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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1622 - (show annotations)
Fri Dec 18 17:08:02 2015 UTC (4 months, 2 weeks ago) by ph10
File size: 129129 byte(s)
Error occurred while calculating annotation data.
Remove no-longer-true statement from the documentation.
1 .TH PCREAPI 3 "18 December 2015" "PCRE 8.39"
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 .sp
120 .B int (*pcre_stack_guard)(void);
121 .fi
122 .
123 .
124 .SH "PCRE 8-BIT, 16-BIT, AND 32-BIT LIBRARIES"
125 .rs
126 .sp
127 As well as support for 8-bit character strings, PCRE also supports 16-bit
128 strings (from release 8.30) and 32-bit strings (from release 8.32), by means of
129 two additional libraries. They can be built as well as, or instead of, the
130 8-bit library. To avoid too much complication, this document describes the
131 8-bit versions of the functions, with only occasional references to the 16-bit
132 and 32-bit libraries.
133 .P
134 The 16-bit and 32-bit functions operate in the same way as their 8-bit
135 counterparts; they just use different data types for their arguments and
136 results, and their names start with \fBpcre16_\fP or \fBpcre32_\fP instead of
137 \fBpcre_\fP. For every option that has UTF8 in its name (for example,
138 PCRE_UTF8), there are corresponding 16-bit and 32-bit names with UTF8 replaced
139 by UTF16 or UTF32, respectively. This facility is in fact just cosmetic; the
140 16-bit and 32-bit option names define the same bit values.
141 .P
142 References to bytes and UTF-8 in this document should be read as references to
143 16-bit data units and UTF-16 when using the 16-bit library, or 32-bit data
144 units and UTF-32 when using the 32-bit library, unless specified otherwise.
145 More details of the specific differences for the 16-bit and 32-bit libraries
146 are given in the
147 .\" HREF
148 \fBpcre16\fP
149 .\"
150 and
151 .\" HREF
152 \fBpcre32\fP
153 .\"
154 pages.
155 .
156 .
157 .SH "PCRE API OVERVIEW"
158 .rs
159 .sp
160 PCRE has its own native API, which is described in this document. There are
161 also some wrapper functions (for the 8-bit library only) that correspond to the
162 POSIX regular expression API, but they do not give access to all the
163 functionality. They are described in the
164 .\" HREF
165 \fBpcreposix\fP
166 .\"
167 documentation. Both of these APIs define a set of C function calls. A C++
168 wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
169 documented in the
170 .\" HREF
171 \fBpcrecpp\fP
172 .\"
173 page.
174 .P
175 The native API C function prototypes are defined in the header file
176 \fBpcre.h\fP, and on Unix-like systems the (8-bit) library itself is called
177 \fBlibpcre\fP. It can normally be accessed by adding \fB-lpcre\fP to the
178 command for linking an application that uses PCRE. The header file defines the
179 macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
180 for the library. Applications can use these to include support for different
181 releases of PCRE.
182 .P
183 In a Windows environment, if you want to statically link an application program
184 against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
185 including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
186 \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
187 \fB__declspec(dllimport)\fP, with unwanted results.
188 .P
189 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
190 and \fBpcre_exec()\fP are used for compiling and matching regular expressions
191 in a Perl-compatible manner. A sample program that demonstrates the simplest
192 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
193 source distribution. A listing of this program is given in the
194 .\" HREF
195 \fBpcredemo\fP
196 .\"
197 documentation, and the
198 .\" HREF
199 \fBpcresample\fP
200 .\"
201 documentation describes how to compile and run it.
202 .P
203 Just-in-time compiler support is an optional feature of PCRE that can be built
204 in appropriate hardware environments. It greatly speeds up the matching
205 performance of many patterns. Simple programs can easily request that it be
206 used if available, by setting an option that is ignored when it is not
207 relevant. More complicated programs might need to make use of the functions
208 \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
209 \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
210 .P
211 From release 8.32 there is also a direct interface for JIT execution, which
212 gives improved performance. The JIT-specific functions are discussed in the
213 .\" HREF
214 \fBpcrejit\fP
215 .\"
216 documentation.
217 .P
218 A second matching function, \fBpcre_dfa_exec()\fP, which is not
219 Perl-compatible, is also provided. This uses a different algorithm for the
220 matching. The alternative algorithm finds all possible matches (at a given
221 point in the subject), and scans the subject just once (unless there are
222 lookbehind assertions). However, this algorithm does not return captured
223 substrings. A description of the two matching algorithms and their advantages
224 and disadvantages is given in the
225 .\" HREF
226 \fBpcrematching\fP
227 .\"
228 documentation.
229 .P
230 In addition to the main compiling and matching functions, there are convenience
231 functions for extracting captured substrings from a subject string that is
232 matched by \fBpcre_exec()\fP. They are:
233 .sp
234 \fBpcre_copy_substring()\fP
235 \fBpcre_copy_named_substring()\fP
236 \fBpcre_get_substring()\fP
237 \fBpcre_get_named_substring()\fP
238 \fBpcre_get_substring_list()\fP
239 \fBpcre_get_stringnumber()\fP
240 \fBpcre_get_stringtable_entries()\fP
241 .sp
242 \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
243 provided, to free the memory used for extracted strings.
244 .P
245 The function \fBpcre_maketables()\fP is used to build a set of character tables
246 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
247 or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
248 specialist use. Most commonly, no special tables are passed, in which case
249 internal tables that are generated when PCRE is built are used.
250 .P
251 The function \fBpcre_fullinfo()\fP is used to find out information about a
252 compiled pattern. The function \fBpcre_version()\fP returns a pointer to a
253 string containing the version of PCRE and its date of release.
254 .P
255 The function \fBpcre_refcount()\fP maintains a reference count in a data block
256 containing a compiled pattern. This is provided for the benefit of
257 object-oriented applications.
258 .P
259 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
260 the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
261 respectively. PCRE calls the memory management functions via these variables,
262 so a calling program can replace them if it wishes to intercept the calls. This
263 should be done before calling any PCRE functions.
264 .P
265 The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
266 indirections to memory management functions. These special functions are used
267 only when PCRE is compiled to use the heap for remembering data, instead of
268 recursive function calls, when running the \fBpcre_exec()\fP function. See the
269 .\" HREF
270 \fBpcrebuild\fP
271 .\"
272 documentation for details of how to do this. It is a non-standard way of
273 building PCRE, for use in environments that have limited stacks. Because of the
274 greater use of memory management, it runs more slowly. Separate functions are
275 provided so that special-purpose external code can be used for this case. When
276 used, these functions always allocate memory blocks of the same size. There is
277 a discussion about PCRE's stack usage in the
278 .\" HREF
279 \fBpcrestack\fP
280 .\"
281 documentation.
282 .P
283 The global variable \fBpcre_callout\fP initially contains NULL. It can be set
284 by the caller to a "callout" function, which PCRE will then call at specified
285 points during a matching operation. Details are given in the
286 .\" HREF
287 \fBpcrecallout\fP
288 .\"
289 documentation.
290 .P
291 The global variable \fBpcre_stack_guard\fP initially contains NULL. It can be
292 set by the caller to a function that is called by PCRE whenever it starts
293 to compile a parenthesized part of a pattern. When parentheses are nested, PCRE
294 uses recursive function calls, which use up the system stack. This function is
295 provided so that applications with restricted stacks can force a compilation
296 error if the stack runs out. The function should return zero if all is well, or
297 non-zero to force an error.
298 .
299 .
300 .\" HTML <a name="newlines"></a>
301 .SH NEWLINES
302 .rs
303 .sp
304 PCRE supports five different conventions for indicating line breaks in
305 strings: a single CR (carriage return) character, a single LF (linefeed)
306 character, the two-character sequence CRLF, any of the three preceding, or any
307 Unicode newline sequence. The Unicode newline sequences are the three just
308 mentioned, plus the single characters VT (vertical tab, U+000B), FF (form feed,
309 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
310 (paragraph separator, U+2029).
311 .P
312 Each of the first three conventions is used by at least one operating system as
313 its standard newline sequence. When PCRE is built, a default can be specified.
314 The default default is LF, which is the Unix standard. When PCRE is run, the
315 default can be overridden, either when a pattern is compiled, or when it is
316 matched.
317 .P
318 At compile time, the newline convention can be specified by the \fIoptions\fP
319 argument of \fBpcre_compile()\fP, or it can be specified by special text at the
320 start of the pattern itself; this overrides any other settings. See the
321 .\" HREF
322 \fBpcrepattern\fP
323 .\"
324 page for details of the special character sequences.
325 .P
326 In the PCRE documentation the word "newline" is used to mean "the character or
327 pair of characters that indicate a line break". The choice of newline
328 convention affects the handling of the dot, circumflex, and dollar
329 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
330 recognized line ending sequence, the match position advancement for a
331 non-anchored pattern. There is more detail about this in the
332 .\" HTML <a href="#execoptions">
333 .\" </a>
334 section on \fBpcre_exec()\fP options
335 .\"
336 below.
337 .P
338 The choice of newline convention does not affect the interpretation of
339 the \en or \er escape sequences, nor does it affect what \eR matches, which is
340 controlled in a similar way, but by separate options.
341 .
342 .
343 .SH MULTITHREADING
344 .rs
345 .sp
346 The PCRE functions can be used in multi-threading applications, with the
347 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
348 \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
349 callout and stack-checking functions pointed to by \fBpcre_callout\fP and
350 \fBpcre_stack_guard\fP, are shared by all threads.
351 .P
352 The compiled form of a regular expression is not altered during matching, so
353 the same compiled pattern can safely be used by several threads at once.
354 .P
355 If the just-in-time optimization feature is being used, it needs separate
356 memory stack areas for each thread. See the
357 .\" HREF
358 \fBpcrejit\fP
359 .\"
360 documentation for more details.
361 .
362 .
363 .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
364 .rs
365 .sp
366 The compiled form of a regular expression can be saved and re-used at a later
367 time, possibly by a different program, and even on a host other than the one on
368 which it was compiled. Details are given in the
369 .\" HREF
370 \fBpcreprecompile\fP
371 .\"
372 documentation, which includes a description of the
373 \fBpcre_pattern_to_host_byte_order()\fP function. However, compiling a regular
374 expression with one version of PCRE for use with a different version is not
375 guaranteed to work and may cause crashes.
376 .
377 .
378 .SH "CHECKING BUILD-TIME OPTIONS"
379 .rs
380 .sp
381 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
382 .PP
383 The function \fBpcre_config()\fP makes it possible for a PCRE client to
384 discover which optional features have been compiled into the PCRE library. The
385 .\" HREF
386 \fBpcrebuild\fP
387 .\"
388 documentation has more details about these optional features.
389 .P
390 The first argument for \fBpcre_config()\fP is an integer, specifying which
391 information is required; the second argument is a pointer to a variable into
392 which the information is placed. The returned value is zero on success, or the
393 negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
394 not recognized. The following information is available:
395 .sp
396 PCRE_CONFIG_UTF8
397 .sp
398 The output is an integer that is set to one if UTF-8 support is available;
399 otherwise it is set to zero. This value should normally be given to the 8-bit
400 version of this function, \fBpcre_config()\fP. If it is given to the 16-bit
401 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
402 .sp
403 PCRE_CONFIG_UTF16
404 .sp
405 The output is an integer that is set to one if UTF-16 support is available;
406 otherwise it is set to zero. This value should normally be given to the 16-bit
407 version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
408 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
409 .sp
410 PCRE_CONFIG_UTF32
411 .sp
412 The output is an integer that is set to one if UTF-32 support is available;
413 otherwise it is set to zero. This value should normally be given to the 32-bit
414 version of this function, \fBpcre32_config()\fP. If it is given to the 8-bit
415 or 16-bit version of this function, the result is PCRE_ERROR_BADOPTION.
416 .sp
417 PCRE_CONFIG_UNICODE_PROPERTIES
418 .sp
419 The output is an integer that is set to one if support for Unicode character
420 properties is available; otherwise it is set to zero.
421 .sp
422 PCRE_CONFIG_JIT
423 .sp
424 The output is an integer that is set to one if support for just-in-time
425 compiling is available; otherwise it is set to zero.
426 .sp
427 PCRE_CONFIG_JITTARGET
428 .sp
429 The output is a pointer to a zero-terminated "const char *" string. If JIT
430 support is available, the string contains the name of the architecture for
431 which the JIT compiler is configured, for example "x86 32bit (little endian +
432 unaligned)". If JIT support is not available, the result is NULL.
433 .sp
434 PCRE_CONFIG_NEWLINE
435 .sp
436 The output is an integer whose value specifies the default character sequence
437 that is recognized as meaning "newline". The values that are supported in
438 ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for
439 ANYCRLF, and -1 for ANY. In EBCDIC environments, CR, ANYCRLF, and ANY yield the
440 same values. However, the value for LF is normally 21, though some EBCDIC
441 environments use 37. The corresponding values for CRLF are 3349 and 3365. The
442 default should normally correspond to the standard sequence for your operating
443 system.
444 .sp
445 PCRE_CONFIG_BSR
446 .sp
447 The output is an integer whose value indicates what character sequences the \eR
448 escape sequence matches by default. A value of 0 means that \eR matches any
449 Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
450 or CRLF. The default can be overridden when a pattern is compiled or matched.
451 .sp
452 PCRE_CONFIG_LINK_SIZE
453 .sp
454 The output is an integer that contains the number of bytes used for internal
455 linkage in compiled regular expressions. For the 8-bit library, the value can
456 be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
457 a number of bytes. For the 32-bit library, the value is either 2 or 4 and is
458 still a number of bytes. The default value of 2 is sufficient for all but the
459 most massive patterns, since it allows the compiled pattern to be up to 64K in
460 size. Larger values allow larger regular expressions to be compiled, at the
461 expense of slower matching.
462 .sp
463 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
464 .sp
465 The output is an integer that contains the threshold above which the POSIX
466 interface uses \fBmalloc()\fP for output vectors. Further details are given in
467 the
468 .\" HREF
469 \fBpcreposix\fP
470 .\"
471 documentation.
472 .sp
473 PCRE_CONFIG_PARENS_LIMIT
474 .sp
475 The output is a long integer that gives the maximum depth of nesting of
476 parentheses (of any kind) in a pattern. This limit is imposed to cap the amount
477 of system stack used when a pattern is compiled. It is specified when PCRE is
478 built; the default is 250. This limit does not take into account the stack that
479 may already be used by the calling application. For finer control over
480 compilation stack usage, you can set a pointer to an external checking function
481 in \fBpcre_stack_guard\fP.
482 .sp
483 PCRE_CONFIG_MATCH_LIMIT
484 .sp
485 The output is a long integer that gives the default limit for the number of
486 internal matching function calls in a \fBpcre_exec()\fP execution. Further
487 details are given with \fBpcre_exec()\fP below.
488 .sp
489 PCRE_CONFIG_MATCH_LIMIT_RECURSION
490 .sp
491 The output is a long integer that gives the default limit for the depth of
492 recursion when calling the internal matching function in a \fBpcre_exec()\fP
493 execution. Further details are given with \fBpcre_exec()\fP below.
494 .sp
495 PCRE_CONFIG_STACKRECURSE
496 .sp
497 The output is an integer that is set to one if internal recursion when running
498 \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
499 to remember their state. This is the usual way that PCRE is compiled. The
500 output is zero if PCRE was compiled to use blocks of data on the heap instead
501 of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
502 \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
503 avoiding the use of the stack.
504 .
505 .
506 .SH "COMPILING A PATTERN"
507 .rs
508 .sp
509 .nf
510 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
511 .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
512 .B " const unsigned char *\fItableptr\fP);"
513 .sp
514 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
515 .B " int *\fIerrorcodeptr\fP,"
516 .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
517 .B " const unsigned char *\fItableptr\fP);"
518 .fi
519 .P
520 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
521 called to compile a pattern into an internal form. The only difference between
522 the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
523 \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
524 too much repetition, we refer just to \fBpcre_compile()\fP below, but the
525 information applies equally to \fBpcre_compile2()\fP.
526 .P
527 The pattern is a C string terminated by a binary zero, and is passed in the
528 \fIpattern\fP argument. A pointer to a single block of memory that is obtained
529 via \fBpcre_malloc\fP is returned. This contains the compiled code and related
530 data. The \fBpcre\fP type is defined for the returned block; this is a typedef
531 for a structure whose contents are not externally defined. It is up to the
532 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
533 .P
534 Although the compiled code of a PCRE regex is relocatable, that is, it does not
535 depend on memory location, the complete \fBpcre\fP data block is not
536 fully relocatable, because it may contain a copy of the \fItableptr\fP
537 argument, which is an address (see below).
538 .P
539 The \fIoptions\fP argument contains various bit settings that affect the
540 compilation. It should be zero if no options are required. The available
541 options are described below. Some of them (in particular, those that are
542 compatible with Perl, but some others as well) can also be set and unset from
543 within the pattern (see the detailed description in the
544 .\" HREF
545 \fBpcrepattern\fP
546 .\"
547 documentation). For those options that can be different in different parts of
548 the pattern, the contents of the \fIoptions\fP argument specifies their
549 settings at the start of compilation and execution. The PCRE_ANCHORED,
550 PCRE_BSR_\fIxxx\fP, PCRE_NEWLINE_\fIxxx\fP, PCRE_NO_UTF8_CHECK, and
551 PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
552 compile time.
553 .P
554 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
555 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
556 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
557 error message. This is a static string that is part of the library. You must
558 not try to free it. Normally, the offset from the start of the pattern to the
559 data unit that was being processed when the error was discovered is placed in
560 the variable pointed to by \fIerroffset\fP, which must not be NULL (if it is,
561 an immediate error is given). However, for an invalid UTF-8 or UTF-16 string,
562 the offset is that of the first data unit of the failing character.
563 .P
564 Some errors are not detected until the whole pattern has been scanned; in these
565 cases, the offset passed back is the length of the pattern. Note that the
566 offset is in data units, not characters, even in a UTF mode. It may sometimes
567 point into the middle of a UTF-8 or UTF-16 character.
568 .P
569 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
570 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
571 returned via this argument in the event of an error. This is in addition to the
572 textual error message. Error codes and messages are listed below.
573 .P
574 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
575 character tables that are built when PCRE is compiled, using the default C
576 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
577 call to \fBpcre_maketables()\fP. This value is stored with the compiled
578 pattern, and used again by \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP when the
579 pattern is matched. For more discussion, see the section on locale support
580 below.
581 .P
582 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
583 .sp
584 pcre *re;
585 const char *error;
586 int erroffset;
587 re = pcre_compile(
588 "^A.*Z", /* the pattern */
589 0, /* default options */
590 &error, /* for error message */
591 &erroffset, /* for error offset */
592 NULL); /* use default character tables */
593 .sp
594 The following names for option bits are defined in the \fBpcre.h\fP header
595 file:
596 .sp
597 PCRE_ANCHORED
598 .sp
599 If this bit is set, the pattern is forced to be "anchored", that is, it is
600 constrained to match only at the first matching point in the string that is
601 being searched (the "subject string"). This effect can also be achieved by
602 appropriate constructs in the pattern itself, which is the only way to do it in
603 Perl.
604 .sp
605 PCRE_AUTO_CALLOUT
606 .sp
607 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
608 all with number 255, before each pattern item. For discussion of the callout
609 facility, see the
610 .\" HREF
611 \fBpcrecallout\fP
612 .\"
613 documentation.
614 .sp
615 PCRE_BSR_ANYCRLF
616 PCRE_BSR_UNICODE
617 .sp
618 These options (which are mutually exclusive) control what the \eR escape
619 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
620 match any Unicode newline sequence. The default is specified when PCRE is
621 built. It can be overridden from within the pattern, or by setting an option
622 when a compiled pattern is matched.
623 .sp
624 PCRE_CASELESS
625 .sp
626 If this bit is set, letters in the pattern match both upper and lower case
627 letters. It is equivalent to Perl's /i option, and it can be changed within a
628 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
629 concept of case for characters whose values are less than 128, so caseless
630 matching is always possible. For characters with higher values, the concept of
631 case is supported if PCRE is compiled with Unicode property support, but not
632 otherwise. If you want to use caseless matching for characters 128 and above,
633 you must ensure that PCRE is compiled with Unicode property support as well as
634 with UTF-8 support.
635 .sp
636 PCRE_DOLLAR_ENDONLY
637 .sp
638 If this bit is set, a dollar metacharacter in the pattern matches only at the
639 end of the subject string. Without this option, a dollar also matches
640 immediately before a newline at the end of the string (but not before any other
641 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
642 There is no equivalent to this option in Perl, and no way to set it within a
643 pattern.
644 .sp
645 PCRE_DOTALL
646 .sp
647 If this bit is set, a dot metacharacter in the pattern matches a character of
648 any value, including one that indicates a newline. However, it only ever
649 matches one character, even if newlines are coded as CRLF. Without this option,
650 a dot does not match when the current position is at a newline. This option is
651 equivalent to Perl's /s option, and it can be changed within a pattern by a
652 (?s) option setting. A negative class such as [^a] always matches newline
653 characters, independent of the setting of this option.
654 .sp
655 PCRE_DUPNAMES
656 .sp
657 If this bit is set, names used to identify capturing subpatterns need not be
658 unique. This can be helpful for certain types of pattern when it is known that
659 only one instance of the named subpattern can ever be matched. There are more
660 details of named subpatterns below; see also the
661 .\" HREF
662 \fBpcrepattern\fP
663 .\"
664 documentation.
665 .sp
666 PCRE_EXTENDED
667 .sp
668 If this bit is set, most white space characters in the pattern are totally
669 ignored except when escaped or inside a character class. However, white space
670 is not allowed within sequences such as (?> that introduce various
671 parenthesized subpatterns, nor within a numerical quantifier such as {1,3}.
672 However, ignorable white space is permitted between an item and a following
673 quantifier and between a quantifier and a following + that indicates
674 possessiveness.
675 .P
676 White space did not used to include the VT character (code 11), because Perl
677 did not treat this character as white space. However, Perl changed at release
678 5.18, so PCRE followed at release 8.34, and VT is now treated as white space.
679 .P
680 PCRE_EXTENDED also causes characters between an unescaped # outside a character
681 class and the next newline, inclusive, to be ignored. PCRE_EXTENDED is
682 equivalent to Perl's /x option, and it can be changed within a pattern by a
683 (?x) option setting.
684 .P
685 Which characters are interpreted as newlines is controlled by the options
686 passed to \fBpcre_compile()\fP or by a special sequence at the start of the
687 pattern, as described in the section entitled
688 .\" HTML <a href="pcrepattern.html#newlines">
689 .\" </a>
690 "Newline conventions"
691 .\"
692 in the \fBpcrepattern\fP documentation. Note that the end of this type of
693 comment is a literal newline sequence in the pattern; escape sequences that
694 happen to represent a newline do not count.
695 .P
696 This option makes it possible to include comments inside complicated patterns.
697 Note, however, that this applies only to data characters. White space characters
698 may never appear within special character sequences in a pattern, for example
699 within the sequence (?( that introduces a conditional subpattern.
700 .sp
701 PCRE_EXTRA
702 .sp
703 This option was invented in order to turn on additional functionality of PCRE
704 that is incompatible with Perl, but it is currently of very little use. When
705 set, any backslash in a pattern that is followed by a letter that has no
706 special meaning causes an error, thus reserving these combinations for future
707 expansion. By default, as in Perl, a backslash followed by a letter with no
708 special meaning is treated as a literal. (Perl can, however, be persuaded to
709 give an error for this, by running it with the -w option.) There are at present
710 no other features controlled by this option. It can also be set by a (?X)
711 option setting within a pattern.
712 .sp
713 PCRE_FIRSTLINE
714 .sp
715 If this option is set, an unanchored pattern is required to match before or at
716 the first newline in the subject string, though the matched text may continue
717 over the newline.
718 .sp
719 PCRE_JAVASCRIPT_COMPAT
720 .sp
721 If this option is set, PCRE's behaviour is changed in some ways so that it is
722 compatible with JavaScript rather than Perl. The changes are as follows:
723 .P
724 (1) A lone closing square bracket in a pattern causes a compile-time error,
725 because this is illegal in JavaScript (by default it is treated as a data
726 character). Thus, the pattern AB]CD becomes illegal when this option is set.
727 .P
728 (2) At run time, a back reference to an unset subpattern group matches an empty
729 string (by default this causes the current matching alternative to fail). A
730 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
731 an "a" in the subject), whereas it fails by default, for Perl compatibility.
732 .P
733 (3) \eU matches an upper case "U" character; by default \eU causes a compile
734 time error (Perl uses \eU to upper case subsequent characters).
735 .P
736 (4) \eu matches a lower case "u" character unless it is followed by four
737 hexadecimal digits, in which case the hexadecimal number defines the code point
738 to match. By default, \eu causes a compile time error (Perl uses it to upper
739 case the following character).
740 .P
741 (5) \ex matches a lower case "x" character unless it is followed by two
742 hexadecimal digits, in which case the hexadecimal number defines the code point
743 to match. By default, as in Perl, a hexadecimal number is always expected after
744 \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
745 binary zero character followed by z).
746 .sp
747 PCRE_MULTILINE
748 .sp
749 By default, for the purposes of matching "start of line" and "end of line",
750 PCRE treats the subject string as consisting of a single line of characters,
751 even if it actually contains newlines. The "start of line" metacharacter (^)
752 matches only at the start of the string, and the "end of line" metacharacter
753 ($) matches only at the end of the string, or before a terminating newline
754 (except when PCRE_DOLLAR_ENDONLY is set). Note, however, that unless
755 PCRE_DOTALL is set, the "any character" metacharacter (.) does not match at a
756 newline. This behaviour (for ^, $, and dot) is the same as Perl.
757 .P
758 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
759 match immediately following or immediately before internal newlines in the
760 subject string, respectively, as well as at the very start and end. This is
761 equivalent to Perl's /m option, and it can be changed within a pattern by a
762 (?m) option setting. If there are no newlines in a subject string, or no
763 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
764 .sp
765 PCRE_NEVER_UTF
766 .sp
767 This option locks out interpretation of the pattern as UTF-8 (or UTF-16 or
768 UTF-32 in the 16-bit and 32-bit libraries). In particular, it prevents the
769 creator of the pattern from switching to UTF interpretation by starting the
770 pattern with (*UTF). This may be useful in applications that process patterns
771 from external sources. The combination of PCRE_UTF8 and PCRE_NEVER_UTF also
772 causes an error.
773 .sp
774 PCRE_NEWLINE_CR
775 PCRE_NEWLINE_LF
776 PCRE_NEWLINE_CRLF
777 PCRE_NEWLINE_ANYCRLF
778 PCRE_NEWLINE_ANY
779 .sp
780 These options override the default newline definition that was chosen when PCRE
781 was built. Setting the first or the second specifies that a newline is
782 indicated by a single character (CR or LF, respectively). Setting
783 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
784 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
785 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
786 that any Unicode newline sequence should be recognized.
787 .P
788 In an ASCII/Unicode environment, the Unicode newline sequences are the three
789 just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
790 feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
791 (paragraph separator, U+2029). For the 8-bit library, the last two are
792 recognized only in UTF-8 mode.
793 .P
794 When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
795 CR is 0x0d, the same as ASCII. However, the character code for LF is normally
796 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
797 not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
798 less than 256. For more details, see the
799 .\" HREF
800 \fBpcrebuild\fP
801 .\"
802 documentation.
803 .P
804 The newline setting in the options word uses three bits that are treated
805 as a number, giving eight possibilities. Currently only six are used (default
806 plus the five values above). This means that if you set more than one newline
807 option, the combination may or may not be sensible. For example,
808 PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
809 other combinations may yield unused numbers and cause an error.
810 .P
811 The only time that a line break in a pattern is specially recognized when
812 compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
813 and so are ignored in this mode. Also, an unescaped # outside a character class
814 indicates a comment that lasts until after the next line break sequence. In
815 other circumstances, line break sequences in patterns are treated as literal
816 data.
817 .P
818 The newline option that is set at compile time becomes the default that is used
819 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
820 .sp
821 PCRE_NO_AUTO_CAPTURE
822 .sp
823 If this option is set, it disables the use of numbered capturing parentheses in
824 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
825 were followed by ?: but named parentheses can still be used for capturing (and
826 they acquire numbers in the usual way). There is no equivalent of this option
827 in Perl.
828 .sp
829 PCRE_NO_AUTO_POSSESS
830 .sp
831 If this option is set, it disables "auto-possessification". This is an
832 optimization that, for example, turns a+b into a++b in order to avoid
833 backtracks into a+ that can never be successful. However, if callouts are in
834 use, auto-possessification means that some of them are never taken. You can set
835 this option if you want the matching functions to do a full unoptimized search
836 and run all the callouts, but it is mainly provided for testing purposes.
837 .sp
838 PCRE_NO_START_OPTIMIZE
839 .sp
840 This is an option that acts at matching time; that is, it is really an option
841 for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
842 it is remembered with the compiled pattern and assumed at matching time. This
843 is necessary if you want to use JIT execution, because the JIT compiler needs
844 to know whether or not this option is set. For details see the discussion of
845 PCRE_NO_START_OPTIMIZE
846 .\" HTML <a href="#execoptions">
847 .\" </a>
848 below.
849 .\"
850 .sp
851 PCRE_UCP
852 .sp
853 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
854 \ew, and some of the POSIX character classes. By default, only ASCII characters
855 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
856 classify characters. More details are given in the section on
857 .\" HTML <a href="pcre.html#genericchartypes">
858 .\" </a>
859 generic character types
860 .\"
861 in the
862 .\" HREF
863 \fBpcrepattern\fP
864 .\"
865 page. If you set PCRE_UCP, matching one of the items it affects takes much
866 longer. The option is available only if PCRE has been compiled with Unicode
867 property support.
868 .sp
869 PCRE_UNGREEDY
870 .sp
871 This option inverts the "greediness" of the quantifiers so that they are not
872 greedy by default, but become greedy if followed by "?". It is not compatible
873 with Perl. It can also be set by a (?U) option setting within the pattern.
874 .sp
875 PCRE_UTF8
876 .sp
877 This option causes PCRE to regard both the pattern and the subject as strings
878 of UTF-8 characters instead of single-byte strings. However, it is available
879 only when PCRE is built to include UTF support. If not, the use of this option
880 provokes an error. Details of how this option changes the behaviour of PCRE are
881 given in the
882 .\" HREF
883 \fBpcreunicode\fP
884 .\"
885 page.
886 .sp
887 PCRE_NO_UTF8_CHECK
888 .sp
889 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
890 automatically checked. There is a discussion about the
891 .\" HTML <a href="pcreunicode.html#utf8strings">
892 .\" </a>
893 validity of UTF-8 strings
894 .\"
895 in the
896 .\" HREF
897 \fBpcreunicode\fP
898 .\"
899 page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
900 error. If you already know that your pattern is valid, and you want to skip
901 this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
902 When it is set, the effect of passing an invalid UTF-8 string as a pattern is
903 undefined. It may cause your program to crash or loop. Note that this option
904 can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
905 the validity checking of subject strings only. If the same string is being
906 matched many times, the option can be safely set for the second and subsequent
907 matchings to improve performance.
908 .
909 .
910 .SH "COMPILATION ERROR CODES"
911 .rs
912 .sp
913 The following table lists the error codes than may be returned by
914 \fBpcre_compile2()\fP, along with the error messages that may be returned by
915 both compiling functions. Note that error messages are always 8-bit ASCII
916 strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
917 have fallen out of use. To avoid confusion, they have not been re-used.
918 .sp
919 0 no error
920 1 \e at end of pattern
921 2 \ec at end of pattern
922 3 unrecognized character follows \e
923 4 numbers out of order in {} quantifier
924 5 number too big in {} quantifier
925 6 missing terminating ] for character class
926 7 invalid escape sequence in character class
927 8 range out of order in character class
928 9 nothing to repeat
929 10 [this code is not in use]
930 11 internal error: unexpected repeat
931 12 unrecognized character after (? or (?-
932 13 POSIX named classes are supported only within a class
933 14 missing )
934 15 reference to non-existent subpattern
935 16 erroffset passed as NULL
936 17 unknown option bit(s) set
937 18 missing ) after comment
938 19 [this code is not in use]
939 20 regular expression is too large
940 21 failed to get memory
941 22 unmatched parentheses
942 23 internal error: code overflow
943 24 unrecognized character after (?<
944 25 lookbehind assertion is not fixed length
945 26 malformed number or name after (?(
946 27 conditional group contains more than two branches
947 28 assertion expected after (?(
948 29 (?R or (?[+-]digits must be followed by )
949 30 unknown POSIX class name
950 31 POSIX collating elements are not supported
951 32 this version of PCRE is compiled without UTF support
952 33 [this code is not in use]
953 34 character value in \ex{} or \eo{} is too large
954 35 invalid condition (?(0)
955 36 \eC not allowed in lookbehind assertion
956 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
957 38 number after (?C is > 255
958 39 closing ) for (?C expected
959 40 recursive call could loop indefinitely
960 41 unrecognized character after (?P
961 42 syntax error in subpattern name (missing terminator)
962 43 two named subpatterns have the same name
963 44 invalid UTF-8 string (specifically UTF-8)
964 45 support for \eP, \ep, and \eX has not been compiled
965 46 malformed \eP or \ep sequence
966 47 unknown property name after \eP or \ep
967 48 subpattern name is too long (maximum 32 characters)
968 49 too many named subpatterns (maximum 10000)
969 50 [this code is not in use]
970 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
971 52 internal error: overran compiling workspace
972 53 internal error: previously-checked referenced subpattern
973 not found
974 54 DEFINE group contains more than one branch
975 55 repeating a DEFINE group is not allowed
976 56 inconsistent NEWLINE options
977 57 \eg is not followed by a braced, angle-bracketed, or quoted
978 name/number or by a plain number
979 58 a numbered reference must not be zero
980 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
981 60 (*VERB) not recognized or malformed
982 61 number is too big
983 62 subpattern name expected
984 63 digit expected after (?+
985 64 ] is an invalid data character in JavaScript compatibility mode
986 65 different names for subpatterns of the same number are
987 not allowed
988 66 (*MARK) must have an argument
989 67 this version of PCRE is not compiled with Unicode property
990 support
991 68 \ec must be followed by an ASCII character
992 69 \ek is not followed by a braced, angle-bracketed, or quoted name
993 70 internal error: unknown opcode in find_fixedlength()
994 71 \eN is not supported in a class
995 72 too many forward references
996 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
997 74 invalid UTF-16 string (specifically UTF-16)
998 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
999 76 character value in \eu.... sequence is too large
1000 77 invalid UTF-32 string (specifically UTF-32)
1001 78 setting UTF is disabled by the application
1002 79 non-hex character in \ex{} (closing brace missing?)
1003 80 non-octal character in \eo{} (closing brace missing?)
1004 81 missing opening brace after \eo
1005 82 parentheses are too deeply nested
1006 83 invalid range in character class
1007 84 group name must start with a non-digit
1008 85 parentheses are too deeply nested (stack check)
1009 .sp
1010 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
1011 be used if the limits were changed when PCRE was built.
1012 .
1013 .
1014 .\" HTML <a name="studyingapattern"></a>
1015 .SH "STUDYING A PATTERN"
1016 .rs
1017 .sp
1018 .nf
1019 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
1020 .B " const char **\fIerrptr\fP);"
1021 .fi
1022 .PP
1023 If a compiled pattern is going to be used several times, it is worth spending
1024 more time analyzing it in order to speed up the time taken for matching. The
1025 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
1026 argument. If studying the pattern produces additional information that will
1027 help speed up matching, \fBpcre_study()\fP returns a pointer to a
1028 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
1029 results of the study.
1030 .P
1031 The returned value from \fBpcre_study()\fP can be passed directly to
1032 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
1033 also contains other fields that can be set by the caller before the block is
1034 passed; these are described
1035 .\" HTML <a href="#extradata">
1036 .\" </a>
1037 below
1038 .\"
1039 in the section on matching a pattern.
1040 .P
1041 If studying the pattern does not produce any useful information,
1042 \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
1043 calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
1044 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
1045 if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
1046 returns a \fBpcre_extra\fP block even if studying did not find any additional
1047 information. It may still return NULL, however, if an error occurs in
1048 \fBpcre_study()\fP.
1049 .P
1050 The second argument of \fBpcre_study()\fP contains option bits. There are three
1051 further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1052 .sp
1053 PCRE_STUDY_JIT_COMPILE
1054 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1055 PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1056 .sp
1057 If any of these are set, and the just-in-time compiler is available, the
1058 pattern is further compiled into machine code that executes much faster than
1059 the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
1060 compiler is not available, these options are ignored. All undefined bits in the
1061 \fIoptions\fP argument must be zero.
1062 .P
1063 JIT compilation is a heavyweight optimization. It can take some time for
1064 patterns to be analyzed, and for one-off matches and simple patterns the
1065 benefit of faster execution might be offset by a much slower study time.
1066 Not all patterns can be optimized by the JIT compiler. For those that cannot be
1067 handled, matching automatically falls back to the \fBpcre_exec()\fP
1068 interpreter. For more details, see the
1069 .\" HREF
1070 \fBpcrejit\fP
1071 .\"
1072 documentation.
1073 .P
1074 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
1075 studying succeeds (even if no data is returned), the variable it points to is
1076 set to NULL. Otherwise it is set to point to a textual error message. This is a
1077 static string that is part of the library. You must not try to free it. You
1078 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
1079 sure that it has run successfully.
1080 .P
1081 When you are finished with a pattern, you can free the memory used for the
1082 study data by calling \fBpcre_free_study()\fP. This function was added to the
1083 API for release 8.20. For earlier versions, the memory could be freed with
1084 \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1085 where JIT optimization is not used, but it is advisable to change to the new
1086 function when convenient.
1087 .P
1088 This is a typical way in which \fBpcre_study\fP() is used (except that in a
1089 real application there should be tests for errors):
1090 .sp
1091 int rc;
1092 pcre *re;
1093 pcre_extra *sd;
1094 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1095 sd = pcre_study(
1096 re, /* result of pcre_compile() */
1097 0, /* no options */
1098 &error); /* set to NULL or points to a message */
1099 rc = pcre_exec( /* see below for details of pcre_exec() options */
1100 re, sd, "subject", 7, 0, 0, ovector, 30);
1101 ...
1102 pcre_free_study(sd);
1103 pcre_free(re);
1104 .sp
1105 Studying a pattern does two things: first, a lower bound for the length of
1106 subject string that is needed to match the pattern is computed. This does not
1107 mean that there are any strings of that length that match, but it does
1108 guarantee that no shorter strings match. The value is used to avoid wasting
1109 time by trying to match strings that are shorter than the lower bound. You can
1110 find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
1111 .P
1112 Studying a pattern is also useful for non-anchored patterns that do not have a
1113 single fixed starting character. A bitmap of possible starting bytes is
1114 created. This speeds up finding a position in the subject at which to start
1115 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1116 In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1117 .P
1118 These two optimizations apply to both \fBpcre_exec()\fP and
1119 \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1120 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option.
1121 You might want to do this if your pattern contains callouts or (*MARK) and you
1122 want to make use of these facilities in cases where matching fails.
1123 .P
1124 PCRE_NO_START_OPTIMIZE can be specified at either compile time or execution
1125 time. However, if PCRE_NO_START_OPTIMIZE is passed to \fBpcre_exec()\fP, (that
1126 is, after any JIT compilation has happened) JIT execution is disabled. For JIT
1127 execution to work with PCRE_NO_START_OPTIMIZE, the option must be set at
1128 compile time.
1129 .P
1130 There is a longer discussion of PCRE_NO_START_OPTIMIZE
1131 .\" HTML <a href="#execoptions">
1132 .\" </a>
1133 below.
1134 .\"
1135 .
1136 .
1137 .\" HTML <a name="localesupport"></a>
1138 .SH "LOCALE SUPPORT"
1139 .rs
1140 .sp
1141 PCRE handles caseless matching, and determines whether characters are letters,
1142 digits, or whatever, by reference to a set of tables, indexed by character
1143 code point. When running in UTF-8 mode, or in the 16- or 32-bit libraries, this
1144 applies only to characters with code points less than 256. By default,
1145 higher-valued code points never match escapes such as \ew or \ed. However, if
1146 PCRE is built with Unicode property support, all characters can be tested with
1147 \ep and \eP, or, alternatively, the PCRE_UCP option can be set when a pattern
1148 is compiled; this causes \ew and friends to use Unicode property support
1149 instead of the built-in tables.
1150 .P
1151 The use of locales with Unicode is discouraged. If you are handling characters
1152 with code points greater than 128, you should either use Unicode support, or
1153 use locales, but not try to mix the two.
1154 .P
1155 PCRE contains an internal set of tables that are used when the final argument
1156 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1157 Normally, the internal tables recognize only ASCII characters. However, when
1158 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1159 default "C" locale of the local system, which may cause them to be different.
1160 .P
1161 The internal tables can always be overridden by tables supplied by the
1162 application that calls PCRE. These may be created in a different locale from
1163 the default. As more and more applications change to using Unicode, the need
1164 for this locale support is expected to die away.
1165 .P
1166 External tables are built by calling the \fBpcre_maketables()\fP function,
1167 which has no arguments, in the relevant locale. The result can then be passed
1168 to \fBpcre_compile()\fP as often as necessary. For example, to build and use
1169 tables that are appropriate for the French locale (where accented characters
1170 with values greater than 128 are treated as letters), the following code could
1171 be used:
1172 .sp
1173 setlocale(LC_CTYPE, "fr_FR");
1174 tables = pcre_maketables();
1175 re = pcre_compile(..., tables);
1176 .sp
1177 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1178 are using Windows, the name for the French locale is "french".
1179 .P
1180 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1181 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1182 that the memory containing the tables remains available for as long as it is
1183 needed.
1184 .P
1185 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1186 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1187 and also by \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP. Thus, for any single
1188 pattern, compilation, studying and matching all happen in the same locale, but
1189 different patterns can be processed in different locales.
1190 .P
1191 It is possible to pass a table pointer or NULL (indicating the use of the
1192 internal tables) to \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP (see the
1193 discussion below in the section on matching a pattern). This facility is
1194 provided for use with pre-compiled patterns that have been saved and reloaded.
1195 Character tables are not saved with patterns, so if a non-standard table was
1196 used at compile time, it must be provided again when the reloaded pattern is
1197 matched. Attempting to use this facility to match a pattern in a different
1198 locale from the one in which it was compiled is likely to lead to anomalous
1199 (usually incorrect) results.
1200 .
1201 .
1202 .\" HTML <a name="infoaboutpattern"></a>
1203 .SH "INFORMATION ABOUT A PATTERN"
1204 .rs
1205 .sp
1206 .nf
1207 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1208 .B " int \fIwhat\fP, void *\fIwhere\fP);"
1209 .fi
1210 .PP
1211 The \fBpcre_fullinfo()\fP function returns information about a compiled
1212 pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1213 library at version 8.30, after more than 10 years of obsolescence.
1214 .P
1215 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1216 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1217 the pattern was not studied. The third argument specifies which piece of
1218 information is required, and the fourth argument is a pointer to a variable
1219 to receive the data. The yield of the function is zero for success, or one of
1220 the following negative numbers:
1221 .sp
1222 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1223 the argument \fIwhere\fP was NULL
1224 PCRE_ERROR_BADMAGIC the "magic number" was not found
1225 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1226 endianness
1227 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1228 PCRE_ERROR_UNSET the requested field is not set
1229 .sp
1230 The "magic number" is placed at the start of each compiled pattern as an simple
1231 check against passing an arbitrary memory pointer. The endianness error can
1232 occur if a compiled pattern is saved and reloaded on a different host. Here is
1233 a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1234 pattern:
1235 .sp
1236 int rc;
1237 size_t length;
1238 rc = pcre_fullinfo(
1239 re, /* result of pcre_compile() */
1240 sd, /* result of pcre_study(), or NULL */
1241 PCRE_INFO_SIZE, /* what is required */
1242 &length); /* where to put the data */
1243 .sp
1244 The possible values for the third argument are defined in \fBpcre.h\fP, and are
1245 as follows:
1246 .sp
1247 PCRE_INFO_BACKREFMAX
1248 .sp
1249 Return the number of the highest back reference in the pattern. The fourth
1250 argument should point to an \fBint\fP variable. Zero is returned if there are
1251 no back references.
1252 .sp
1253 PCRE_INFO_CAPTURECOUNT
1254 .sp
1255 Return the number of capturing subpatterns in the pattern. The fourth argument
1256 should point to an \fBint\fP variable.
1257 .sp
1258 PCRE_INFO_DEFAULT_TABLES
1259 .sp
1260 Return a pointer to the internal default character tables within PCRE. The
1261 fourth argument should point to an \fBunsigned char *\fP variable. This
1262 information call is provided for internal use by the \fBpcre_study()\fP
1263 function. External callers can cause PCRE to use its internal tables by passing
1264 a NULL table pointer.
1265 .sp
1266 PCRE_INFO_FIRSTBYTE (deprecated)
1267 .sp
1268 Return information about the first data unit of any matched string, for a
1269 non-anchored pattern. The name of this option refers to the 8-bit library,
1270 where data units are bytes. The fourth argument should point to an \fBint\fP
1271 variable. Negative values are used for special cases. However, this means that
1272 when the 32-bit library is in non-UTF-32 mode, the full 32-bit range of
1273 characters cannot be returned. For this reason, this value is deprecated; use
1274 PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER instead.
1275 .P
1276 If there is a fixed first value, for example, the letter "c" from a pattern
1277 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1278 value is always less than 256. In the 16-bit library the value can be up to
1279 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1280 .P
1281 If there is no fixed first value, and if either
1282 .sp
1283 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1284 starts with "^", or
1285 .sp
1286 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1287 (if it were set, the pattern would be anchored),
1288 .sp
1289 -1 is returned, indicating that the pattern matches only at the start of a
1290 subject string or after any newline within the string. Otherwise -2 is
1291 returned. For anchored patterns, -2 is returned.
1292 .sp
1293 PCRE_INFO_FIRSTCHARACTER
1294 .sp
1295 Return the value of the first data unit (non-UTF character) of any matched
1296 string in the situation where PCRE_INFO_FIRSTCHARACTERFLAGS returns 1;
1297 otherwise return 0. The fourth argument should point to an \fBuint_t\fP
1298 variable.
1299 .P
1300 In the 8-bit library, the value is always less than 256. In the 16-bit library
1301 the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1302 can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1303 .sp
1304 PCRE_INFO_FIRSTCHARACTERFLAGS
1305 .sp
1306 Return information about the first data unit of any matched string, for a
1307 non-anchored pattern. The fourth argument should point to an \fBint\fP
1308 variable.
1309 .P
1310 If there is a fixed first value, for example, the letter "c" from a pattern
1311 such as (cat|cow|coyote), 1 is returned, and the character value can be
1312 retrieved using PCRE_INFO_FIRSTCHARACTER. If there is no fixed first value, and
1313 if either
1314 .sp
1315 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1316 starts with "^", or
1317 .sp
1318 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1319 (if it were set, the pattern would be anchored),
1320 .sp
1321 2 is returned, indicating that the pattern matches only at the start of a
1322 subject string or after any newline within the string. Otherwise 0 is
1323 returned. For anchored patterns, 0 is returned.
1324 .sp
1325 PCRE_INFO_FIRSTTABLE
1326 .sp
1327 If the pattern was studied, and this resulted in the construction of a 256-bit
1328 table indicating a fixed set of values for the first data unit in any matching
1329 string, a pointer to the table is returned. Otherwise NULL is returned. The
1330 fourth argument should point to an \fBunsigned char *\fP variable.
1331 .sp
1332 PCRE_INFO_HASCRORLF
1333 .sp
1334 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1335 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1336 explicit match is either a literal CR or LF character, or \er or \en.
1337 .sp
1338 PCRE_INFO_JCHANGED
1339 .sp
1340 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1341 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1342 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1343 .sp
1344 PCRE_INFO_JIT
1345 .sp
1346 Return 1 if the pattern was studied with one of the JIT options, and
1347 just-in-time compiling was successful. The fourth argument should point to an
1348 \fBint\fP variable. A return value of 0 means that JIT support is not available
1349 in this version of PCRE, or that the pattern was not studied with a JIT option,
1350 or that the JIT compiler could not handle this particular pattern. See the
1351 .\" HREF
1352 \fBpcrejit\fP
1353 .\"
1354 documentation for details of what can and cannot be handled.
1355 .sp
1356 PCRE_INFO_JITSIZE
1357 .sp
1358 If the pattern was successfully studied with a JIT option, return the size of
1359 the JIT compiled code, otherwise return zero. The fourth argument should point
1360 to a \fBsize_t\fP variable.
1361 .sp
1362 PCRE_INFO_LASTLITERAL
1363 .sp
1364 Return the value of the rightmost literal data unit that must exist in any
1365 matched string, other than at its start, if such a value has been recorded. The
1366 fourth argument should point to an \fBint\fP variable. If there is no such
1367 value, -1 is returned. For anchored patterns, a last literal value is recorded
1368 only if it follows something of variable length. For example, for the pattern
1369 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1370 is -1.
1371 .P
1372 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1373 to return the full 32-bit range of characters, this value is deprecated;
1374 instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1375 be used.
1376 .sp
1377 PCRE_INFO_MATCH_EMPTY
1378 .sp
1379 Return 1 if the pattern can match an empty string, otherwise 0. The fourth
1380 argument should point to an \fBint\fP variable.
1381 .sp
1382 PCRE_INFO_MATCHLIMIT
1383 .sp
1384 If the pattern set a match limit by including an item of the form
1385 (*LIMIT_MATCH=nnnn) at the start, the value is returned. The fourth argument
1386 should point to an unsigned 32-bit integer. If no such value has been set, the
1387 call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1388 .sp
1389 PCRE_INFO_MAXLOOKBEHIND
1390 .sp
1391 Return the number of characters (NB not data units) in the longest lookbehind
1392 assertion in the pattern. This information is useful when doing multi-segment
1393 matching using the partial matching facilities. Note that the simple assertions
1394 \eb and \eB require a one-character lookbehind. \eA also registers a
1395 one-character lookbehind, though it does not actually inspect the previous
1396 character. This is to ensure that at least one character from the old segment
1397 is retained when a new segment is processed. Otherwise, if there are no
1398 lookbehinds in the pattern, \eA might match incorrectly at the start of a new
1399 segment.
1400 .sp
1401 PCRE_INFO_MINLENGTH
1402 .sp
1403 If the pattern was studied and a minimum length for matching subject strings
1404 was computed, its value is returned. Otherwise the returned value is -1. The
1405 value is a number of characters, which in UTF mode may be different from the
1406 number of data units. The fourth argument should point to an \fBint\fP
1407 variable. A non-negative value is a lower bound to the length of any matching
1408 string. There may not be any strings of that length that do actually match, but
1409 every string that does match is at least that long.
1410 .sp
1411 PCRE_INFO_NAMECOUNT
1412 PCRE_INFO_NAMEENTRYSIZE
1413 PCRE_INFO_NAMETABLE
1414 .sp
1415 PCRE supports the use of named as well as numbered capturing parentheses. The
1416 names are just an additional way of identifying the parentheses, which still
1417 acquire numbers. Several convenience functions such as
1418 \fBpcre_get_named_substring()\fP are provided for extracting captured
1419 substrings by name. It is also possible to extract the data directly, by first
1420 converting the name to a number in order to access the correct pointers in the
1421 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1422 you need to use the name-to-number map, which is described by these three
1423 values.
1424 .P
1425 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1426 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1427 entry; both of these return an \fBint\fP value. The entry size depends on the
1428 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1429 entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1430 the first two bytes of each entry are the number of the capturing parenthesis,
1431 most significant byte first. In the 16-bit library, the pointer points to
1432 16-bit data units, the first of which contains the parenthesis number. In the
1433 32-bit library, the pointer points to 32-bit data units, the first of which
1434 contains the parenthesis number. The rest of the entry is the corresponding
1435 name, zero terminated.
1436 .P
1437 The names are in alphabetical order. If (?| is used to create multiple groups
1438 with the same number, as described in the
1439 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1440 .\" </a>
1441 section on duplicate subpattern numbers
1442 .\"
1443 in the
1444 .\" HREF
1445 \fBpcrepattern\fP
1446 .\"
1447 page, the groups may be given the same name, but there is only one entry in the
1448 table. Different names for groups of the same number are not permitted.
1449 Duplicate names for subpatterns with different numbers are permitted,
1450 but only if PCRE_DUPNAMES is set. They appear in the table in the order in
1451 which they were found in the pattern. In the absence of (?| this is the order
1452 of increasing number; when (?| is used this is not necessarily the case because
1453 later subpatterns may have lower numbers.
1454 .P
1455 As a simple example of the name/number table, consider the following pattern
1456 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1457 space - including newlines - is ignored):
1458 .sp
1459 .\" JOIN
1460 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1461 (?<month>\ed\ed) - (?<day>\ed\ed) )
1462 .sp
1463 There are four named subpatterns, so the table has four entries, and each entry
1464 in the table is eight bytes long. The table is as follows, with non-printing
1465 bytes shows in hexadecimal, and undefined bytes shown as ??:
1466 .sp
1467 00 01 d a t e 00 ??
1468 00 05 d a y 00 ?? ??
1469 00 04 m o n t h 00
1470 00 02 y e a r 00 ??
1471 .sp
1472 When writing code to extract data from named subpatterns using the
1473 name-to-number map, remember that the length of the entries is likely to be
1474 different for each compiled pattern.
1475 .sp
1476 PCRE_INFO_OKPARTIAL
1477 .sp
1478 Return 1 if the pattern can be used for partial matching with
1479 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1480 \fBint\fP variable. From release 8.00, this always returns 1, because the
1481 restrictions that previously applied to partial matching have been lifted. The
1482 .\" HREF
1483 \fBpcrepartial\fP
1484 .\"
1485 documentation gives details of partial matching.
1486 .sp
1487 PCRE_INFO_OPTIONS
1488 .sp
1489 Return a copy of the options with which the pattern was compiled. The fourth
1490 argument should point to an \fBunsigned long int\fP variable. These option bits
1491 are those specified in the call to \fBpcre_compile()\fP, modified by any
1492 top-level option settings at the start of the pattern itself. In other words,
1493 they are the options that will be in force when matching starts. For example,
1494 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1495 result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1496 .P
1497 A pattern is automatically anchored by PCRE if all of its top-level
1498 alternatives begin with one of the following:
1499 .sp
1500 ^ unless PCRE_MULTILINE is set
1501 \eA always
1502 \eG always
1503 .\" JOIN
1504 .* if PCRE_DOTALL is set and there are no back
1505 references to the subpattern in which .* appears
1506 .sp
1507 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1508 \fBpcre_fullinfo()\fP.
1509 .sp
1510 PCRE_INFO_RECURSIONLIMIT
1511 .sp
1512 If the pattern set a recursion limit by including an item of the form
1513 (*LIMIT_RECURSION=nnnn) at the start, the value is returned. The fourth
1514 argument should point to an unsigned 32-bit integer. If no such value has been
1515 set, the call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1516 .sp
1517 PCRE_INFO_SIZE
1518 .sp
1519 Return the size of the compiled pattern in bytes (for all three libraries). The
1520 fourth argument should point to a \fBsize_t\fP variable. This value does not
1521 include the size of the \fBpcre\fP structure that is returned by
1522 \fBpcre_compile()\fP. The value that is passed as the argument to
1523 \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1524 place the compiled data is the value returned by this option plus the size of
1525 the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1526 does not alter the value returned by this option.
1527 .sp
1528 PCRE_INFO_STUDYSIZE
1529 .sp
1530 Return the size in bytes (for all three libraries) of the data block pointed to
1531 by the \fIstudy_data\fP field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP
1532 is NULL, or there is no study data, zero is returned. The fourth argument
1533 should point to a \fBsize_t\fP variable. The \fIstudy_data\fP field is set by
1534 \fBpcre_study()\fP to record information that will speed up matching (see the
1535 section entitled
1536 .\" HTML <a href="#studyingapattern">
1537 .\" </a>
1538 "Studying a pattern"
1539 .\"
1540 above). The format of the \fIstudy_data\fP block is private, but its length
1541 is made available via this option so that it can be saved and restored (see the
1542 .\" HREF
1543 \fBpcreprecompile\fP
1544 .\"
1545 documentation for details).
1546 .sp
1547 PCRE_INFO_REQUIREDCHARFLAGS
1548 .sp
1549 Returns 1 if there is a rightmost literal data unit that must exist in any
1550 matched string, other than at its start. The fourth argument should point to
1551 an \fBint\fP variable. If there is no such value, 0 is returned. If returning
1552 1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1553 .P
1554 For anchored patterns, a last literal value is recorded only if it follows
1555 something of variable length. For example, for the pattern /^a\ed+z\ed+/ the
1556 returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1557 /^a\edz\ed/ the returned value is 0.
1558 .sp
1559 PCRE_INFO_REQUIREDCHAR
1560 .sp
1561 Return the value of the rightmost literal data unit that must exist in any
1562 matched string, other than at its start, if such a value has been recorded. The
1563 fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1564 value, 0 is returned.
1565 .
1566 .
1567 .SH "REFERENCE COUNTS"
1568 .rs
1569 .sp
1570 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1571 .PP
1572 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1573 data block that contains a compiled pattern. It is provided for the benefit of
1574 applications that operate in an object-oriented manner, where different parts
1575 of the application may be using the same compiled pattern, but you want to free
1576 the block when they are all done.
1577 .P
1578 When a pattern is compiled, the reference count field is initialized to zero.
1579 It is changed only by calling this function, whose action is to add the
1580 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1581 function is the new value. However, the value of the count is constrained to
1582 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1583 it is forced to the appropriate limit value.
1584 .P
1585 Except when it is zero, the reference count is not correctly preserved if a
1586 pattern is compiled on one host and then transferred to a host whose byte-order
1587 is different. (This seems a highly unlikely scenario.)
1588 .
1589 .
1590 .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1591 .rs
1592 .sp
1593 .nf
1594 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1595 .B " const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1596 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);"
1597 .fi
1598 .P
1599 The function \fBpcre_exec()\fP is called to match a subject string against a
1600 compiled pattern, which is passed in the \fIcode\fP argument. If the
1601 pattern was studied, the result of the study should be passed in the
1602 \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1603 and \fIextra\fP arguments as many times as you like, in order to match
1604 different subject strings with the same pattern.
1605 .P
1606 This function is the main matching facility of the library, and it operates in
1607 a Perl-like manner. For specialist use there is also an alternative matching
1608 function, which is described
1609 .\" HTML <a href="#dfamatch">
1610 .\" </a>
1611 below
1612 .\"
1613 in the section about the \fBpcre_dfa_exec()\fP function.
1614 .P
1615 In most applications, the pattern will have been compiled (and optionally
1616 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1617 possible to save compiled patterns and study data, and then use them later
1618 in different processes, possibly even on different hosts. For a discussion
1619 about this, see the
1620 .\" HREF
1621 \fBpcreprecompile\fP
1622 .\"
1623 documentation.
1624 .P
1625 Here is an example of a simple call to \fBpcre_exec()\fP:
1626 .sp
1627 int rc;
1628 int ovector[30];
1629 rc = pcre_exec(
1630 re, /* result of pcre_compile() */
1631 NULL, /* we didn't study the pattern */
1632 "some string", /* the subject string */
1633 11, /* the length of the subject string */
1634 0, /* start at offset 0 in the subject */
1635 0, /* default options */
1636 ovector, /* vector of integers for substring information */
1637 30); /* number of elements (NOT size in bytes) */
1638 .
1639 .
1640 .\" HTML <a name="extradata"></a>
1641 .SS "Extra data for \fBpcre_exec()\fR"
1642 .rs
1643 .sp
1644 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1645 data block. The \fBpcre_study()\fP function returns such a block (when it
1646 doesn't return NULL), but you can also create one for yourself, and pass
1647 additional information in it. The \fBpcre_extra\fP block contains the following
1648 fields (not necessarily in this order):
1649 .sp
1650 unsigned long int \fIflags\fP;
1651 void *\fIstudy_data\fP;
1652 void *\fIexecutable_jit\fP;
1653 unsigned long int \fImatch_limit\fP;
1654 unsigned long int \fImatch_limit_recursion\fP;
1655 void *\fIcallout_data\fP;
1656 const unsigned char *\fItables\fP;
1657 unsigned char **\fImark\fP;
1658 .sp
1659 In the 16-bit version of this structure, the \fImark\fP field has type
1660 "PCRE_UCHAR16 **".
1661 .sp
1662 In the 32-bit version of this structure, the \fImark\fP field has type
1663 "PCRE_UCHAR32 **".
1664 .P
1665 The \fIflags\fP field is used to specify which of the other fields are set. The
1666 flag bits are:
1667 .sp
1668 PCRE_EXTRA_CALLOUT_DATA
1669 PCRE_EXTRA_EXECUTABLE_JIT
1670 PCRE_EXTRA_MARK
1671 PCRE_EXTRA_MATCH_LIMIT
1672 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1673 PCRE_EXTRA_STUDY_DATA
1674 PCRE_EXTRA_TABLES
1675 .sp
1676 Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1677 the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1678 returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1679 should not set these yourself, but you may add to the block by setting other
1680 fields and their corresponding flag bits.
1681 .P
1682 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1683 vast amount of resources when running patterns that are not going to match,
1684 but which have a very large number of possibilities in their search trees. The
1685 classic example is a pattern that uses nested unlimited repeats.
1686 .P
1687 Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1688 calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1689 imposed on the number of times this function is called during a match, which
1690 has the effect of limiting the amount of backtracking that can take place. For
1691 patterns that are not anchored, the count restarts from zero for each position
1692 in the subject string.
1693 .P
1694 When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1695 with a JIT option, the way that the matching is executed is entirely different.
1696 However, there is still the possibility of runaway matching that goes on for a
1697 very long time, and so the \fImatch_limit\fP value is also used in this case
1698 (but in a different way) to limit how long the matching can continue.
1699 .P
1700 The default value for the limit can be set when PCRE is built; the default
1701 default is 10 million, which handles all but the most extreme cases. You can
1702 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1703 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1704 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1705 PCRE_ERROR_MATCHLIMIT.
1706 .P
1707 A value for the match limit may also be supplied by an item at the start of a
1708 pattern of the form
1709 .sp
1710 (*LIMIT_MATCH=d)
1711 .sp
1712 where d is a decimal number. However, such a setting is ignored unless d is
1713 less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1714 is set, less than the default.
1715 .P
1716 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1717 instead of limiting the total number of times that \fBmatch()\fP is called, it
1718 limits the depth of recursion. The recursion depth is a smaller number than the
1719 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1720 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1721 .P
1722 Limiting the recursion depth limits the amount of machine stack that can be
1723 used, or, when PCRE has been compiled to use memory on the heap instead of the
1724 stack, the amount of heap memory that can be used. This limit is not relevant,
1725 and is ignored, when matching is done using JIT compiled code.
1726 .P
1727 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1728 built; the default default is the same value as the default for
1729 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1730 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1731 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1732 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1733 .P
1734 A value for the recursion limit may also be supplied by an item at the start of
1735 a pattern of the form
1736 .sp
1737 (*LIMIT_RECURSION=d)
1738 .sp
1739 where d is a decimal number. However, such a setting is ignored unless d is
1740 less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1741 is set, less than the default.
1742 .P
1743 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1744 and is described in the
1745 .\" HREF
1746 \fBpcrecallout\fP
1747 .\"
1748 documentation.
1749 .P
1750 The \fItables\fP field is provided for use with patterns that have been
1751 pre-compiled using custom character tables, saved to disc or elsewhere, and
1752 then reloaded, because the tables that were used to compile a pattern are not
1753 saved with it. See the
1754 .\" HREF
1755 \fBpcreprecompile\fP
1756 .\"
1757 documentation for a discussion of saving compiled patterns for later use. If
1758 NULL is passed using this mechanism, it forces PCRE's internal tables to be
1759 used.
1760 .P
1761 \fBWarning:\fP The tables that \fBpcre_exec()\fP uses must be the same as those
1762 that were used when the pattern was compiled. If this is not the case, the
1763 behaviour of \fBpcre_exec()\fP is undefined. Therefore, when a pattern is
1764 compiled and matched in the same process, this field should never be set. In
1765 this (the most common) case, the correct table pointer is automatically passed
1766 with the compiled pattern from \fBpcre_compile()\fP to \fBpcre_exec()\fP.
1767 .P
1768 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1769 be set to point to a suitable variable. If the pattern contains any
1770 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1771 a name to pass back, a pointer to the name string (zero terminated) is placed
1772 in the variable pointed to by the \fImark\fP field. The names are within the
1773 compiled pattern; if you wish to retain such a name you must copy it before
1774 freeing the memory of a compiled pattern. If there is no name to pass back, the
1775 variable pointed to by the \fImark\fP field is set to NULL. For details of the
1776 backtracking control verbs, see the section entitled
1777 .\" HTML <a href="pcrepattern#backtrackcontrol">
1778 .\" </a>
1779 "Backtracking control"
1780 .\"
1781 in the
1782 .\" HREF
1783 \fBpcrepattern\fP
1784 .\"
1785 documentation.
1786 .
1787 .
1788 .\" HTML <a name="execoptions"></a>
1789 .SS "Option bits for \fBpcre_exec()\fP"
1790 .rs
1791 .sp
1792 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1793 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1794 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1795 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1796 PCRE_PARTIAL_SOFT.
1797 .P
1798 If the pattern was successfully studied with one of the just-in-time (JIT)
1799 compile options, the only supported options for JIT execution are
1800 PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1801 PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1802 unsupported option is used, JIT execution is disabled and the normal
1803 interpretive code in \fBpcre_exec()\fP is run.
1804 .sp
1805 PCRE_ANCHORED
1806 .sp
1807 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1808 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1809 to be anchored by virtue of its contents, it cannot be made unachored at
1810 matching time.
1811 .sp
1812 PCRE_BSR_ANYCRLF
1813 PCRE_BSR_UNICODE
1814 .sp
1815 These options (which are mutually exclusive) control what the \eR escape
1816 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1817 match any Unicode newline sequence. These options override the choice that was
1818 made or defaulted when the pattern was compiled.
1819 .sp
1820 PCRE_NEWLINE_CR
1821 PCRE_NEWLINE_LF
1822 PCRE_NEWLINE_CRLF
1823 PCRE_NEWLINE_ANYCRLF
1824 PCRE_NEWLINE_ANY
1825 .sp
1826 These options override the newline definition that was chosen or defaulted when
1827 the pattern was compiled. For details, see the description of
1828 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1829 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1830 the way the match position is advanced after a match failure for an unanchored
1831 pattern.
1832 .P
1833 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1834 match attempt for an unanchored pattern fails when the current position is at a
1835 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1836 characters, the match position is advanced by two characters instead of one, in
1837 other words, to after the CRLF.
1838 .P
1839 The above rule is a compromise that makes the most common cases work as
1840 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1841 set), it does not match the string "\er\enA" because, after failing at the
1842 start, it skips both the CR and the LF before retrying. However, the pattern
1843 [\er\en]A does match that string, because it contains an explicit CR or LF
1844 reference, and so advances only by one character after the first failure.
1845 .P
1846 An explicit match for CR of LF is either a literal appearance of one of those
1847 characters, or one of the \er or \en escape sequences. Implicit matches such as
1848 [^X] do not count, nor does \es (which includes CR and LF in the characters
1849 that it matches).
1850 .P
1851 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1852 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1853 .sp
1854 PCRE_NOTBOL
1855 .sp
1856 This option specifies that first character of the subject string is not the
1857 beginning of a line, so the circumflex metacharacter should not match before
1858 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1859 never to match. This option affects only the behaviour of the circumflex
1860 metacharacter. It does not affect \eA.
1861 .sp
1862 PCRE_NOTEOL
1863 .sp
1864 This option specifies that the end of the subject string is not the end of a
1865 line, so the dollar metacharacter should not match it nor (except in multiline
1866 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1867 compile time) causes dollar never to match. This option affects only the
1868 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1869 .sp
1870 PCRE_NOTEMPTY
1871 .sp
1872 An empty string is not considered to be a valid match if this option is set. If
1873 there are alternatives in the pattern, they are tried. If all the alternatives
1874 match the empty string, the entire match fails. For example, if the pattern
1875 .sp
1876 a?b?
1877 .sp
1878 is applied to a string not beginning with "a" or "b", it matches an empty
1879 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1880 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1881 .sp
1882 PCRE_NOTEMPTY_ATSTART
1883 .sp
1884 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1885 the start of the subject is permitted. If the pattern is anchored, such a match
1886 can occur only if the pattern contains \eK.
1887 .P
1888 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1889 does make a special case of a pattern match of the empty string within its
1890 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1891 emulate Perl's behaviour after matching a null string by first trying the match
1892 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1893 if that fails, by advancing the starting offset (see below) and trying an
1894 ordinary match again. There is some code that demonstrates how to do this in
1895 the
1896 .\" HREF
1897 \fBpcredemo\fP
1898 .\"
1899 sample program. In the most general case, you have to check to see if the
1900 newline convention recognizes CRLF as a newline, and if so, and the current
1901 character is CR followed by LF, advance the starting offset by two characters
1902 instead of one.
1903 .sp
1904 PCRE_NO_START_OPTIMIZE
1905 .sp
1906 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1907 a match, in order to speed up the process. For example, if it is known that an
1908 unanchored match must start with a specific character, it searches the subject
1909 for that character, and fails immediately if it cannot find it, without
1910 actually running the main matching function. This means that a special item
1911 such as (*COMMIT) at the start of a pattern is not considered until after a
1912 suitable starting point for the match has been found. Also, when callouts or
1913 (*MARK) items are in use, these "start-up" optimizations can cause them to be
1914 skipped if the pattern is never actually used. The start-up optimizations are
1915 in effect a pre-scan of the subject that takes place before the pattern is run.
1916 .P
1917 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1918 causing performance to suffer, but ensuring that in cases where the result is
1919 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1920 are considered at every possible starting position in the subject string. If
1921 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1922 time. The use of PCRE_NO_START_OPTIMIZE at matching time (that is, passing it
1923 to \fBpcre_exec()\fP) disables JIT execution; in this situation, matching is
1924 always done using interpretively.
1925 .P
1926 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1927 Consider the pattern
1928 .sp
1929 (*COMMIT)ABC
1930 .sp
1931 When this is compiled, PCRE records the fact that a match must start with the
1932 character "A". Suppose the subject string is "DEFABC". The start-up
1933 optimization scans along the subject, finds "A" and runs the first match
1934 attempt from there. The (*COMMIT) item means that the pattern must match the
1935 current starting position, which in this case, it does. However, if the same
1936 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1937 subject string does not happen. The first match attempt is run starting from
1938 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1939 the overall result is "no match". If the pattern is studied, more start-up
1940 optimizations may be used. For example, a minimum length for the subject may be
1941 recorded. Consider the pattern
1942 .sp
1943 (*MARK:A)(X|Y)
1944 .sp
1945 The minimum length for a match is one character. If the subject is "ABC", there
1946 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1947 If the pattern is studied, the final attempt does not take place, because PCRE
1948 knows that the subject is too short, and so the (*MARK) is never encountered.
1949 In this case, studying the pattern does not affect the overall match result,
1950 which is still "no match", but it does affect the auxiliary information that is
1951 returned.
1952 .sp
1953 PCRE_NO_UTF8_CHECK
1954 .sp
1955 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1956 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1957 The entire string is checked before any other processing takes place. The value
1958 of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1959 UTF-8 character. There is a discussion about the
1960 .\" HTML <a href="pcreunicode.html#utf8strings">
1961 .\" </a>
1962 validity of UTF-8 strings
1963 .\"
1964 in the
1965 .\" HREF
1966 \fBpcreunicode\fP
1967 .\"
1968 page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1969 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1970 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1971 cases, information about the precise nature of the error may also be returned
1972 (see the descriptions of these errors in the section entitled \fIError return
1973 values from\fP \fBpcre_exec()\fP
1974 .\" HTML <a href="#errorlist">
1975 .\" </a>
1976 below).
1977 .\"
1978 If \fIstartoffset\fP contains a value that does not point to the start of a
1979 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1980 returned.
1981 .P
1982 If you already know that your subject is valid, and you want to skip these
1983 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1984 calling \fBpcre_exec()\fP. You might want to do this for the second and
1985 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1986 all the matches in a single subject string. However, you should be sure that
1987 the value of \fIstartoffset\fP points to the start of a character (or the end
1988 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1989 invalid string as a subject or an invalid value of \fIstartoffset\fP is
1990 undefined. Your program may crash or loop.
1991 .sp
1992 PCRE_PARTIAL_HARD
1993 PCRE_PARTIAL_SOFT
1994 .sp
1995 These options turn on the partial matching feature. For backwards
1996 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1997 occurs if the end of the subject string is reached successfully, but there are
1998 not enough subject characters to complete the match. If this happens when
1999 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
2000 testing any remaining alternatives. Only if no complete match can be found is
2001 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
2002 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
2003 but only if no complete match can be found.
2004 .P
2005 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
2006 partial match is found, \fBpcre_exec()\fP immediately returns
2007 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
2008 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
2009 important that an alternative complete match.
2010 .P
2011 In both cases, the portion of the string that was inspected when the partial
2012 match was found is set as the first matching string. There is a more detailed
2013 discussion of partial and multi-segment matching, with examples, in the
2014 .\" HREF
2015 \fBpcrepartial\fP
2016 .\"
2017 documentation.
2018 .
2019 .
2020 .SS "The string to be matched by \fBpcre_exec()\fP"
2021 .rs
2022 .sp
2023 The subject string is passed to \fBpcre_exec()\fP as a pointer in
2024 \fIsubject\fP, a length in \fIlength\fP, and a starting offset in
2025 \fIstartoffset\fP. The units for \fIlength\fP and \fIstartoffset\fP are bytes
2026 for the 8-bit library, 16-bit data items for the 16-bit library, and 32-bit
2027 data items for the 32-bit library.
2028 .P
2029 If \fIstartoffset\fP is negative or greater than the length of the subject,
2030 \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting offset is
2031 zero, the search for a match starts at the beginning of the subject, and this
2032 is by far the most common case. In UTF-8 or UTF-16 mode, the offset must point
2033 to the start of a character, or the end of the subject (in UTF-32 mode, one
2034 data unit equals one character, so all offsets are valid). Unlike the pattern
2035 string, the subject may contain binary zeroes.
2036 .P
2037 A non-zero starting offset is useful when searching for another match in the
2038 same subject by calling \fBpcre_exec()\fP again after a previous success.
2039 Setting \fIstartoffset\fP differs from just passing over a shortened string and
2040 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
2041 lookbehind. For example, consider the pattern
2042 .sp
2043 \eBiss\eB
2044 .sp
2045 which finds occurrences of "iss" in the middle of words. (\eB matches only if
2046 the current position in the subject is not a word boundary.) When applied to
2047 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
2048 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
2049 subject, namely "issipi", it does not match, because \eB is always false at the
2050 start of the subject, which is deemed to be a word boundary. However, if
2051 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
2052 set to 4, it finds the second occurrence of "iss" because it is able to look
2053 behind the starting point to discover that it is preceded by a letter.
2054 .P
2055 Finding all the matches in a subject is tricky when the pattern can match an
2056 empty string. It is possible to emulate Perl's /g behaviour by first trying the
2057 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
2058 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
2059 and trying an ordinary match again. There is some code that demonstrates how to
2060 do this in the
2061 .\" HREF
2062 \fBpcredemo\fP
2063 .\"
2064 sample program. In the most general case, you have to check to see if the
2065 newline convention recognizes CRLF as a newline, and if so, and the current
2066 character is CR followed by LF, advance the starting offset by two characters
2067 instead of one.
2068 .P
2069 If a non-zero starting offset is passed when the pattern is anchored, one
2070 attempt to match at the given offset is made. This can only succeed if the
2071 pattern does not require the match to be at the start of the subject.
2072 .
2073 .
2074 .SS "How \fBpcre_exec()\fP returns captured substrings"
2075 .rs
2076 .sp
2077 In general, a pattern matches a certain portion of the subject, and in
2078 addition, further substrings from the subject may be picked out by parts of the
2079 pattern. Following the usage in Jeffrey Friedl's book, this is called
2080 "capturing" in what follows, and the phrase "capturing subpattern" is used for
2081 a fragment of a pattern that picks out a substring. PCRE supports several other
2082 kinds of parenthesized subpattern that do not cause substrings to be captured.
2083 .P
2084 Captured substrings are returned to the caller via a vector of integers whose
2085 address is passed in \fIovector\fP. The number of elements in the vector is
2086 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
2087 argument is NOT the size of \fIovector\fP in bytes.
2088 .P
2089 The first two-thirds of the vector is used to pass back captured substrings,
2090 each substring using a pair of integers. The remaining third of the vector is
2091 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
2092 and is not available for passing back information. The number passed in
2093 \fIovecsize\fP should always be a multiple of three. If it is not, it is
2094 rounded down.
2095 .P
2096 When a match is successful, information about captured substrings is returned
2097 in pairs of integers, starting at the beginning of \fIovector\fP, and
2098 continuing up to two-thirds of its length at the most. The first element of
2099 each pair is set to the offset of the first character in a substring, and the
2100 second is set to the offset of the first character after the end of a
2101 substring. These values are always data unit offsets, even in UTF mode. They
2102 are byte offsets in the 8-bit library, 16-bit data item offsets in the 16-bit
2103 library, and 32-bit data item offsets in the 32-bit library. \fBNote\fP: they
2104 are not character counts.
2105 .P
2106 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
2107 portion of the subject string matched by the entire pattern. The next pair is
2108 used for the first capturing subpattern, and so on. The value returned by
2109 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
2110 For example, if two substrings have been captured, the returned value is 3. If
2111 there are no capturing subpatterns, the return value from a successful match is
2112 1, indicating that just the first pair of offsets has been set.
2113 .P
2114 If a capturing subpattern is matched repeatedly, it is the last portion of the
2115 string that it matched that is returned.
2116 .P
2117 If the vector is too small to hold all the captured substring offsets, it is
2118 used as far as possible (up to two-thirds of its length), and the function
2119 returns a value of zero. If neither the actual string matched nor any captured
2120 substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2121 passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2122 back references and the \fIovector\fP is not big enough to remember the related
2123 substrings, PCRE has to get additional memory for use during matching. Thus it
2124 is usually advisable to supply an \fIovector\fP of reasonable size.
2125 .P
2126 There are some cases where zero is returned (indicating vector overflow) when
2127 in fact the vector is exactly the right size for the final match. For example,
2128 consider the pattern
2129 .sp
2130 (a)(?:(b)c|bd)
2131 .sp
2132 If a vector of 6 elements (allowing for only 1 captured substring) is given
2133 with subject string "abd", \fBpcre_exec()\fP will try to set the second
2134 captured string, thereby recording a vector overflow, before failing to match
2135 "c" and backing up to try the second alternative. The zero return, however,
2136 does correctly indicate that the maximum number of slots (namely 2) have been
2137 filled. In similar cases where there is temporary overflow, but the final
2138 number of used slots is actually less than the maximum, a non-zero value is
2139 returned.
2140 .P
2141 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
2142 subpatterns there are in a compiled pattern. The smallest size for
2143 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
2144 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
2145 .P
2146 It is possible for capturing subpattern number \fIn+1\fP to match some part of
2147 the subject when subpattern \fIn\fP has not been used at all. For example, if
2148 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2149 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2150 happens, both values in the offset pairs corresponding to unused subpatterns
2151 are set to -1.
2152 .P
2153 Offset values that correspond to unused subpatterns at the end of the
2154 expression are also set to -1. For example, if the string "abc" is matched
2155 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2156 return from the function is 2, because the highest used capturing subpattern
2157 number is 1, and the offsets for for the second and third capturing subpatterns
2158 (assuming the vector is large enough, of course) are set to -1.
2159 .P
2160 \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
2161 correspond to capturing parentheses in the pattern are never changed. That is,
2162 if a pattern contains \fIn\fP capturing parentheses, no more than
2163 \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
2164 elements (in the first two-thirds) retain whatever values they previously had.
2165 .P
2166 Some convenience functions are provided for extracting the captured substrings
2167 as separate strings. These are described below.
2168 .
2169 .
2170 .\" HTML <a name="errorlist"></a>
2171 .SS "Error return values from \fBpcre_exec()\fP"
2172 .rs
2173 .sp
2174 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
2175 defined in the header file:
2176 .sp
2177 PCRE_ERROR_NOMATCH (-1)
2178 .sp
2179 The subject string did not match the pattern.
2180 .sp
2181 PCRE_ERROR_NULL (-2)
2182 .sp
2183 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
2184 NULL and \fIovecsize\fP was not zero.
2185 .sp
2186 PCRE_ERROR_BADOPTION (-3)
2187 .sp
2188 An unrecognized bit was set in the \fIoptions\fP argument.
2189 .sp
2190 PCRE_ERROR_BADMAGIC (-4)
2191 .sp
2192 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2193 the case when it is passed a junk pointer and to detect when a pattern that was
2194 compiled in an environment of one endianness is run in an environment with the
2195 other endianness. This is the error that PCRE gives when the magic number is
2196 not present.
2197 .sp
2198 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2199 .sp
2200 While running the pattern match, an unknown item was encountered in the
2201 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2202 of the compiled pattern.
2203 .sp
2204 PCRE_ERROR_NOMEMORY (-6)
2205 .sp
2206 If a pattern contains back references, but the \fIovector\fP that is passed to
2207 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2208 gets a block of memory at the start of matching to use for this purpose. If the
2209 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2210 automatically freed at the end of matching.
2211 .P
2212 This error is also given if \fBpcre_stack_malloc()\fP fails in
2213 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2214 \fB--disable-stack-for-recursion\fP.
2215 .sp
2216 PCRE_ERROR_NOSUBSTRING (-7)
2217 .sp
2218 This error is used by the \fBpcre_copy_substring()\fP,
2219 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2220 below). It is never returned by \fBpcre_exec()\fP.
2221 .sp
2222 PCRE_ERROR_MATCHLIMIT (-8)
2223 .sp
2224 The backtracking limit, as specified by the \fImatch_limit\fP field in a
2225 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2226 above.
2227 .sp
2228 PCRE_ERROR_CALLOUT (-9)
2229 .sp
2230 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2231 use by callout functions that want to yield a distinctive error code. See the
2232 .\" HREF
2233 \fBpcrecallout\fP
2234 .\"
2235 documentation for details.
2236 .sp
2237 PCRE_ERROR_BADUTF8 (-10)
2238 .sp
2239 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2240 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2241 (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2242 UTF-8 character is placed in the first element, and a reason code is placed in
2243 the second element. The reason codes are listed in the
2244 .\" HTML <a href="#badutf8reasons">
2245 .\" </a>
2246 following section.
2247 .\"
2248 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2249 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2250 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2251 .sp
2252 PCRE_ERROR_BADUTF8_OFFSET (-11)
2253 .sp
2254 The UTF-8 byte sequence that was passed as a subject was checked and found to
2255 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2256 \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2257 end of the subject.
2258 .sp
2259 PCRE_ERROR_PARTIAL (-12)
2260 .sp
2261 The subject string did not match, but it did match partially. See the
2262 .\" HREF
2263 \fBpcrepartial\fP
2264 .\"
2265 documentation for details of partial matching.
2266 .sp
2267 PCRE_ERROR_BADPARTIAL (-13)
2268 .sp
2269 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2270 option was used with a compiled pattern containing items that were not
2271 supported for partial matching. From release 8.00 onwards, there are no
2272 restrictions on partial matching.
2273 .sp
2274 PCRE_ERROR_INTERNAL (-14)
2275 .sp
2276 An unexpected internal error has occurred. This error could be caused by a bug
2277 in PCRE or by overwriting of the compiled pattern.
2278 .sp
2279 PCRE_ERROR_BADCOUNT (-15)
2280 .sp
2281 This error is given if the value of the \fIovecsize\fP argument is negative.
2282 .sp
2283 PCRE_ERROR_RECURSIONLIMIT (-21)
2284 .sp
2285 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2286 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2287 description above.
2288 .sp
2289 PCRE_ERROR_BADNEWLINE (-23)
2290 .sp
2291 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2292 .sp
2293 PCRE_ERROR_BADOFFSET (-24)
2294 .sp
2295 The value of \fIstartoffset\fP was negative or greater than the length of the
2296 subject, that is, the value in \fIlength\fP.
2297 .sp
2298 PCRE_ERROR_SHORTUTF8 (-25)
2299 .sp
2300 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2301 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2302 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2303 fact sufficient to detect this case, but this special error code for
2304 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2305 retained for backwards compatibility.
2306 .sp
2307 PCRE_ERROR_RECURSELOOP (-26)
2308 .sp
2309 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2310 the pattern. Specifically, it means that either the whole pattern or a
2311 subpattern has been called recursively for the second time at the same position
2312 in the subject string. Some simple patterns that might do this are detected and
2313 faulted at compile time, but more complicated cases, in particular mutual
2314 recursions between two different subpatterns, cannot be detected until run
2315 time.
2316 .sp
2317 PCRE_ERROR_JIT_STACKLIMIT (-27)
2318 .sp
2319 This error is returned when a pattern that was successfully studied using a
2320 JIT compile option is being matched, but the memory available for the
2321 just-in-time processing stack is not large enough. See the
2322 .\" HREF
2323 \fBpcrejit\fP
2324 .\"
2325 documentation for more details.
2326 .sp
2327 PCRE_ERROR_BADMODE (-28)
2328 .sp
2329 This error is given if a pattern that was compiled by the 8-bit library is
2330 passed to a 16-bit or 32-bit library function, or vice versa.
2331 .sp
2332 PCRE_ERROR_BADENDIANNESS (-29)
2333 .sp
2334 This error is given if a pattern that was compiled and saved is reloaded on a
2335 host with different endianness. The utility function
2336 \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2337 so that it runs on the new host.
2338 .sp
2339 PCRE_ERROR_JIT_BADOPTION
2340 .sp
2341 This error is returned when a pattern that was successfully studied using a JIT
2342 compile option is being matched, but the matching mode (partial or complete
2343 match) does not correspond to any JIT compilation mode. When the JIT fast path
2344 function is used, this error may be also given for invalid options. See the
2345 .\" HREF
2346 \fBpcrejit\fP
2347 .\"
2348 documentation for more details.
2349 .sp
2350 PCRE_ERROR_BADLENGTH (-32)
2351 .sp
2352 This error is given if \fBpcre_exec()\fP is called with a negative value for
2353 the \fIlength\fP argument.
2354 .P
2355 Error numbers -16 to -20, -22, and 30 are not used by \fBpcre_exec()\fP.
2356 .
2357 .
2358 .\" HTML <a name="badutf8reasons"></a>
2359 .SS "Reason codes for invalid UTF-8 strings"
2360 .rs
2361 .sp
2362 This section applies only to the 8-bit library. The corresponding information
2363 for the 16-bit and 32-bit libraries is given in the
2364 .\" HREF
2365 \fBpcre16\fP
2366 .\"
2367 and
2368 .\" HREF
2369 \fBpcre32\fP
2370 .\"
2371 pages.
2372 .P
2373 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2374 PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2375 least 2, the offset of the start of the invalid UTF-8 character is placed in
2376 the first output vector element (\fIovector[0]\fP) and a reason code is placed
2377 in the second element (\fIovector[1]\fP). The reason codes are given names in
2378 the \fBpcre.h\fP header file:
2379 .sp
2380 PCRE_UTF8_ERR1
2381 PCRE_UTF8_ERR2
2382 PCRE_UTF8_ERR3
2383 PCRE_UTF8_ERR4
2384 PCRE_UTF8_ERR5
2385 .sp
2386 The string ends with a truncated UTF-8 character; the code specifies how many
2387 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2388 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2389 allows for up to 6 bytes, and this is checked first; hence the possibility of
2390 4 or 5 missing bytes.
2391 .sp
2392 PCRE_UTF8_ERR6
2393 PCRE_UTF8_ERR7
2394 PCRE_UTF8_ERR8
2395 PCRE_UTF8_ERR9
2396 PCRE_UTF8_ERR10
2397 .sp
2398 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2399 character do not have the binary value 0b10 (that is, either the most
2400 significant bit is 0, or the next bit is 1).
2401 .sp
2402 PCRE_UTF8_ERR11
2403 PCRE_UTF8_ERR12
2404 .sp
2405 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2406 these code points are excluded by RFC 3629.
2407 .sp
2408 PCRE_UTF8_ERR13
2409 .sp
2410 A 4-byte character has a value greater than 0x10fff; these code points are
2411 excluded by RFC 3629.
2412 .sp
2413 PCRE_UTF8_ERR14
2414 .sp
2415 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2416 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2417 from UTF-8.
2418 .sp
2419 PCRE_UTF8_ERR15
2420 PCRE_UTF8_ERR16
2421 PCRE_UTF8_ERR17
2422 PCRE_UTF8_ERR18
2423 PCRE_UTF8_ERR19
2424 .sp
2425 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2426 value that can be represented by fewer bytes, which is invalid. For example,
2427 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2428 one byte.
2429 .sp
2430 PCRE_UTF8_ERR20
2431 .sp
2432 The two most significant bits of the first byte of a character have the binary
2433 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2434 byte can only validly occur as the second or subsequent byte of a multi-byte
2435 character.
2436 .sp
2437 PCRE_UTF8_ERR21
2438 .sp
2439 The first byte of a character has the value 0xfe or 0xff. These values can
2440 never occur in a valid UTF-8 string.
2441 .sp
2442 PCRE_UTF8_ERR22
2443 .sp
2444 This error code was formerly used when the presence of a so-called
2445 "non-character" caused an error. Unicode corrigendum #9 makes it clear that
2446 such characters should not cause a string to be rejected, and so this code is
2447 no longer in use and is never returned.
2448 .
2449 .
2450 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2451 .rs
2452 .sp
2453 .nf
2454 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2455 .B " int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,"
2456 .B " int \fIbuffersize\fP);"
2457 .sp
2458 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2459 .B " int \fIstringcount\fP, int \fIstringnumber\fP,"
2460 .B " const char **\fIstringptr\fP);"
2461 .sp
2462 .B int pcre_get_substring_list(const char *\fIsubject\fP,
2463 .B " int *\fIovector\fP, int \fIstringcount\fP, const char ***\fIlistptr\fP);"
2464 .fi
2465 .PP
2466 Captured substrings can be accessed directly by using the offsets returned by
2467 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2468 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2469 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2470 as new, separate, zero-terminated strings. These functions identify substrings
2471 by number. The next section describes functions for extracting named
2472 substrings.
2473 .P
2474 A substring that contains a binary zero is correctly extracted and has a
2475 further zero added on the end, but the result is not, of course, a C string.
2476 However, you can process such a string by referring to the length that is
2477 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2478 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2479 for handling strings containing binary zeros, because the end of the final
2480 string is not independently indicated.
2481 .P
2482 The first three arguments are the same for all three of these functions:
2483 \fIsubject\fP is the subject string that has just been successfully matched,
2484 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2485 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2486 captured by the match, including the substring that matched the entire regular
2487 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2488 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2489 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2490 number of elements in the vector divided by three.
2491 .P
2492 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2493 extract a single substring, whose number is given as \fIstringnumber\fP. A
2494 value of zero extracts the substring that matched the entire pattern, whereas
2495 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2496 the string is placed in \fIbuffer\fP, whose length is given by
2497 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2498 obtained via \fBpcre_malloc\fP, and its address is returned via
2499 \fIstringptr\fP. The yield of the function is the length of the string, not
2500 including the terminating zero, or one of these error codes:
2501 .sp
2502 PCRE_ERROR_NOMEMORY (-6)
2503 .sp
2504 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2505 memory failed for \fBpcre_get_substring()\fP.
2506 .sp
2507 PCRE_ERROR_NOSUBSTRING (-7)
2508 .sp
2509 There is no substring whose number is \fIstringnumber\fP.
2510 .P
2511 The \fBpcre_get_substring_list()\fP function extracts all available substrings
2512 and builds a list of pointers to them. All this is done in a single block of
2513 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2514 is returned via \fIlistptr\fP, which is also the start of the list of string
2515 pointers. The end of the list is marked by a NULL pointer. The yield of the
2516 function is zero if all went well, or the error code
2517 .sp
2518 PCRE_ERROR_NOMEMORY (-6)
2519 .sp
2520 if the attempt to get the memory block failed.
2521 .P
2522 When any of these functions encounter a substring that is unset, which can
2523 happen when capturing subpattern number \fIn+1\fP matches some part of the
2524 subject, but subpattern \fIn\fP has not been used at all, they return an empty
2525 string. This can be distinguished from a genuine zero-length substring by
2526 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2527 substrings.
2528 .P
2529 The two convenience functions \fBpcre_free_substring()\fP and
2530 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2531 a previous call of \fBpcre_get_substring()\fP or
2532 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2533 the function pointed to by \fBpcre_free\fP, which of course could be called
2534 directly from a C program. However, PCRE is used in some situations where it is
2535 linked via a special interface to another programming language that cannot use
2536 \fBpcre_free\fP directly; it is for these cases that the functions are
2537 provided.
2538 .
2539 .
2540 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2541 .rs
2542 .sp
2543 .nf
2544 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2545 .B " const char *\fIname\fP);"
2546 .sp
2547 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2548 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
2549 .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
2550 .B " char *\fIbuffer\fP, int \fIbuffersize\fP);"
2551 .sp
2552 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2553 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
2554 .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
2555 .B " const char **\fIstringptr\fP);"
2556 .fi
2557 .PP
2558 To extract a substring by name, you first have to find associated number.
2559 For example, for this pattern
2560 .sp
2561 (a+)b(?<xxx>\ed+)...
2562 .sp
2563 the number of the subpattern called "xxx" is 2. If the name is known to be
2564 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2565 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2566 pattern, and the second is the name. The yield of the function is the
2567 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2568 that name.
2569 .P
2570 Given the number, you can extract the substring directly, or use one of the
2571 functions described in the previous section. For convenience, there are also
2572 two functions that do the whole job.
2573 .P
2574 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2575 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2576 functions that extract by number. As these are described in the previous
2577 section, they are not re-described here. There are just two differences:
2578 .P
2579 First, instead of a substring number, a substring name is given. Second, there
2580 is an extra argument, given at the start, which is a pointer to the compiled
2581 pattern. This is needed in order to gain access to the name-to-number
2582 translation table.
2583 .P
2584 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2585 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2586 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2587 the behaviour may not be what you want (see the next section).
2588 .P
2589 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2590 subpatterns with the same number, as described in the
2591 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2592 .\" </a>
2593 section on duplicate subpattern numbers
2594 .\"
2595 in the
2596 .\" HREF
2597 \fBpcrepattern\fP
2598 .\"
2599 page, you cannot use names to distinguish the different subpatterns, because
2600 names are not included in the compiled code. The matching process uses only
2601 numbers. For this reason, the use of different names for subpatterns of the
2602 same number causes an error at compile time.
2603 .
2604 .
2605 .SH "DUPLICATE SUBPATTERN NAMES"
2606 .rs
2607 .sp
2608 .nf
2609 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2610 .B " const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);"
2611 .fi
2612 .PP
2613 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2614 are not required to be unique. (Duplicate names are always allowed for
2615 subpatterns with the same number, created by using the (?| feature. Indeed, if
2616 such subpatterns are named, they are required to use the same names.)
2617 .P
2618 Normally, patterns with duplicate names are such that in any one match, only
2619 one of the named subpatterns participates. An example is shown in the
2620 .\" HREF
2621 \fBpcrepattern\fP
2622 .\"
2623 documentation.
2624 .P
2625 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2626 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2627 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2628 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2629 returns one of the numbers that are associated with the name, but it is not
2630 defined which it is.
2631 .P
2632 If you want to get full details of all captured substrings for a given name,
2633 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2634 argument is the compiled pattern, and the second is the name. The third and
2635 fourth are pointers to variables which are updated by the function. After it
2636 has run, they point to the first and last entries in the name-to-number table
2637 for the given name. The function itself returns the length of each entry, or
2638 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2639 described above in the section entitled \fIInformation about a pattern\fP
2640 .\" HTML <a href="#infoaboutpattern">
2641 .\" </a>
2642 above.
2643 .\"
2644 Given all the relevant entries for the name, you can extract each of their
2645 numbers, and hence the captured data, if any.
2646 .
2647 .
2648 .SH "FINDING ALL POSSIBLE MATCHES"
2649 .rs
2650 .sp
2651 The traditional matching function uses a similar algorithm to Perl, which stops
2652 when it finds the first match, starting at a given point in the subject. If you
2653 want to find all possible matches, or the longest possible match, consider
2654 using the alternative matching function (see below) instead. If you cannot use
2655 the alternative function, but still need to find all possible matches, you
2656 can kludge it up by making use of the callout facility, which is described in
2657 the
2658 .\" HREF
2659 \fBpcrecallout\fP
2660 .\"
2661 documentation.
2662 .P
2663 What you have to do is to insert a callout right at the end of the pattern.
2664 When your callout function is called, extract and save the current matched
2665 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2666 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2667 will yield PCRE_ERROR_NOMATCH.
2668 .
2669 .
2670 .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2671 .rs
2672 .sp
2673 Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2674 stack, which in certain environments can be rather limited in size. Some users
2675 find it helpful to have an estimate of the amount of stack that is used by
2676 \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2677 .\" HREF
2678 \fBpcrestack\fP
2679 .\"
2680 documentation. The estimate that is output by \fBpcretest\fP when called with
2681 the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2682 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2683 .P
2684 Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2685 the negative error code PCRE_ERROR_NULL, but with this special combination of
2686 arguments, it returns instead a negative number whose absolute value is the
2687 approximate stack frame size in bytes. (A negative number is used so that it is
2688 clear that no match has happened.) The value is approximate because in some
2689 cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2690 additional variables on the stack.
2691 .P
2692 If PCRE has been compiled to use the heap instead of the stack for recursion,
2693 the value returned is the size of each block that is obtained from the heap.
2694 .
2695 .
2696 .\" HTML <a name="dfamatch"></a>
2697 .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2698 .rs
2699 .sp
2700 .nf
2701 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2702 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
2703 .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
2704 .B " int *\fIworkspace\fP, int \fIwscount\fP);"
2705 .fi
2706 .P
2707 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2708 a compiled pattern, using a matching algorithm that scans the subject string
2709 just once, and does not backtrack. This has different characteristics to the
2710 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2711 patterns are not supported. Nevertheless, there are times when this kind of
2712 matching can be useful. For a discussion of the two matching algorithms, and a
2713 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2714 .\" HREF
2715 \fBpcrematching\fP
2716 .\"
2717 documentation.
2718 .P
2719 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2720 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2721 different way, and this is described below. The other common arguments are used
2722 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2723 here.
2724 .P
2725 The two additional arguments provide workspace for the function. The workspace
2726 vector should contain at least 20 elements. It is used for keeping track of
2727 multiple paths through the pattern tree. More workspace will be needed for
2728 patterns and subjects where there are a lot of potential matches.
2729 .P
2730 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2731 .sp
2732 int rc;
2733 int ovector[10];
2734 int wspace[20];
2735 rc = pcre_dfa_exec(
2736 re, /* result of pcre_compile() */
2737 NULL, /* we didn't study the pattern */
2738 "some string", /* the subject string */
2739 11, /* the length of the subject string */
2740 0, /* start at offset 0 in the subject */
2741 0, /* default options */
2742 ovector, /* vector of integers for substring information */
2743 10, /* number of elements (NOT size in bytes) */
2744 wspace, /* working space vector */
2745 20); /* number of elements (NOT size in bytes) */
2746 .
2747 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2748 .rs
2749 .sp
2750 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2751 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2752 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2753 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2754 PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2755 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2756 so their description is not repeated here.
2757 .sp
2758 PCRE_PARTIAL_HARD
2759 PCRE_PARTIAL_SOFT
2760 .sp
2761 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2762 details are slightly different. When PCRE_PARTIAL_HARD is set for
2763 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2764 is reached and there is still at least one matching possibility that requires
2765 additional characters. This happens even if some complete matches have also
2766 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2767 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2768 there have been no complete matches, but there is still at least one matching
2769 possibility. The portion of the string that was inspected when the longest
2770 partial match was found is set as the first matching string in both cases.
2771 There is a more detailed discussion of partial and multi-segment matching, with
2772 examples, in the
2773 .\" HREF
2774 \fBpcrepartial\fP
2775 .\"
2776 documentation.
2777 .sp
2778 PCRE_DFA_SHORTEST
2779 .sp
2780 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2781 soon as it has found one match. Because of the way the alternative algorithm
2782 works, this is necessarily the shortest possible match at the first possible
2783 matching point in the subject string.
2784 .sp
2785 PCRE_DFA_RESTART
2786 .sp
2787 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2788 again, with additional subject characters, and have it continue with the same
2789 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2790 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2791 before because data about the match so far is left in them after a partial
2792 match. There is more discussion of this facility in the
2793 .\" HREF
2794 \fBpcrepartial\fP
2795 .\"
2796 documentation.
2797 .
2798 .
2799 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2800 .rs
2801 .sp
2802 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2803 substring in the subject. Note, however, that all the matches from one run of
2804 the function start at the same point in the subject. The shorter matches are
2805 all initial substrings of the longer matches. For example, if the pattern
2806 .sp
2807 <.*>
2808 .sp
2809 is matched against the string
2810 .sp
2811 This is <something> <something else> <something further> no more
2812 .sp
2813 the three matched strings are
2814 .sp
2815 <something>
2816 <something> <something else>
2817 <something> <something else> <something further>
2818 .sp
2819 On success, the yield of the function is a number greater than zero, which is
2820 the number of matched substrings. The substrings themselves are returned in
2821 \fIovector\fP. Each string uses two elements; the first is the offset to the
2822 start, and the second is the offset to the end. In fact, all the strings have
2823 the same start offset. (Space could have been saved by giving this only once,
2824 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2825 returns data, even though the meaning of the strings is different.)
2826 .P
2827 The strings are returned in reverse order of length; that is, the longest
2828 matching string is given first. If there were too many matches to fit into
2829 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2830 the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2831 the entire \fIovector\fP for returning matched strings.
2832 .P
2833 NOTE: PCRE's "auto-possessification" optimization usually applies to character
2834 repeats at the end of a pattern (as well as internally). For example, the
2835 pattern "a\ed+" is compiled as if it were "a\ed++" because there is no point
2836 even considering the possibility of backtracking into the repeated digits. For
2837 DFA matching, this means that only one possible match is found. If you really
2838 do want multiple matches in such cases, either use an ungreedy repeat
2839 ("a\ed+?") or set the PCRE_NO_AUTO_POSSESS option when compiling.
2840 .
2841 .
2842 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2843 .rs
2844 .sp
2845 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2846 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2847 described
2848 .\" HTML <a href="#errorlist">
2849 .\" </a>
2850 above.
2851 .\"
2852 There are in addition the following errors that are specific to
2853 \fBpcre_dfa_exec()\fP:
2854 .sp
2855 PCRE_ERROR_DFA_UITEM (-16)
2856 .sp
2857 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2858 that it does not support, for instance, the use of \eC or a back reference.
2859 .sp
2860 PCRE_ERROR_DFA_UCOND (-17)
2861 .sp
2862 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2863 uses a back reference for the condition, or a test for recursion in a specific
2864 group. These are not supported.
2865 .sp
2866 PCRE_ERROR_DFA_UMLIMIT (-18)
2867 .sp
2868 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2869 block that contains a setting of the \fImatch_limit\fP or
2870 \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2871 meaningless for DFA matching).
2872 .sp
2873 PCRE_ERROR_DFA_WSSIZE (-19)
2874 .sp
2875 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2876 \fIworkspace\fP vector.
2877 .sp
2878 PCRE_ERROR_DFA_RECURSE (-20)
2879 .sp
2880 When a recursive subpattern is processed, the matching function calls itself
2881 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2882 error is given if the output vector is not large enough. This should be
2883 extremely rare, as a vector of size 1000 is used.
2884 .sp
2885 PCRE_ERROR_DFA_BADRESTART (-30)
2886 .sp
2887 When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2888 some plausibility checks are made on the contents of the workspace, which
2889 should contain data about the previous partial match. If any of these checks
2890 fail, this error is given.
2891 .
2892 .
2893 .SH "SEE ALSO"
2894 .rs
2895 .sp
2896 \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2897 \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2898 \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2899 \fBpcrestack\fP(3).
2900 .
2901 .
2902 .SH AUTHOR
2903 .rs
2904 .sp
2905 .nf
2906 Philip Hazel
2907 University Computing Service
2908 Cambridge CB2 3QH, England.
2909 .fi
2910 .
2911 .
2912 .SH REVISION
2913 .rs
2914 .sp
2915 .nf
2916 Last updated: 18 December 2015
2917 Copyright (c) 1997-2015 University of Cambridge.
2918 .fi

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5