ASCII character
961 69 \ek is not followed by a braced, angle-bracketed, or quoted name
962 70 internal error: unknown opcode in find_fixedlength()
963 71 \eN is not supported in a class
964 72 too many forward references
965 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
966 74 invalid UTF-16 string (specifically UTF-16)
967 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
968 76 character value in \eu.... sequence is too large
969 77 invalid UTF-32 string (specifically UTF-32)
970 78 setting UTF is disabled by the application
971 79 non-hex character in \ex{} (closing brace missing?)
972 80 non-octal character in \eo{} (closing brace missing?)
973 81 missing opening brace after \eo
974 .sp
975 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
976 be used if the limits were changed when PCRE was built.
977 .
978 .
979 .\" HTML <a name="studyingapattern"></a>
980 .SH "STUDYING A PATTERN"
981 .rs
982 .sp
983 .nf
984 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
985 .B " const char **\fIerrptr\fP);"
986 .fi
987 .PP
988 If a compiled pattern is going to be used several times, it is worth spending
989 more time analyzing it in order to speed up the time taken for matching. The
990 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
991 argument. If studying the pattern produces additional information that will
992 help speed up matching, \fBpcre_study()\fP returns a pointer to a
993 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
994 results of the study.
995 .P
996 The returned value from \fBpcre_study()\fP can be passed directly to
997 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
998 also contains other fields that can be set by the caller before the block is
999 passed; these are described
1000 .\" HTML <a href="#extradata">
1001 .\" </a>
1002 below
1003 .\"
1004 in the section on matching a pattern.
1005 .P
1006 If studying the pattern does not produce any useful information,
1007 \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
1008 calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
1009 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
1010 if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
1011 returns a \fBpcre_extra\fP block even if studying did not find any additional
1012 information. It may still return NULL, however, if an error occurs in
1013 \fBpcre_study()\fP.
1014 .P
1015 The second argument of \fBpcre_study()\fP contains option bits. There are three
1016 further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1017 .sp
1018 PCRE_STUDY_JIT_COMPILE
1019 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1020 PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1021 .sp
1022 If any of these are set, and the just-in-time compiler is available, the
1023 pattern is further compiled into machine code that executes much faster than
1024 the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
1025 compiler is not available, these options are ignored. All undefined bits in the
1026 \fIoptions\fP argument must be zero.
1027 .P
1028 JIT compilation is a heavyweight optimization. It can take some time for
1029 patterns to be analyzed, and for one-off matches and simple patterns the
1030 benefit of faster execution might be offset by a much slower study time.
1031 Not all patterns can be optimized by the JIT compiler. For those that cannot be
1032 handled, matching automatically falls back to the \fBpcre_exec()\fP
1033 interpreter. For more details, see the
1034 .\" HREF
1035 \fBpcrejit\fP
1036 .\"
1037 documentation.
1038 .P
1039 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
1040 studying succeeds (even if no data is returned), the variable it points to is
1041 set to NULL. Otherwise it is set to point to a textual error message. This is a
1042 static string that is part of the library. You must not try to free it. You
1043 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
1044 sure that it has run successfully.
1045 .P
1046 When you are finished with a pattern, you can free the memory used for the
1047 study data by calling \fBpcre_free_study()\fP. This function was added to the
1048 API for release 8.20. For earlier versions, the memory could be freed with
1049 \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1050 where JIT optimization is not used, but it is advisable to change to the new
1051 function when convenient.
1052 .P
1053 This is a typical way in which \fBpcre_study\fP() is used (except that in a
1054 real application there should be tests for errors):
1055 .sp
1056 int rc;
1057 pcre *re;
1058 pcre_extra *sd;
1059 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1060 sd = pcre_study(
1061 re, /* result of pcre_compile() */
1062 0, /* no options */
1063 &error); /* set to NULL or points to a message */
1064 rc = pcre_exec( /* see below for details of pcre_exec() options */
1065 re, sd, "subject", 7, 0, 0, ovector, 30);
1066 ...
1067 pcre_free_study(sd);
1068 pcre_free(re);
1069 .sp
1070 Studying a pattern does two things: first, a lower bound for the length of
1071 subject string that is needed to match the pattern is computed. This does not
1072 mean that there are any strings of that length that match, but it does
1073 guarantee that no shorter strings match. The value is used to avoid wasting
1074 time by trying to match strings that are shorter than the lower bound. You can
1075 find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
1076 .P
1077 Studying a pattern is also useful for non-anchored patterns that do not have a
1078 single fixed starting character. A bitmap of possible starting bytes is
1079 created. This speeds up finding a position in the subject at which to start
1080 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1081 In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1082 .P
1083 These two optimizations apply to both \fBpcre_exec()\fP and
1084 \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1085 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option.
1086 You might want to do this if your pattern contains callouts or (*MARK) and you
1087 want to make use of these facilities in cases where matching fails.
1088 .P
1089 PCRE_NO_START_OPTIMIZE can be specified at either compile time or execution
1090 time. However, if PCRE_NO_START_OPTIMIZE is passed to \fBpcre_exec()\fP, (that
1091 is, after any JIT compilation has happened) JIT execution is disabled. For JIT
1092 execution to work with PCRE_NO_START_OPTIMIZE, the option must be set at
1093 compile time.
1094 .P
1095 There is a longer discussion of PCRE_NO_START_OPTIMIZE
1096 .\" HTML <a href="#execoptions">
1097 .\" </a>
1098 below.
1099 .\"
1100 .
1101 .
1102 .\" HTML <a name="localesupport"></a>
1103 .SH "LOCALE SUPPORT"
1104 .rs
1105 .sp
1106 PCRE handles caseless matching, and determines whether characters are letters,
1107 digits, or whatever, by reference to a set of tables, indexed by character
1108 value. When running in UTF-8 mode, this applies only to characters
1109 with codes less than 128. By default, higher-valued codes never match escapes
1110 such as \ew or \ed, but they can be tested with \ep if PCRE is built with
1111 Unicode character property support. Alternatively, the PCRE_UCP option can be
1112 set at compile time; this causes \ew and friends to use Unicode property
1113 support instead of built-in tables. The use of locales with Unicode is
1114 discouraged. If you are handling characters with codes greater than 128, you
1115 should either use UTF-8 and Unicode, or use locales, but not try to mix the
1116 two.
1117 .P
1118 PCRE contains an internal set of tables that are used when the final argument
1119 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1120 Normally, the internal tables recognize only ASCII characters. However, when
1121 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1122 default "C" locale of the local system, which may cause them to be different.
1123 .P
1124 The internal tables can always be overridden by tables supplied by the
1125 application that calls PCRE. These may be created in a different locale from
1126 the default. As more and more applications change to using Unicode, the need
1127 for this locale support is expected to die away.
1128 .P
1129 External tables are built by calling the \fBpcre_maketables()\fP function,
1130 which has no arguments, in the relevant locale. The result can then be passed
1131 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
1132 example, to build and use tables that are appropriate for the French locale
1133 (where accented characters with values greater than 128 are treated as letters),
1134 the following code could be used:
1135 .sp
1136 setlocale(LC_CTYPE, "fr_FR");
1137 tables = pcre_maketables();
1138 re = pcre_compile(..., tables);
1139 .sp
1140 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1141 are using Windows, the name for the French locale is "french".
1142 .P
1143 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1144 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1145 that the memory containing the tables remains available for as long as it is
1146 needed.
1147 .P
1148 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1149 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1150 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
1151 pattern, compilation, studying and matching all happen in the same locale, but
1152 different patterns can be compiled in different locales.
1153 .P
1154 It is possible to pass a table pointer or NULL (indicating the use of the
1155 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
1156 this facility could be used to match a pattern in a different locale from the
1157 one in which it was compiled. Passing table pointers at run time is discussed
1158 below in the section on matching a pattern.
1159 .
1160 .
1161 .\" HTML <a name="infoaboutpattern"></a>
1162 .SH "INFORMATION ABOUT A PATTERN"
1163 .rs
1164 .sp
1165 .nf
1166 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1167 .B " int \fIwhat\fP, void *\fIwhere\fP);"
1168 .fi
1169 .PP
1170 The \fBpcre_fullinfo()\fP function returns information about a compiled
1171 pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1172 library at version 8.30, after more than 10 years of obsolescence.
1173 .P
1174 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1175 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1176 the pattern was not studied. The third argument specifies which piece of
1177 information is required, and the fourth argument is a pointer to a variable
1178 to receive the data. The yield of the function is zero for success, or one of
1179 the following negative numbers:
1180 .sp
1181 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1182 the argument \fIwhere\fP was NULL
1183 PCRE_ERROR_BADMAGIC the "magic number" was not found
1184 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1185 endianness
1186 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1187 PCRE_ERROR_UNSET the requested field is not set
1188 .sp
1189 The "magic number" is placed at the start of each compiled pattern as an simple
1190 check against passing an arbitrary memory pointer. The endianness error can
1191 occur if a compiled pattern is saved and reloaded on a different host. Here is
1192 a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1193 pattern:
1194 .sp
1195 int rc;
1196 size_t length;
1197 rc = pcre_fullinfo(
1198 re, /* result of pcre_compile() */
1199 sd, /* result of pcre_study(), or NULL */
1200 PCRE_INFO_SIZE, /* what is required */
1201 &length); /* where to put the data */
1202 .sp
1203 The possible values for the third argument are defined in \fBpcre.h\fP, and are
1204 as follows:
1205 .sp
1206 PCRE_INFO_BACKREFMAX
1207 .sp
1208 Return the number of the highest back reference in the pattern. The fourth
1209 argument should point to an \fBint\fP variable. Zero is returned if there are
1210 no back references.
1211 .sp
1212 PCRE_INFO_CAPTURECOUNT
1213 .sp
1214 Return the number of capturing subpatterns in the pattern. The fourth argument
1215 should point to an \fBint\fP variable.
1216 .sp
1217 PCRE_INFO_DEFAULT_TABLES
1218 .sp
1219 Return a pointer to the internal default character tables within PCRE. The
1220 fourth argument should point to an \fBunsigned char *\fP variable. This
1221 information call is provided for internal use by the \fBpcre_study()\fP
1222 function. External callers can cause PCRE to use its internal tables by passing
1223 a NULL table pointer.
1224 .sp
1225 PCRE_INFO_FIRSTBYTE
1226 .sp
1227 Return information about the first data unit of any matched string, for a
1228 non-anchored pattern. (The name of this option refers to the 8-bit library,
1229 where data units are bytes.) The fourth argument should point to an \fBint\fP
1230 variable.
1231 .P
1232 If there is a fixed first value, for example, the letter "c" from a pattern
1233 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1234 value is always less than 256. In the 16-bit library the value can be up to
1235 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1236 .P
1237 If there is no fixed first value, and if either
1238 .sp
1239 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1240 starts with "^", or
1241 .sp
1242 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1243 (if it were set, the pattern would be anchored),
1244 .sp
1245 -1 is returned, indicating that the pattern matches only at the start of a
1246 subject string or after any newline within the string. Otherwise -2 is
1247 returned. For anchored patterns, -2 is returned.
1248 .P
1249 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1250 to return the full 32-bit range of the character, this value is deprecated;
1251 instead the PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER values
1252 should be used.
1253 .sp
1254 PCRE_INFO_FIRSTTABLE
1255 .sp
1256 If the pattern was studied, and this resulted in the construction of a 256-bit
1257 table indicating a fixed set of values for the first data unit in any matching
1258 string, a pointer to the table is returned. Otherwise NULL is returned. The
1259 fourth argument should point to an \fBunsigned char *\fP variable.
1260 .sp
1261 PCRE_INFO_HASCRORLF
1262 .sp
1263 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1264 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1265 explicit match is either a literal CR or LF character, or \er or \en.
1266 .sp
1267 PCRE_INFO_JCHANGED
1268 .sp
1269 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1270 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1271 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1272 .sp
1273 PCRE_INFO_JIT
1274 .sp
1275 Return 1 if the pattern was studied with one of the JIT options, and
1276 just-in-time compiling was successful. The fourth argument should point to an
1277 \fBint\fP variable. A return value of 0 means that JIT support is not available
1278 in this version of PCRE, or that the pattern was not studied with a JIT option,
1279 or that the JIT compiler could not handle this particular pattern. See the
1280 .\" HREF
1281 \fBpcrejit\fP
1282 .\"
1283 documentation for details of what can and cannot be handled.
1284 .sp
1285 PCRE_INFO_JITSIZE
1286 .sp
1287 If the pattern was successfully studied with a JIT option, return the size of
1288 the JIT compiled code, otherwise return zero. The fourth argument should point
1289 to a \fBsize_t\fP variable.
1290 .sp
1291 PCRE_INFO_LASTLITERAL
1292 .sp
1293 Return the value of the rightmost literal data unit that must exist in any
1294 matched string, other than at its start, if such a value has been recorded. The
1295 fourth argument should point to an \fBint\fP variable. If there is no such
1296 value, -1 is returned. For anchored patterns, a last literal value is recorded
1297 only if it follows something of variable length. For example, for the pattern
1298 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1299 is -1.
1300 .P
1301 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1302 to return the full 32-bit range of characters, this value is deprecated;
1303 instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1304 be used.
1305 .sp
1306 PCRE_INFO_MATCH_EMPTY
1307 .sp
1308 Return 1 if the pattern can match an empty string, otherwise 0. The fourth
1309 argument should point to an \fBint\fP variable.
1310 .sp
1311 PCRE_INFO_MATCHLIMIT
1312 .sp
1313 If the pattern set a match limit by including an item of the form
1314 (*LIMIT_MATCH=nnnn) at the start, the value is returned. The fourth argument
1315 should point to an unsigned 32-bit integer. If no such value has been set, the
1316 call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1317 .sp
1318 PCRE_INFO_MAXLOOKBEHIND
1319 .sp
1320 Return the number of characters (NB not data units) in the longest lookbehind
1321 assertion in the pattern. This information is useful when doing multi-segment
1322 matching using the partial matching facilities. Note that the simple assertions
1323 \eb and \eB require a one-character lookbehind. \eA also registers a
1324 one-character lookbehind, though it does not actually inspect the previous
1325 character. This is to ensure that at least one character from the old segment
1326 is retained when a new segment is processed. Otherwise, if there are no
1327 lookbehinds in the pattern, \eA might match incorrectly at the start of a new
1328 segment.
1329 .sp
1330 PCRE_INFO_MINLENGTH
1331 .sp
1332 If the pattern was studied and a minimum length for matching subject strings
1333 was computed, its value is returned. Otherwise the returned value is -1. The
1334 value is a number of characters, which in UTF mode may be different from the
1335 number of data units. The fourth argument should point to an \fBint\fP
1336 variable. A non-negative value is a lower bound to the length of any matching
1337 string. There may not be any strings of that length that do actually match, but
1338 every string that does match is at least that long.
1339 .sp
1340 PCRE_INFO_NAMECOUNT
1341 PCRE_INFO_NAMEENTRYSIZE
1342 PCRE_INFO_NAMETABLE
1343 .sp
1344 PCRE supports the use of named as well as numbered capturing parentheses. The
1345 names are just an additional way of identifying the parentheses, which still
1346 acquire numbers. Several convenience functions such as
1347 \fBpcre_get_named_substring()\fP are provided for extracting captured
1348 substrings by name. It is also possible to extract the data directly, by first
1349 converting the name to a number in order to access the correct pointers in the
1350 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1351 you need to use the name-to-number map, which is described by these three
1352 values.
1353 .P
1354 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1355 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1356 entry; both of these return an \fBint\fP value. The entry size depends on the
1357 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1358 entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1359 the first two bytes of each entry are the number of the capturing parenthesis,
1360 most significant byte first. In the 16-bit library, the pointer points to
1361 16-bit data units, the first of which contains the parenthesis number. In the
1362 32-bit library, the pointer points to 32-bit data units, the first of which
1363 contains the parenthesis number. The rest of the entry is the corresponding
1364 name, zero terminated.
1365 .P
1366 The names are in alphabetical order. If (?| is used to create multiple groups
1367 with the same number, as described in the
1368 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1369 .\" </a>
1370 section on duplicate subpattern numbers
1371 .\"
1372 in the
1373 .\" HREF
1374 \fBpcrepattern\fP
1375 .\"
1376 page, the groups may be given the same name, but there is only one entry in the
1377 table. Different names for groups of the same number are not permitted.
1378 Duplicate names for subpatterns with different numbers are permitted,
1379 but only if PCRE_DUPNAMES is set. They appear in the table in the order in
1380 which they were found in the pattern. In the absence of (?| this is the order
1381 of increasing number; when (?| is used this is not necessarily the case because
1382 later subpatterns may have lower numbers.
1383 .P
1384 As a simple example of the name/number table, consider the following pattern
1385 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1386 space - including newlines - is ignored):
1387 .sp
1388 .\" JOIN
1389 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1390 (?<month>\ed\ed) - (?<day>\ed\ed) )
1391 .sp
1392 There are four named subpatterns, so the table has four entries, and each entry
1393 in the table is eight bytes long. The table is as follows, with non-printing
1394 bytes shows in hexadecimal, and undefined bytes shown as ??:
1395 .sp
1396 00 01 d a t e 00 ??
1397 00 05 d a y 00 ?? ??
1398 00 04 m o n t h 00
1399 00 02 y e a r 00 ??
1400 .sp
1401 When writing code to extract data from named subpatterns using the
1402 name-to-number map, remember that the length of the entries is likely to be
1403 different for each compiled pattern.
1404 .sp
1405 PCRE_INFO_OKPARTIAL
1406 .sp
1407 Return 1 if the pattern can be used for partial matching with
1408 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1409 \fBint\fP variable. From release 8.00, this always returns 1, because the
1410 restrictions that previously applied to partial matching have been lifted. The
1411 .\" HREF
1412 \fBpcrepartial\fP
1413 .\"
1414 documentation gives details of partial matching.
1415 .sp
1416 PCRE_INFO_OPTIONS
1417 .sp
1418 Return a copy of the options with which the pattern was compiled. The fourth
1419 argument should point to an \fBunsigned long int\fP variable. These option bits
1420 are those specified in the call to \fBpcre_compile()\fP, modified by any
1421 top-level option settings at the start of the pattern itself. In other words,
1422 they are the options that will be in force when matching starts. For example,
1423 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1424 result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1425 .P
1426 A pattern is automatically anchored by PCRE if all of its top-level
1427 alternatives begin with one of the following:
1428 .sp
1429 ^ unless PCRE_MULTILINE is set
1430 \eA always
1431 \eG always
1432 .\" JOIN
1433 .* if PCRE_DOTALL is set and there are no back
1434 references to the subpattern in which .* appears
1435 .sp
1436 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1437 \fBpcre_fullinfo()\fP.
1438 .sp
1439 PCRE_INFO_RECURSIONLIMIT
1440 .sp
1441 If the pattern set a recursion limit by including an item of the form
1442 (*LIMIT_RECURSION=nnnn) at the start, the value is returned. The fourth
1443 argument should point to an unsigned 32-bit integer. If no such value has been
1444 set, the call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1445 .sp
1446 PCRE_INFO_SIZE
1447 .sp
1448 Return the size of the compiled pattern in bytes (for all three libraries). The
1449 fourth argument should point to a \fBsize_t\fP variable. This value does not
1450 include the size of the \fBpcre\fP structure that is returned by
1451 \fBpcre_compile()\fP. The value that is passed as the argument to
1452 \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1453 place the compiled data is the value returned by this option plus the size of
1454 the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1455 does not alter the value returned by this option.
1456 .sp
1457 PCRE_INFO_STUDYSIZE
1458 .sp
1459 Return the size in bytes (for all three libraries) of the data block pointed to
1460 by the \fIstudy_data\fP field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP
1461 is NULL, or there is no study data, zero is returned. The fourth argument
1462 should point to a \fBsize_t\fP variable. The \fIstudy_data\fP field is set by
1463 \fBpcre_study()\fP to record information that will speed up matching (see the
1464 section entitled
1465 .\" HTML <a href="#studyingapattern">
1466 .\" </a>
1467 "Studying a pattern"
1468 .\"
1469 above). The format of the \fIstudy_data\fP block is private, but its length
1470 is made available via this option so that it can be saved and restored (see the
1471 .\" HREF
1472 \fBpcreprecompile\fP
1473 .\"
1474 documentation for details).
1475 .sp
1476 PCRE_INFO_FIRSTCHARACTERFLAGS
1477 .sp
1478 Return information about the first data unit of any matched string, for a
1479 non-anchored pattern. The fourth argument should point to an \fBint\fP
1480 variable.
1481 .P
1482 If there is a fixed first value, for example, the letter "c" from a pattern
1483 such as (cat|cow|coyote), 1 is returned, and the character value can be
1484 retrieved using PCRE_INFO_FIRSTCHARACTER.
1485 .P
1486 If there is no fixed first value, and if either
1487 .sp
1488 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1489 starts with "^", or
1490 .sp
1491 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1492 (if it were set, the pattern would be anchored),
1493 .sp
1494 2 is returned, indicating that the pattern matches only at the start of a
1495 subject string or after any newline within the string. Otherwise 0 is
1496 returned. For anchored patterns, 0 is returned.
1497 .sp
1498 PCRE_INFO_FIRSTCHARACTER
1499 .sp
1500 Return the fixed first character value in the situation where
1501 PCRE_INFO_FIRSTCHARACTERFLAGS returns 1; otherwise return 0. The fourth
1502 argument should point to an \fBuint_t\fP variable.
1503 .P
1504 In the 8-bit library, the value is always less than 256. In the 16-bit library
1505 the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1506 can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1507 .sp
1508 PCRE_INFO_REQUIREDCHARFLAGS
1509 .sp
1510 Returns 1 if there is a rightmost literal data unit that must exist in any
1511 matched string, other than at its start. The fourth argument should point to
1512 an \fBint\fP variable. If there is no such value, 0 is returned. If returning
1513 1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1514 .P
1515 For anchored patterns, a last literal value is recorded only if it follows
1516 something of variable length. For example, for the pattern /^a\ed+z\ed+/ the
1517 returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1518 /^a\edz\ed/ the returned value is 0.
1519 .sp
1520 PCRE_INFO_REQUIREDCHAR
1521 .sp
1522 Return the value of the rightmost literal data unit that must exist in any
1523 matched string, other than at its start, if such a value has been recorded. The
1524 fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1525 value, 0 is returned.
1526 .
1527 .
1528 .SH "REFERENCE COUNTS"
1529 .rs
1530 .sp
1531 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1532 .PP
1533 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1534 data block that contains a compiled pattern. It is provided for the benefit of
1535 applications that operate in an object-oriented manner, where different parts
1536 of the application may be using the same compiled pattern, but you want to free
1537 the block when they are all done.
1538 .P
1539 When a pattern is compiled, the reference count field is initialized to zero.
1540 It is changed only by calling this function, whose action is to add the
1541 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1542 function is the new value. However, the value of the count is constrained to
1543 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1544 it is forced to the appropriate limit value.
1545 .P
1546 Except when it is zero, the reference count is not correctly preserved if a
1547 pattern is compiled on one host and then transferred to a host whose byte-order
1548 is different. (This seems a highly unlikely scenario.)
1549 .
1550 .
1551 .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1552 .rs
1553 .sp
1554 .nf
1555 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1556 .B " const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1557 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);"
1558 .fi
1559 .P
1560 The function \fBpcre_exec()\fP is called to match a subject string against a
1561 compiled pattern, which is passed in the \fIcode\fP argument. If the
1562 pattern was studied, the result of the study should be passed in the
1563 \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1564 and \fIextra\fP arguments as many times as you like, in order to match
1565 different subject strings with the same pattern.
1566 .P
1567 This function is the main matching facility of the library, and it operates in
1568 a Perl-like manner. For specialist use there is also an alternative matching
1569 function, which is described
1570 .\" HTML <a href="#dfamatch">
1571 .\" </a>
1572 below
1573 .\"
1574 in the section about the \fBpcre_dfa_exec()\fP function.
1575 .P
1576 In most applications, the pattern will have been compiled (and optionally
1577 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1578 possible to save compiled patterns and study data, and then use them later
1579 in different processes, possibly even on different hosts. For a discussion
1580 about this, see the
1581 .\" HREF
1582 \fBpcreprecompile\fP
1583 .\"
1584 documentation.
1585 .P
1586 Here is an example of a simple call to \fBpcre_exec()\fP:
1587 .sp
1588 int rc;
1589 int ovector[30];
1590 rc = pcre_exec(
1591 re, /* result of pcre_compile() */
1592 NULL, /* we didn't study the pattern */
1593 "some string", /* the subject string */
1594 11, /* the length of the subject string */
1595 0, /* start at offset 0 in the subject */
1596 0, /* default options */
1597 ovector, /* vector of integers for substring information */
1598 30); /* number of elements (NOT size in bytes) */
1599 .
1600 .
1601 .\" HTML <a name="extradata"></a>
1602 .SS "Extra data for \fBpcre_exec()\fR"
1603 .rs
1604 .sp
1605 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1606 data block. The \fBpcre_study()\fP function returns such a block (when it
1607 doesn't return NULL), but you can also create one for yourself, and pass
1608 additional information in it. The \fBpcre_extra\fP block contains the following
1609 fields (not necessarily in this order):
1610 .sp
1611 unsigned long int \fIflags\fP;
1612 void *\fIstudy_data\fP;
1613 void *\fIexecutable_jit\fP;
1614 unsigned long int \fImatch_limit\fP;
1615 unsigned long int \fImatch_limit_recursion\fP;
1616 void *\fIcallout_data\fP;
1617 const unsigned char *\fItables\fP;
1618 unsigned char **\fImark\fP;
1619 .sp
1620 In the 16-bit version of this structure, the \fImark\fP field has type
1621 "PCRE_UCHAR16 **".
1622 .sp
1623 In the 32-bit version of this structure, the \fImark\fP field has type
1624 "PCRE_UCHAR32 **".
1625 .P
1626 The \fIflags\fP field is used to specify which of the other fields are set. The
1627 flag bits are:
1628 .sp
1629 PCRE_EXTRA_CALLOUT_DATA
1630 PCRE_EXTRA_EXECUTABLE_JIT
1631 PCRE_EXTRA_MARK
1632 PCRE_EXTRA_MATCH_LIMIT
1633 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1634 PCRE_EXTRA_STUDY_DATA
1635 PCRE_EXTRA_TABLES
1636 .sp
1637 Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1638 the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1639 returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1640 should not set these yourself, but you may add to the block by setting other
1641 fields and their corresponding flag bits.
1642 .P
1643 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1644 vast amount of resources when running patterns that are not going to match,
1645 but which have a very large number of possibilities in their search trees. The
1646 classic example is a pattern that uses nested unlimited repeats.
1647 .P
1648 Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1649 calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1650 imposed on the number of times this function is called during a match, which
1651 has the effect of limiting the amount of backtracking that can take place. For
1652 patterns that are not anchored, the count restarts from zero for each position
1653 in the subject string.
1654 .P
1655 When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1656 with a JIT option, the way that the matching is executed is entirely different.
1657 However, there is still the possibility of runaway matching that goes on for a
1658 very long time, and so the \fImatch_limit\fP value is also used in this case
1659 (but in a different way) to limit how long the matching can continue.
1660 .P
1661 The default value for the limit can be set when PCRE is built; the default
1662 default is 10 million, which handles all but the most extreme cases. You can
1663 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1664 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1665 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1666 PCRE_ERROR_MATCHLIMIT.
1667 .P
1668 A value for the match limit may also be supplied by an item at the start of a
1669 pattern of the form
1670 .sp
1671 (*LIMIT_MATCH=d)
1672 .sp
1673 where d is a decimal number. However, such a setting is ignored unless d is
1674 less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1675 is set, less than the default.
1676 .P
1677 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1678 instead of limiting the total number of times that \fBmatch()\fP is called, it
1679 limits the depth of recursion. The recursion depth is a smaller number than the
1680 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1681 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1682 .P
1683 Limiting the recursion depth limits the amount of machine stack that can be
1684 used, or, when PCRE has been compiled to use memory on the heap instead of the
1685 stack, the amount of heap memory that can be used. This limit is not relevant,
1686 and is ignored, when matching is done using JIT compiled code.
1687 .P
1688 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1689 built; the default default is the same value as the default for
1690 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1691 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1692 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1693 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1694 .P
1695 A value for the recursion limit may also be supplied by an item at the start of
1696 a pattern of the form
1697 .sp
1698 (*LIMIT_RECURSION=d)
1699 .sp
1700 where d is a decimal number. However, such a setting is ignored unless d is
1701 less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1702 is set, less than the default.
1703 .P
1704 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1705 and is described in the
1706 .\" HREF
1707 \fBpcrecallout\fP
1708 .\"
1709 documentation.
1710 .P
1711 The \fItables\fP field is used to pass a character tables pointer to
1712 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1713 pattern. A non-NULL value is stored with the compiled pattern only if custom
1714 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1715 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1716 internal tables to be used. This facility is helpful when re-using patterns
1717 that have been saved after compiling with an external set of tables, because
1718 the external tables might be at a different address when \fBpcre_exec()\fP is
1719 called. See the
1720 .\" HREF
1721 \fBpcreprecompile\fP
1722 .\"
1723 documentation for a discussion of saving compiled patterns for later use.
1724 .P
1725 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1726 be set to point to a suitable variable. If the pattern contains any
1727 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1728 a name to pass back, a pointer to the name string (zero terminated) is placed
1729 in the variable pointed to by the \fImark\fP field. The names are within the
1730 compiled pattern; if you wish to retain such a name you must copy it before
1731 freeing the memory of a compiled pattern. If there is no name to pass back, the
1732 variable pointed to by the \fImark\fP field is set to NULL. For details of the
1733 backtracking control verbs, see the section entitled
1734 .\" HTML <a href="pcrepattern#backtrackcontrol">
1735 .\" </a>
1736 "Backtracking control"
1737 .\"
1738 in the
1739 .\" HREF
1740 \fBpcrepattern\fP
1741 .\"
1742 documentation.
1743 .
1744 .
1745 .\" HTML <a name="execoptions"></a>
1746 .SS "Option bits for \fBpcre_exec()\fP"
1747 .rs
1748 .sp
1749 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1750 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1751 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1752 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1753 PCRE_PARTIAL_SOFT.
1754 .P
1755 If the pattern was successfully studied with one of the just-in-time (JIT)
1756 compile options, the only supported options for JIT execution are
1757 PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1758 PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1759 unsupported option is used, JIT execution is disabled and the normal
1760 interpretive code in \fBpcre_exec()\fP is run.
1761 .sp
1762 PCRE_ANCHORED
1763 .sp
1764 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1765 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1766 to be anchored by virtue of its contents, it cannot be made unachored at
1767 matching time.
1768 .sp
1769 PCRE_BSR_ANYCRLF
1770 PCRE_BSR_UNICODE
1771 .sp
1772 These options (which are mutually exclusive) control what the \eR escape
1773 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1774 match any Unicode newline sequence. These options override the choice that was
1775 made or defaulted when the pattern was compiled.
1776 .sp
1777 PCRE_NEWLINE_CR
1778 PCRE_NEWLINE_LF
1779 PCRE_NEWLINE_CRLF
1780 PCRE_NEWLINE_ANYCRLF
1781 PCRE_NEWLINE_ANY
1782 .sp
1783 These options override the newline definition that was chosen or defaulted when
1784 the pattern was compiled. For details, see the description of
1785 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1786 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1787 the way the match position is advanced after a match failure for an unanchored
1788 pattern.
1789 .P
1790 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1791 match attempt for an unanchored pattern fails when the current position is at a
1792 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1793 characters, the match position is advanced by two characters instead of one, in
1794 other words, to after the CRLF.
1795 .P
1796 The above rule is a compromise that makes the most common cases work as
1797 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1798 set), it does not match the string "\er\enA" because, after failing at the
1799 start, it skips both the CR and the LF before retrying. However, the pattern
1800 [\er\en]A does match that string, because it contains an explicit CR or LF
1801 reference, and so advances only by one character after the first failure.
1802 .P
1803 An explicit match for CR of LF is either a literal appearance of one of those
1804 characters, or one of the \er or \en escape sequences. Implicit matches such as
1805 [^X] do not count, nor does \es (which includes CR and LF in the characters
1806 that it matches).
1807 .P
1808 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1809 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1810 .sp
1811 PCRE_NOTBOL
1812 .sp
1813 This option specifies that first character of the subject string is not the
1814 beginning of a line, so the circumflex metacharacter should not match before
1815 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1816 never to match. This option affects only the behaviour of the circumflex
1817 metacharacter. It does not affect \eA.
1818 .sp
1819 PCRE_NOTEOL
1820 .sp
1821 This option specifies that the end of the subject string is not the end of a
1822 line, so the dollar metacharacter should not match it nor (except in multiline
1823 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1824 compile time) causes dollar never to match. This option affects only the
1825 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1826 .sp
1827 PCRE_NOTEMPTY
1828 .sp
1829 An empty string is not considered to be a valid match if this option is set. If
1830 there are alternatives in the pattern, they are tried. If all the alternatives
1831 match the empty string, the entire match fails. For example, if the pattern
1832 .sp
1833 a?b?
1834 .sp
1835 is applied to a string not beginning with "a" or "b", it matches an empty
1836 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1837 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1838 .sp
1839 PCRE_NOTEMPTY_ATSTART
1840 .sp
1841 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1842 the start of the subject is permitted. If the pattern is anchored, such a match
1843 can occur only if the pattern contains \eK.
1844 .P
1845 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1846 does make a special case of a pattern match of the empty string within its
1847 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1848 emulate Perl's behaviour after matching a null string by first trying the match
1849 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1850 if that fails, by advancing the starting offset (see below) and trying an
1851 ordinary match again. There is some code that demonstrates how to do this in
1852 the
1853 .\" HREF
1854 \fBpcredemo\fP
1855 .\"
1856 sample program. In the most general case, you have to check to see if the
1857 newline convention recognizes CRLF as a newline, and if so, and the current
1858 character is CR followed by LF, advance the starting offset by two characters
1859 instead of one.
1860 .sp
1861 PCRE_NO_START_OPTIMIZE
1862 .sp
1863 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1864 a match, in order to speed up the process. For example, if it is known that an
1865 unanchored match must start with a specific character, it searches the subject
1866 for that character, and fails immediately if it cannot find it, without
1867 actually running the main matching function. This means that a special item
1868 such as (*COMMIT) at the start of a pattern is not considered until after a
1869 suitable starting point for the match has been found. Also, when callouts or
1870 (*MARK) items are in use, these "start-up" optimizations can cause them to be
1871 skipped if the pattern is never actually used. The start-up optimizations are
1872 in effect a pre-scan of the subject that takes place before the pattern is run.
1873 .P
1874 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1875 causing performance to suffer, but ensuring that in cases where the result is
1876 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1877 are considered at every possible starting position in the subject string. If
1878 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1879 time. The use of PCRE_NO_START_OPTIMIZE at matching time (that is, passing it
1880 to \fBpcre_exec()\fP) disables JIT execution; in this situation, matching is
1881 always done using interpretively.
1882 .P
1883 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1884 Consider the pattern
1885 .sp
1886 (*COMMIT)ABC
1887 .sp
1888 When this is compiled, PCRE records the fact that a match must start with the
1889 character "A". Suppose the subject string is "DEFABC". The start-up
1890 optimization scans along the subject, finds "A" and runs the first match
1891 attempt from there. The (*COMMIT) item means that the pattern must match the
1892 current starting position, which in this case, it does. However, if the same
1893 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1894 subject string does not happen. The first match attempt is run starting from
1895 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1896 the overall result is "no match". If the pattern is studied, more start-up
1897 optimizations may be used. For example, a minimum length for the subject may be
1898 recorded. Consider the pattern
1899 .sp
1900 (*MARK:A)(X|Y)
1901 .sp
1902 The minimum length for a match is one character. If the subject is "ABC", there
1903 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1904 If the pattern is studied, the final attempt does not take place, because PCRE
1905 knows that the subject is too short, and so the (*MARK) is never encountered.
1906 In this case, studying the pattern does not affect the overall match result,
1907 which is still "no match", but it does affect the auxiliary information that is
1908 returned.
1909 .sp
1910 PCRE_NO_UTF8_CHECK
1911 .sp
1912 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1913 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1914 The entire string is checked before any other processing takes place. The value
1915 of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1916 UTF-8 character. There is a discussion about the
1917 .\" HTML <a href="pcreunicode.html#utf8strings">
1918 .\" </a>
1919 validity of UTF-8 strings
1920 .\"
1921 in the
1922 .\" HREF
1923 \fBpcreunicode\fP
1924 .\"
1925 page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1926 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1927 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1928 cases, information about the precise nature of the error may also be returned
1929 (see the descriptions of these errors in the section entitled \fIError return
1930 values from\fP \fBpcre_exec()\fP
1931 .\" HTML <a href="#errorlist">
1932 .\" </a>
1933 below).
1934 .\"
1935 If \fIstartoffset\fP contains a value that does not point to the start of a
1936 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1937 returned.
1938 .P
1939 If you already know that your subject is valid, and you want to skip these
1940 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1941 calling \fBpcre_exec()\fP. You might want to do this for the second and
1942 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1943 all the matches in a single subject string. However, you should be sure that
1944 the value of \fIstartoffset\fP points to the start of a character (or the end
1945 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1946 invalid string as a subject or an invalid value of \fIstartoffset\fP is
1947 undefined. Your program may crash or loop.
1948 .sp
1949 PCRE_PARTIAL_HARD
1950 PCRE_PARTIAL_SOFT
1951 .sp
1952 These options turn on the partial matching feature. For backwards
1953 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1954 occurs if the end of the subject string is reached successfully, but there are
1955 not enough subject characters to complete the match. If this happens when
1956 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1957 testing any remaining alternatives. Only if no complete match can be found is
1958 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1959 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1960 but only if no complete match can be found.
1961 .P
1962 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1963 partial match is found, \fBpcre_exec()\fP immediately returns
1964 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1965 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1966 important that an alternative complete match.
1967 .P
1968 In both cases, the portion of the string that was inspected when the partial
1969 match was found is set as the first matching string. There is a more detailed
1970 discussion of partial and multi-segment matching, with examples, in the
1971 .\" HREF
1972 \fBpcrepartial\fP
1973 .\"
1974 documentation.
1975 .
1976 .
1977 .SS "The string to be matched by \fBpcre_exec()\fP"
1978 .rs
1979 .sp
1980 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1981 \fIsubject\fP, a length in \fIlength\fP, and a starting offset in
1982 \fIstartoffset\fP. The units for \fIlength\fP and \fIstartoffset\fP are bytes
1983 for the 8-bit library, 16-bit data items for the 16-bit library, and 32-bit
1984 data items for the 32-bit library.
1985 .P
1986 If \fIstartoffset\fP is negative or greater than the length of the subject,
1987 \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting offset is
1988 zero, the search for a match starts at the beginning of the subject, and this
1989 is by far the most common case. In UTF-8 or UTF-16 mode, the offset must point
1990 to the start of a character, or the end of the subject (in UTF-32 mode, one
1991 data unit equals one character, so all offsets are valid). Unlike the pattern
1992 string, the subject may contain binary zeroes.
1993 .P
1994 A non-zero starting offset is useful when searching for another match in the
1995 same subject by calling \fBpcre_exec()\fP again after a previous success.
1996 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1997 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1998 lookbehind. For example, consider the pattern
1999 .sp
2000 \eBiss\eB
2001 .sp
2002 which finds occurrences of "iss" in the middle of words. (\eB matches only if
2003 the current position in the subject is not a word boundary.) When applied to
2004 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
2005 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
2006 subject, namely "issipi", it does not match, because \eB is always false at the
2007 start of the subject, which is deemed to be a word boundary. However, if
2008 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
2009 set to 4, it finds the second occurrence of "iss" because it is able to look
2010 behind the starting point to discover that it is preceded by a letter.
2011 .P
2012 Finding all the matches in a subject is tricky when the pattern can match an
2013 empty string. It is possible to emulate Perl's /g behaviour by first trying the
2014 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
2015 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
2016 and trying an ordinary match again. There is some code that demonstrates how to
2017 do this in the
2018 .\" HREF
2019 \fBpcredemo\fP
2020 .\"
2021 sample program. In the most general case, you have to check to see if the
2022 newline convention recognizes CRLF as a newline, and if so, and the current
2023 character is CR followed by LF, advance the starting offset by two characters
2024 instead of one.
2025 .P
2026 If a non-zero starting offset is passed when the pattern is anchored, one
2027 attempt to match at the given offset is made. This can only succeed if the
2028 pattern does not require the match to be at the start of the subject.
2029 .
2030 .
2031 .SS "How \fBpcre_exec()\fP returns captured substrings"
2032 .rs
2033 .sp
2034 In general, a pattern matches a certain portion of the subject, and in
2035 addition, further substrings from the subject may be picked out by parts of the
2036 pattern. Following the usage in Jeffrey Friedl's book, this is called
2037 "capturing" in what follows, and the phrase "capturing subpattern" is used for
2038 a fragment of a pattern that picks out a substring. PCRE supports several other
2039 kinds of parenthesized subpattern that do not cause substrings to be captured.
2040 .P
2041 Captured substrings are returned to the caller via a vector of integers whose
2042 address is passed in \fIovector\fP. The number of elements in the vector is
2043 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
2044 argument is NOT the size of \fIovector\fP in bytes.
2045 .P
2046 The first two-thirds of the vector is used to pass back captured substrings,
2047 each substring using a pair of integers. The remaining third of the vector is
2048 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
2049 and is not available for passing back information. The number passed in
2050 \fIovecsize\fP should always be a multiple of three. If it is not, it is
2051 rounded down.
2052 .P
2053 When a match is successful, information about captured substrings is returned
2054 in pairs of integers, starting at the beginning of \fIovector\fP, and
2055 continuing up to two-thirds of its length at the most. The first element of
2056 each pair is set to the offset of the first character in a substring, and the
2057 second is set to the offset of the first character after the end of a
2058 substring. These values are always data unit offsets, even in UTF mode. They
2059 are byte offsets in the 8-bit library, 16-bit data item offsets in the 16-bit
2060 library, and 32-bit data item offsets in the 32-bit library. \fBNote\fP: they
2061 are not character counts.
2062 .P
2063 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
2064 portion of the subject string matched by the entire pattern. The next pair is
2065 used for the first capturing subpattern, and so on. The value returned by
2066 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
2067 For example, if two substrings have been captured, the returned value is 3. If
2068 there are no capturing subpatterns, the return value from a successful match is
2069 1, indicating that just the first pair of offsets has been set.
2070 .P
2071 If a capturing subpattern is matched repeatedly, it is the last portion of the
2072 string that it matched that is returned.
2073 .P
2074 If the vector is too small to hold all the captured substring offsets, it is
2075 used as far as possible (up to two-thirds of its length), and the function
2076 returns a value of zero. If neither the actual string matched nor any captured
2077 substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2078 passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2079 back references and the \fIovector\fP is not big enough to remember the related
2080 substrings, PCRE has to get additional memory for use during matching. Thus it
2081 is usually advisable to supply an \fIovector\fP of reasonable size.
2082 .P
2083 There are some cases where zero is returned (indicating vector overflow) when
2084 in fact the vector is exactly the right size for the final match. For example,
2085 consider the pattern
2086 .sp
2087 (a)(?:(b)c|bd)
2088 .sp
2089 If a vector of 6 elements (allowing for only 1 captured substring) is given
2090 with subject string "abd", \fBpcre_exec()\fP will try to set the second
2091 captured string, thereby recording a vector overflow, before failing to match
2092 "c" and backing up to try the second alternative. The zero return, however,
2093 does correctly indicate that the maximum number of slots (namely 2) have been
2094 filled. In similar cases where there is temporary overflow, but the final
2095 number of used slots is actually less than the maximum, a non-zero value is
2096 returned.
2097 .P
2098 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
2099 subpatterns there are in a compiled pattern. The smallest size for
2100 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
2101 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
2102 .P
2103 It is possible for capturing subpattern number \fIn+1\fP to match some part of
2104 the subject when subpattern \fIn\fP has not been used at all. For example, if
2105 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2106 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2107 happens, both values in the offset pairs corresponding to unused subpatterns
2108 are set to -1.
2109 .P
2110 Offset values that correspond to unused subpatterns at the end of the
2111 expression are also set to -1. For example, if the string "abc" is matched
2112 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2113 return from the function is 2, because the highest used capturing subpattern
2114 number is 1, and the offsets for for the second and third capturing subpatterns
2115 (assuming the vector is large enough, of course) are set to -1.
2116 .P
2117 \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
2118 correspond to capturing parentheses in the pattern are never changed. That is,
2119 if a pattern contains \fIn\fP capturing parentheses, no more than
2120 \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
2121 elements (in the first two-thirds) retain whatever values they previously had.
2122 .P
2123 Some convenience functions are provided for extracting the captured substrings
2124 as separate strings. These are described below.
2125 .
2126 .
2127 .\" HTML <a name="errorlist"></a>
2128 .SS "Error return values from \fBpcre_exec()\fP"
2129 .rs
2130 .sp
2131 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
2132 defined in the header file:
2133 .sp
2134 PCRE_ERROR_NOMATCH (-1)
2135 .sp
2136 The subject string did not match the pattern.
2137 .sp
2138 PCRE_ERROR_NULL (-2)
2139 .sp
2140 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
2141 NULL and \fIovecsize\fP was not zero.
2142 .sp
2143 PCRE_ERROR_BADOPTION (-3)
2144 .sp
2145 An unrecognized bit was set in the \fIoptions\fP argument.
2146 .sp
2147 PCRE_ERROR_BADMAGIC (-4)
2148 .sp
2149 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2150 the case when it is passed a junk pointer and to detect when a pattern that was
2151 compiled in an environment of one endianness is run in an environment with the
2152 other endianness. This is the error that PCRE gives when the magic number is
2153 not present.
2154 .sp
2155 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2156 .sp
2157 While running the pattern match, an unknown item was encountered in the
2158 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2159 of the compiled pattern.
2160 .sp
2161 PCRE_ERROR_NOMEMORY (-6)
2162 .sp
2163 If a pattern contains back references, but the \fIovector\fP that is passed to
2164 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2165 gets a block of memory at the start of matching to use for this purpose. If the
2166 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2167 automatically freed at the end of matching.
2168 .P
2169 This error is also given if \fBpcre_stack_malloc()\fP fails in
2170 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2171 \fB--disable-stack-for-recursion\fP.
2172 .sp
2173 PCRE_ERROR_NOSUBSTRING (-7)
2174 .sp
2175 This error is used by the \fBpcre_copy_substring()\fP,
2176 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2177 below). It is never returned by \fBpcre_exec()\fP.
2178 .sp
2179 PCRE_ERROR_MATCHLIMIT (-8)
2180 .sp
2181 The backtracking limit, as specified by the \fImatch_limit\fP field in a
2182 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2183 above.
2184 .sp
2185 PCRE_ERROR_CALLOUT (-9)
2186 .sp
2187 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2188 use by callout functions that want to yield a distinctive error code. See the
2189 .\" HREF
2190 \fBpcrecallout\fP
2191 .\"
2192 documentation for details.
2193 .sp
2194 PCRE_ERROR_BADUTF8 (-10)
2195 .sp
2196 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2197 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2198 (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2199 UTF-8 character is placed in the first element, and a reason code is placed in
2200 the second element. The reason codes are listed in the
2201 .\" HTML <a href="#badutf8reasons">
2202 .\" </a>
2203 following section.
2204 .\"
2205 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2206 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2207 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2208 .sp
2209 PCRE_ERROR_BADUTF8_OFFSET (-11)
2210 .sp
2211 The UTF-8 byte sequence that was passed as a subject was checked and found to
2212 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2213 \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2214 end of the subject.
2215 .sp
2216 PCRE_ERROR_PARTIAL (-12)
2217 .sp
2218 The subject string did not match, but it did match partially. See the
2219 .\" HREF
2220 \fBpcrepartial\fP
2221 .\"
2222 documentation for details of partial matching.
2223 .sp
2224 PCRE_ERROR_BADPARTIAL (-13)
2225 .sp
2226 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2227 option was used with a compiled pattern containing items that were not
2228 supported for partial matching. From release 8.00 onwards, there are no
2229 restrictions on partial matching.
2230 .sp
2231 PCRE_ERROR_INTERNAL (-14)
2232 .sp
2233 An unexpected internal error has occurred. This error could be caused by a bug
2234 in PCRE or by overwriting of the compiled pattern.
2235 .sp
2236 PCRE_ERROR_BADCOUNT (-15)
2237 .sp
2238 This error is given if the value of the \fIovecsize\fP argument is negative.
2239 .sp
2240 PCRE_ERROR_RECURSIONLIMIT (-21)
2241 .sp
2242 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2243 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2244 description above.
2245 .sp
2246 PCRE_ERROR_BADNEWLINE (-23)
2247 .sp
2248 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2249 .sp
2250 PCRE_ERROR_BADOFFSET (-24)
2251 .sp
2252 The value of \fIstartoffset\fP was negative or greater than the length of the
2253 subject, that is, the value in \fIlength\fP.
2254 .sp
2255 PCRE_ERROR_SHORTUTF8 (-25)
2256 .sp
2257 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2258 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2259 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2260 fact sufficient to detect this case, but this special error code for
2261 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2262 retained for backwards compatibility.
2263 .sp
2264 PCRE_ERROR_RECURSELOOP (-26)
2265 .sp
2266 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2267 the pattern. Specifically, it means that either the whole pattern or a
2268 subpattern has been called recursively for the second time at the same position
2269 in the subject string. Some simple patterns that might do this are detected and
2270 faulted at compile time, but more complicated cases, in particular mutual
2271 recursions between two different subpatterns, cannot be detected until run
2272 time.
2273 .sp
2274 PCRE_ERROR_JIT_STACKLIMIT (-27)
2275 .sp
2276 This error is returned when a pattern that was successfully studied using a
2277 JIT compile option is being matched, but the memory available for the
2278 just-in-time processing stack is not large enough. See the
2279 .\" HREF
2280 \fBpcrejit\fP
2281 .\"
2282 documentation for more details.
2283 .sp
2284 PCRE_ERROR_BADMODE (-28)
2285 .sp
2286 This error is given if a pattern that was compiled by the 8-bit library is
2287 passed to a 16-bit or 32-bit library function, or vice versa.
2288 .sp
2289 PCRE_ERROR_BADENDIANNESS (-29)
2290 .sp
2291 This error is given if a pattern that was compiled and saved is reloaded on a
2292 host with different endianness. The utility function
2293 \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2294 so that it runs on the new host.
2295 .sp
2296 PCRE_ERROR_JIT_BADOPTION
2297 .sp
2298 This error is returned when a pattern that was successfully studied using a JIT
2299 compile option is being matched, but the matching mode (partial or complete
2300 match) does not correspond to any JIT compilation mode. When the JIT fast path
2301 function is used, this error may be also given for invalid options. See the
2302 .\" HREF
2303 \fBpcrejit\fP
2304 .\"
2305 documentation for more details.
2306 .sp
2307 PCRE_ERROR_BADLENGTH (-32)
2308 .sp
2309 This error is given if \fBpcre_exec()\fP is called with a negative value for
2310 the \fIlength\fP argument.
2311 .P
2312 Error numbers -16 to -20, -22, and 30 are not used by \fBpcre_exec()\fP.
2313 .
2314 .
2315 .\" HTML <a name="badutf8reasons"></a>
2316 .SS "Reason codes for invalid UTF-8 strings"
2317 .rs
2318 .sp
2319 This section applies only to the 8-bit library. The corresponding information
2320 for the 16-bit and 32-bit libraries is given in the
2321 .\" HREF
2322 \fBpcre16\fP
2323 .\"
2324 and
2325 .\" HREF
2326 \fBpcre32\fP
2327 .\"
2328 pages.
2329 .P
2330 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2331 PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2332 least 2, the offset of the start of the invalid UTF-8 character is placed in
2333 the first output vector element (\fIovector[0]\fP) and a reason code is placed
2334 in the second element (\fIovector[1]\fP). The reason codes are given names in
2335 the \fBpcre.h\fP header file:
2336 .sp
2337 PCRE_UTF8_ERR1
2338 PCRE_UTF8_ERR2
2339 PCRE_UTF8_ERR3
2340 PCRE_UTF8_ERR4
2341 PCRE_UTF8_ERR5
2342 .sp
2343 The string ends with a truncated UTF-8 character; the code specifies how many
2344 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2345 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2346 allows for up to 6 bytes, and this is checked first; hence the possibility of
2347 4 or 5 missing bytes.
2348 .sp
2349 PCRE_UTF8_ERR6
2350 PCRE_UTF8_ERR7
2351 PCRE_UTF8_ERR8
2352 PCRE_UTF8_ERR9
2353 PCRE_UTF8_ERR10
2354 .sp
2355 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2356 character do not have the binary value 0b10 (that is, either the most
2357 significant bit is 0, or the next bit is 1).
2358 .sp
2359 PCRE_UTF8_ERR11
2360 PCRE_UTF8_ERR12
2361 .sp
2362 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2363 these code points are excluded by RFC 3629.
2364 .sp
2365 PCRE_UTF8_ERR13
2366 .sp
2367 A 4-byte character has a value greater than 0x10fff; these code points are
2368 excluded by RFC 3629.
2369 .sp
2370 PCRE_UTF8_ERR14
2371 .sp
2372 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2373 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2374 from UTF-8.
2375 .sp
2376 PCRE_UTF8_ERR15
2377 PCRE_UTF8_ERR16
2378 PCRE_UTF8_ERR17
2379 PCRE_UTF8_ERR18
2380 PCRE_UTF8_ERR19
2381 .sp
2382 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2383 value that can be represented by fewer bytes, which is invalid. For example,
2384 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2385 one byte.
2386 .sp
2387 PCRE_UTF8_ERR20
2388 .sp
2389 The two most significant bits of the first byte of a character have the binary
2390 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2391 byte can only validly occur as the second or subsequent byte of a multi-byte
2392 character.
2393 .sp
2394 PCRE_UTF8_ERR21
2395 .sp
2396 The first byte of a character has the value 0xfe or 0xff. These values can
2397 never occur in a valid UTF-8 string.
2398 .sp
2399 PCRE_UTF8_ERR22
2400 .sp
2401 This error code was formerly used when the presence of a so-called
2402 "non-character" caused an error. Unicode corrigendum #9 makes it clear that
2403 such characters should not cause a string to be rejected, and so this code is
2404 no longer in use and is never returned.
2405 .
2406 .
2407 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2408 .rs
2409 .sp
2410 .nf
2411 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2412 .B " int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,"
2413 .B " int \fIbuffersize\fP);"
2414 .sp
2415 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2416 .B " int \fIstringcount\fP, int \fIstringnumber\fP,"
2417 .B " const char **\fIstringptr\fP);"
2418 .sp
2419 .B int pcre_get_substring_list(const char *\fIsubject\fP,
2420 .B " int *\fIovector\fP, int \fIstringcount\fP, const char ***\fIlistptr\fP);"
2421 .fi
2422 .PP
2423 Captured substrings can be accessed directly by using the offsets returned by
2424 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2425 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2426 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2427 as new, separate, zero-terminated strings. These functions identify substrings
2428 by number. The next section describes functions for extracting named
2429 substrings.
2430 .P
2431 A substring that contains a binary zero is correctly extracted and has a
2432 further zero added on the end, but the result is not, of course, a C string.
2433 However, you can process such a string by referring to the length that is
2434 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2435 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2436 for handling strings containing binary zeros, because the end of the final
2437 string is not independently indicated.
2438 .P
2439 The first three arguments are the same for all three of these functions:
2440 \fIsubject\fP is the subject string that has just been successfully matched,
2441 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2442 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2443 captured by the match, including the substring that matched the entire regular
2444 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2445 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2446 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2447 number of elements in the vector divided by three.
2448 .P
2449 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2450 extract a single substring, whose number is given as \fIstringnumber\fP. A
2451 value of zero extracts the substring that matched the entire pattern, whereas
2452 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2453 the string is placed in \fIbuffer\fP, whose length is given by
2454 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2455 obtained via \fBpcre_malloc\fP, and its address is returned via
2456 \fIstringptr\fP. The yield of the function is the length of the string, not
2457 including the terminating zero, or one of these error codes:
2458 .sp
2459 PCRE_ERROR_NOMEMORY (-6)
2460 .sp
2461 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2462 memory failed for \fBpcre_get_substring()\fP.
2463 .sp
2464 PCRE_ERROR_NOSUBSTRING (-7)
2465 .sp
2466 There is no substring whose number is \fIstringnumber\fP.
2467 .P
2468 The \fBpcre_get_substring_list()\fP function extracts all available substrings
2469 and builds a list of pointers to them. All this is done in a single block of
2470 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2471 is returned via \fIlistptr\fP, which is also the start of the list of string
2472 pointers. The end of the list is marked by a NULL pointer. The yield of the
2473 function is zero if all went well, or the error code
2474 .sp
2475 PCRE_ERROR_NOMEMORY (-6)
2476 .sp
2477 if the attempt to get the memory block failed.
2478 .P
2479 When any of these functions encounter a substring that is unset, which can
2480 happen when capturing subpattern number \fIn+1\fP matches some part of the
2481 subject, but subpattern \fIn\fP has not been used at all, they return an empty
2482 string. This can be distinguished from a genuine zero-length substring by
2483 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2484 substrings.
2485 .P
2486 The two convenience functions \fBpcre_free_substring()\fP and
2487 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2488 a previous call of \fBpcre_get_substring()\fP or
2489 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2490 the function pointed to by \fBpcre_free\fP, which of course could be called
2491 directly from a C program. However, PCRE is used in some situations where it is
2492 linked via a special interface to another programming language that cannot use
2493 \fBpcre_free\fP directly; it is for these cases that the functions are
2494 provided.
2495 .
2496 .
2497 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2498 .rs
2499 .sp
2500 .nf
2501 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2502 .B " const char *\fIname\fP);"
2503 .sp
2504 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2505 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
2506 .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
2507 .B " char *\fIbuffer\fP, int \fIbuffersize\fP);"
2508 .sp
2509 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2510 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
2511 .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
2512 .B " const char **\fIstringptr\fP);"
2513 .fi
2514 .PP
2515 To extract a substring by name, you first have to find associated number.
2516 For example, for this pattern
2517 .sp
2518 (a+)b(?<xxx>\ed+)...
2519 .sp
2520 the number of the subpattern called "xxx" is 2. If the name is known to be
2521 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2522 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2523 pattern, and the second is the name. The yield of the function is the
2524 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2525 that name.
2526 .P
2527 Given the number, you can extract the substring directly, or use one of the
2528 functions described in the previous section. For convenience, there are also
2529 two functions that do the whole job.
2530 .P
2531 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2532 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2533 functions that extract by number. As these are described in the previous
2534 section, they are not re-described here. There are just two differences:
2535 .P
2536 First, instead of a substring number, a substring name is given. Second, there
2537 is an extra argument, given at the start, which is a pointer to the compiled
2538 pattern. This is needed in order to gain access to the name-to-number
2539 translation table.
2540 .P
2541 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2542 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2543 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2544 the behaviour may not be what you want (see the next section).
2545 .P
2546 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2547 subpatterns with the same number, as described in the
2548 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2549 .\" </a>
2550 section on duplicate subpattern numbers
2551 .\"
2552 in the
2553 .\" HREF
2554 \fBpcrepattern\fP
2555 .\"
2556 page, you cannot use names to distinguish the different subpatterns, because
2557 names are not included in the compiled code. The matching process uses only
2558 numbers. For this reason, the use of different names for subpatterns of the
2559 same number causes an error at compile time.
2560 .
2561 .
2562 .SH "DUPLICATE SUBPATTERN NAMES"
2563 .rs
2564 .sp
2565 .nf
2566 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2567 .B " const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);"
2568 .fi
2569 .PP
2570 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2571 are not required to be unique. (Duplicate names are always allowed for
2572 subpatterns with the same number, created by using the (?| feature. Indeed, if
2573 such subpatterns are named, they are required to use the same names.)
2574 .P
2575 Normally, patterns with duplicate names are such that in any one match, only
2576 one of the named subpatterns participates. An example is shown in the
2577 .\" HREF
2578 \fBpcrepattern\fP
2579 .\"
2580 documentation.
2581 .P
2582 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2583 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2584 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2585 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2586 returns one of the numbers that are associated with the name, but it is not
2587 defined which it is.
2588 .P
2589 If you want to get full details of all captured substrings for a given name,
2590 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2591 argument is the compiled pattern, and the second is the name. The third and
2592 fourth are pointers to variables which are updated by the function. After it
2593 has run, they point to the first and last entries in the name-to-number table
2594 for the given name. The function itself returns the length of each entry, or
2595 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2596 described above in the section entitled \fIInformation about a pattern\fP
2597 .\" HTML <a href="#infoaboutpattern">
2598 .\" </a>
2599 above.
2600 .\"
2601 Given all the relevant entries for the name, you can extract each of their
2602 numbers, and hence the captured data, if any.
2603 .
2604 .
2605 .SH "FINDING ALL POSSIBLE MATCHES"
2606 .rs
2607 .sp
2608 The traditional matching function uses a similar algorithm to Perl, which stops
2609 when it finds the first match, starting at a given point in the subject. If you
2610 want to find all possible matches, or the longest possible match, consider
2611 using the alternative matching function (see below) instead. If you cannot use
2612 the alternative function, but still need to find all possible matches, you
2613 can kludge it up by making use of the callout facility, which is described in
2614 the
2615 .\" HREF
2616 \fBpcrecallout\fP
2617 .\"
2618 documentation.
2619 .P
2620 What you have to do is to insert a callout right at the end of the pattern.
2621 When your callout function is called, extract and save the current matched
2622 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2623 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2624 will yield PCRE_ERROR_NOMATCH.
2625 .
2626 .
2627 .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2628 .rs
2629 .sp
2630 Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2631 stack, which in certain environments can be rather limited in size. Some users
2632 find it helpful to have an estimate of the amount of stack that is used by
2633 \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2634 .\" HREF
2635 \fBpcrestack\fP
2636 .\"
2637 documentation. The estimate that is output by \fBpcretest\fP when called with
2638 the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2639 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2640 .P
2641 Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2642 the negative error code PCRE_ERROR_NULL, but with this special combination of
2643 arguments, it returns instead a negative number whose absolute value is the
2644 approximate stack frame size in bytes. (A negative number is used so that it is
2645 clear that no match has happened.) The value is approximate because in some
2646 cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2647 additional variables on the stack.
2648 .P
2649 If PCRE has been compiled to use the heap instead of the stack for recursion,
2650 the value returned is the size of each block that is obtained from the heap.
2651 .
2652 .
2653 .\" HTML <a name="dfamatch"></a>
2654 .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2655 .rs
2656 .sp
2657 .nf
2658 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2659 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
2660 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
2661 .B " int *\fIworkspace\fP, int \fIwscount\fP);"
2662 .fi
2663 .P
2664 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2665 a compiled pattern, using a matching algorithm that scans the subject string
2666 just once, and does not backtrack. This has different characteristics to the
2667 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2668 patterns are not supported. Nevertheless, there are times when this kind of
2669 matching can be useful. For a discussion of the two matching algorithms, and a
2670 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2671 .\" HREF
2672 \fBpcrematching\fP
2673 .\"
2674 documentation.
2675 .P
2676 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2677 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2678 different way, and this is described below. The other common arguments are used
2679 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2680 here.
2681 .P
2682 The two additional arguments provide workspace for the function. The workspace
2683 vector should contain at least 20 elements. It is used for keeping track of
2684 multiple paths through the pattern tree. More workspace will be needed for
2685 patterns and subjects where there are a lot of potential matches.
2686 .P
2687 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2688 .sp
2689 int rc;
2690 int ovector[10];
2691 int wspace[20];
2692 rc = pcre_dfa_exec(
2693 re, /* result of pcre_compile() */
2694 NULL, /* we didn't study the pattern */
2695 "some string", /* the subject string */
2696 11, /* the length of the subject string */
2697 0, /* start at offset 0 in the subject */
2698 0, /* default options */
2699 ovector, /* vector of integers for substring information */
2700 10, /* number of elements (NOT size in bytes) */
2701 wspace, /* working space vector */
2702 20); /* number of elements (NOT size in bytes) */
2703 .
2704 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2705 .rs
2706 .sp
2707 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2708 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2709 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2710 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2711 PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2712 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2713 so their description is not repeated here.
2714 .sp
2715 PCRE_PARTIAL_HARD
2716 PCRE_PARTIAL_SOFT
2717 .sp
2718 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2719 details are slightly different. When PCRE_PARTIAL_HARD is set for
2720 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2721 is reached and there is still at least one matching possibility that requires
2722 additional characters. This happens even if some complete matches have also
2723 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2724 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2725 there have been no complete matches, but there is still at least one matching
2726 possibility. The portion of the string that was inspected when the longest
2727 partial match was found is set as the first matching string in both cases.
2728 There is a more detailed discussion of partial and multi-segment matching, with
2729 examples, in the
2730 .\" HREF
2731 \fBpcrepartial\fP
2732 .\"
2733 documentation.
2734 .sp
2735 PCRE_DFA_SHORTEST
2736 .sp
2737 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2738 soon as it has found one match. Because of the way the alternative algorithm
2739 works, this is necessarily the shortest possible match at the first possible
2740 matching point in the subject string.
2741 .sp
2742 PCRE_DFA_RESTART
2743 .sp
2744 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2745 again, with additional subject characters, and have it continue with the same
2746 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2747 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2748 before because data about the match so far is left in them after a partial
2749 match. There is more discussion of this facility in the
2750 .\" HREF
2751 \fBpcrepartial\fP
2752 .\"
2753 documentation.
2754 .
2755 .
2756 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2757 .rs
2758 .sp
2759 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2760 substring in the subject. Note, however, that all the matches from one run of
2761 the function start at the same point in the subject. The shorter matches are
2762 all initial substrings of the longer matches. For example, if the pattern
2763 .sp
2764 <.*>
2765 .sp
2766 is matched against the string
2767 .sp
2768 This is <something> <something else> <something further> no more
2769 .sp
2770 the three matched strings are
2771 .sp
2772 <something>
2773 <something> <something else>
2774 <something> <something else> <something further>
2775 .sp
2776 On success, the yield of the function is a number greater than zero, which is
2777 the number of matched substrings. The substrings themselves are returned in
2778 \fIovector\fP. Each string uses two elements; the first is the offset to the
2779 start, and the second is the offset to the end. In fact, all the strings have
2780 the same start offset. (Space could have been saved by giving this only once,
2781 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2782 returns data, even though the meaning of the strings is different.)
2783 .P
2784 The strings are returned in reverse order of length; that is, the longest
2785 matching string is given first. If there were too many matches to fit into
2786 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2787 the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2788 the entire \fIovector\fP for returning matched strings.
2789
2790 NOTE: PCRE's "auto-possessification" optimization usually applies to character
2791 repeats at the end of a pattern (as well as internally). For example, the
2792 pattern "a\ed+" is compiled as if it were "a\ed++" because there is no point
2793 even considering the possibility of backtracking into the repeated digits. For
2794 DFA matching, this means that only one possible match is found. If you really
2795 do want multiple matches in such cases, either use an ungreedy repeat
2796 ("a\ed+?") or set the PCRE_NO_AUTO_POSSESSIFY option when compiling.
2797 .
2798 .
2799 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2800 .rs
2801 .sp
2802 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2803 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2804 described
2805 .\" HTML <a href="#errorlist">
2806 .\" </a>
2807 above.
2808 .\"
2809 There are in addition the following errors that are specific to
2810 \fBpcre_dfa_exec()\fP:
2811 .sp
2812 PCRE_ERROR_DFA_UITEM (-16)
2813 .sp
2814 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2815 that it does not support, for instance, the use of \eC or a back reference.
2816 .sp
2817 PCRE_ERROR_DFA_UCOND (-17)
2818 .sp
2819 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2820 uses a back reference for the condition, or a test for recursion in a specific
2821 group. These are not supported.
2822 .sp
2823 PCRE_ERROR_DFA_UMLIMIT (-18)
2824 .sp
2825 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2826 block that contains a setting of the \fImatch_limit\fP or
2827 \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2828 meaningless for DFA matching).
2829 .sp
2830 PCRE_ERROR_DFA_WSSIZE (-19)
2831 .sp
2832 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2833 \fIworkspace\fP vector.
2834 .sp
2835 PCRE_ERROR_DFA_RECURSE (-20)
2836 .sp
2837 When a recursive subpattern is processed, the matching function calls itself
2838 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2839 error is given if the output vector is not large enough. This should be
2840 extremely rare, as a vector of size 1000 is used.
2841 .sp
2842 PCRE_ERROR_DFA_BADRESTART (-30)
2843 .sp
2844 When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2845 some plausibility checks are made on the contents of the workspace, which
2846 should contain data about the previous partial match. If any of these checks
2847 fail, this error is given.
2848 .
2849 .
2850 .SH "SEE ALSO"
2851 .rs
2852 .sp
2853 \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2854 \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2855 \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2856 \fBpcrestack\fP(3).
2857 .
2858 .
2859 .SH AUTHOR
2860 .rs
2861 .sp
2862 .nf
2863 Philip Hazel
2864 University Computing Service
2865 Cambridge CB2 3QH, England.
2866 .fi
2867 .
2868 .
2869 .SH REVISION
2870 .rs
2871 .sp
2872 .nf
2873 Last updated: 08 October 2013
2874 Copyright (c) 1997-2013 University of Cambridge.
2875 .fi

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5