ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log

Revision 692 - (show annotations)
Mon Sep 12 11:38:03 2011 UTC (8 years, 9 months ago) by ph10
File MIME type: text/html
File size: 114631 byte(s)
Fix CMake for release candidate; doc tidies.
1 <html>
2 <head>
3 <title>pcreapi specification</title>
4 </head>
5 <body bgcolor="#FFFFFF" text="#00005A" link="#0066FF" alink="#3399FF" vlink="#2222BB">
6 <h1>pcreapi man page</h1>
7 <p>
8 Return to the <a href="index.html">PCRE index page</a>.
9 </p>
10 <p>
11 This page is part of the PCRE HTML documentation. It was generated automatically
12 from the original man page. If there is any nonsense in it, please consult the
13 man page, in case the conversion went wrong.
14 <br>
15 <ul>
16 <li><a name="TOC1" href="#SEC1">PCRE NATIVE API BASIC FUNCTIONS</a>
17 <li><a name="TOC2" href="#SEC2">PCRE NATIVE API AUXILIARY FUNCTIONS</a>
19 <li><a name="TOC4" href="#SEC4">PCRE API OVERVIEW</a>
20 <li><a name="TOC5" href="#SEC5">NEWLINES</a>
21 <li><a name="TOC6" href="#SEC6">MULTITHREADING</a>
23 <li><a name="TOC8" href="#SEC8">CHECKING BUILD-TIME OPTIONS</a>
24 <li><a name="TOC9" href="#SEC9">COMPILING A PATTERN</a>
25 <li><a name="TOC10" href="#SEC10">COMPILATION ERROR CODES</a>
26 <li><a name="TOC11" href="#SEC11">STUDYING A PATTERN</a>
27 <li><a name="TOC12" href="#SEC12">LOCALE SUPPORT</a>
28 <li><a name="TOC13" href="#SEC13">INFORMATION ABOUT A PATTERN</a>
29 <li><a name="TOC14" href="#SEC14">OBSOLETE INFO FUNCTION</a>
30 <li><a name="TOC15" href="#SEC15">REFERENCE COUNTS</a>
34 <li><a name="TOC19" href="#SEC19">DUPLICATE SUBPATTERN NAMES</a>
35 <li><a name="TOC20" href="#SEC20">FINDING ALL POSSIBLE MATCHES</a>
37 <li><a name="TOC22" href="#SEC22">SEE ALSO</a>
38 <li><a name="TOC23" href="#SEC23">AUTHOR</a>
39 <li><a name="TOC24" href="#SEC24">REVISION</a>
40 </ul>
41 <br><a name="SEC1" href="#TOC1">PCRE NATIVE API BASIC FUNCTIONS</a><br>
42 <P>
43 <b>#include &#60;pcre.h&#62;</b>
44 </P>
45 <P>
46 <b>pcre *pcre_compile(const char *<i>pattern</i>, int <i>options</i>,</b>
47 <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
48 <b>const unsigned char *<i>tableptr</i>);</b>
49 </P>
50 <P>
51 <b>pcre *pcre_compile2(const char *<i>pattern</i>, int <i>options</i>,</b>
52 <b>int *<i>errorcodeptr</i>,</b>
53 <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
54 <b>const unsigned char *<i>tableptr</i>);</b>
55 </P>
56 <P>
57 <b>pcre_extra *pcre_study(const pcre *<i>code</i>, int <i>options</i>,</b>
58 <b>const char **<i>errptr</i>);</b>
59 </P>
60 <P>
61 <b>void pcre_free_study(pcre_extra *<i>extra</i>);</b>
62 </P>
63 <P>
64 <b>int pcre_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
65 <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
66 <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>);</b>
67 </P>
68 <br><a name="SEC2" href="#TOC1">PCRE NATIVE API AUXILIARY FUNCTIONS</a><br>
69 <P>
70 <b>pcre_jit_stack *pcre_jit_stack_alloc(int <i>startsize</i>, int <i>maxsize</i>);</b>
71 </P>
72 <P>
73 <b>void pcre_jit_stack_free(pcre_jit_stack *<i>stack</i>);</b>
74 </P>
75 <P>
76 <b>void pcre_assign_jit_stack(pcre_extra *<i>extra</i>,</b>
77 <b>pcre_jit_callback <i>callback</i>, void *<i>data</i>);</b>
78 </P>
79 <P>
80 <b>int pcre_dfa_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
81 <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
82 <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>,</b>
83 <b>int *<i>workspace</i>, int <i>wscount</i>);</b>
84 </P>
85 <P>
86 <b>int pcre_copy_named_substring(const pcre *<i>code</i>,</b>
87 <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
88 <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
89 <b>char *<i>buffer</i>, int <i>buffersize</i>);</b>
90 </P>
91 <P>
92 <b>int pcre_copy_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
93 <b>int <i>stringcount</i>, int <i>stringnumber</i>, char *<i>buffer</i>,</b>
94 <b>int <i>buffersize</i>);</b>
95 </P>
96 <P>
97 <b>int pcre_get_named_substring(const pcre *<i>code</i>,</b>
98 <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
99 <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
100 <b>const char **<i>stringptr</i>);</b>
101 </P>
102 <P>
103 <b>int pcre_get_stringnumber(const pcre *<i>code</i>,</b>
104 <b>const char *<i>name</i>);</b>
105 </P>
106 <P>
107 <b>int pcre_get_stringtable_entries(const pcre *<i>code</i>,</b>
108 <b>const char *<i>name</i>, char **<i>first</i>, char **<i>last</i>);</b>
109 </P>
110 <P>
111 <b>int pcre_get_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
112 <b>int <i>stringcount</i>, int <i>stringnumber</i>,</b>
113 <b>const char **<i>stringptr</i>);</b>
114 </P>
115 <P>
116 <b>int pcre_get_substring_list(const char *<i>subject</i>,</b>
117 <b>int *<i>ovector</i>, int <i>stringcount</i>, const char ***<i>listptr</i>);</b>
118 </P>
119 <P>
120 <b>void pcre_free_substring(const char *<i>stringptr</i>);</b>
121 </P>
122 <P>
123 <b>void pcre_free_substring_list(const char **<i>stringptr</i>);</b>
124 </P>
125 <P>
126 <b>const unsigned char *pcre_maketables(void);</b>
127 </P>
128 <P>
129 <b>int pcre_fullinfo(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
130 <b>int <i>what</i>, void *<i>where</i>);</b>
131 </P>
132 <P>
133 <b>int pcre_info(const pcre *<i>code</i>, int *<i>optptr</i>, int</b>
134 <b>*<i>firstcharptr</i>);</b>
135 </P>
136 <P>
137 <b>int pcre_refcount(pcre *<i>code</i>, int <i>adjust</i>);</b>
138 </P>
139 <P>
140 <b>int pcre_config(int <i>what</i>, void *<i>where</i>);</b>
141 </P>
142 <P>
143 <b>char *pcre_version(void);</b>
144 </P>
145 <br><a name="SEC3" href="#TOC1">PCRE NATIVE API INDIRECTED FUNCTIONS</a><br>
146 <P>
147 <b>void *(*pcre_malloc)(size_t);</b>
148 </P>
149 <P>
150 <b>void (*pcre_free)(void *);</b>
151 </P>
152 <P>
153 <b>void *(*pcre_stack_malloc)(size_t);</b>
154 </P>
155 <P>
156 <b>void (*pcre_stack_free)(void *);</b>
157 </P>
158 <P>
159 <b>int (*pcre_callout)(pcre_callout_block *);</b>
160 </P>
161 <br><a name="SEC4" href="#TOC1">PCRE API OVERVIEW</a><br>
162 <P>
163 PCRE has its own native API, which is described in this document. There are
164 also some wrapper functions that correspond to the POSIX regular expression
165 API, but they do not give access to all the functionality. They are described
166 in the
167 <a href="pcreposix.html"><b>pcreposix</b></a>
168 documentation. Both of these APIs define a set of C function calls. A C++
169 wrapper is also distributed with PCRE. It is documented in the
170 <a href="pcrecpp.html"><b>pcrecpp</b></a>
171 page.
172 </P>
173 <P>
174 The native API C function prototypes are defined in the header file
175 <b>pcre.h</b>, and on Unix systems the library itself is called <b>libpcre</b>.
176 It can normally be accessed by adding <b>-lpcre</b> to the command for linking
177 an application that uses PCRE. The header file defines the macros PCRE_MAJOR
178 and PCRE_MINOR to contain the major and minor release numbers for the library.
179 Applications can use these to include support for different releases of PCRE.
180 </P>
181 <P>
182 In a Windows environment, if you want to statically link an application program
183 against a non-dll <b>pcre.a</b> file, you must define PCRE_STATIC before
184 including <b>pcre.h</b> or <b>pcrecpp.h</b>, because otherwise the
185 <b>pcre_malloc()</b> and <b>pcre_free()</b> exported functions will be declared
186 <b>__declspec(dllimport)</b>, with unwanted results.
187 </P>
188 <P>
189 The functions <b>pcre_compile()</b>, <b>pcre_compile2()</b>, <b>pcre_study()</b>,
190 and <b>pcre_exec()</b> 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 <i>pcredemo.c</i> in the PCRE
193 source distribution. A listing of this program is given in the
194 <a href="pcredemo.html"><b>pcredemo</b></a>
195 documentation, and the
196 <a href="pcresample.html"><b>pcresample</b></a>
197 documentation describes how to compile and run it.
198 </P>
199 <P>
200 Just-in-time compiler support is an optional feature of PCRE that can be built
201 in appropriate hardware environments. It greatly speeds up the matching
202 performance of many patterns. Simple programs can easily request that it be
203 used if available, by setting an option that is ignored when it is not
204 relevant. More complicated programs might need to make use of the functions
205 <b>pcre_jit_stack_alloc()</b>, <b>pcre_jit_stack_free()</b>, and
206 <b>pcre_assign_jit_stack()</b> in order to control the JIT code's memory usage.
207 These functions are discussed in the
208 <a href="pcrejit.html"><b>pcrejit</b></a>
209 documentation.
210 </P>
211 <P>
212 A second matching function, <b>pcre_dfa_exec()</b>, which is not
213 Perl-compatible, is also provided. This uses a different algorithm for the
214 matching. The alternative algorithm finds all possible matches (at a given
215 point in the subject), and scans the subject just once (unless there are
216 lookbehind assertions). However, this algorithm does not return captured
217 substrings. A description of the two matching algorithms and their advantages
218 and disadvantages is given in the
219 <a href="pcrematching.html"><b>pcrematching</b></a>
220 documentation.
221 </P>
222 <P>
223 In addition to the main compiling and matching functions, there are convenience
224 functions for extracting captured substrings from a subject string that is
225 matched by <b>pcre_exec()</b>. They are:
226 <pre>
227 <b>pcre_copy_substring()</b>
228 <b>pcre_copy_named_substring()</b>
229 <b>pcre_get_substring()</b>
230 <b>pcre_get_named_substring()</b>
231 <b>pcre_get_substring_list()</b>
232 <b>pcre_get_stringnumber()</b>
233 <b>pcre_get_stringtable_entries()</b>
234 </pre>
235 <b>pcre_free_substring()</b> and <b>pcre_free_substring_list()</b> are also
236 provided, to free the memory used for extracted strings.
237 </P>
238 <P>
239 The function <b>pcre_maketables()</b> is used to build a set of character tables
240 in the current locale for passing to <b>pcre_compile()</b>, <b>pcre_exec()</b>,
241 or <b>pcre_dfa_exec()</b>. This is an optional facility that is provided for
242 specialist use. Most commonly, no special tables are passed, in which case
243 internal tables that are generated when PCRE is built are used.
244 </P>
245 <P>
246 The function <b>pcre_fullinfo()</b> is used to find out information about a
247 compiled pattern; <b>pcre_info()</b> is an obsolete version that returns only
248 some of the available information, but is retained for backwards compatibility.
249 The function <b>pcre_version()</b> returns a pointer to a string containing the
250 version of PCRE and its date of release.
251 </P>
252 <P>
253 The function <b>pcre_refcount()</b> maintains a reference count in a data block
254 containing a compiled pattern. This is provided for the benefit of
255 object-oriented applications.
256 </P>
257 <P>
258 The global variables <b>pcre_malloc</b> and <b>pcre_free</b> initially contain
259 the entry points of the standard <b>malloc()</b> and <b>free()</b> functions,
260 respectively. PCRE calls the memory management functions via these variables,
261 so a calling program can replace them if it wishes to intercept the calls. This
262 should be done before calling any PCRE functions.
263 </P>
264 <P>
265 The global variables <b>pcre_stack_malloc</b> and <b>pcre_stack_free</b> 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 <b>pcre_exec()</b> function. See the
269 <a href="pcrebuild.html"><b>pcrebuild</b></a>
270 documentation for details of how to do this. It is a non-standard way of
271 building PCRE, for use in environments that have limited stacks. Because of the
272 greater use of memory management, it runs more slowly. Separate functions are
273 provided so that special-purpose external code can be used for this case. When
274 used, these functions are always called in a stack-like manner (last obtained,
275 first freed), and always for memory blocks of the same size. There is a
276 discussion about PCRE's stack usage in the
277 <a href="pcrestack.html"><b>pcrestack</b></a>
278 documentation.
279 </P>
280 <P>
281 The global variable <b>pcre_callout</b> initially contains NULL. It can be set
282 by the caller to a "callout" function, which PCRE will then call at specified
283 points during a matching operation. Details are given in the
284 <a href="pcrecallout.html"><b>pcrecallout</b></a>
285 documentation.
286 <a name="newlines"></a></P>
287 <br><a name="SEC5" href="#TOC1">NEWLINES</a><br>
288 <P>
289 PCRE supports five different conventions for indicating line breaks in
290 strings: a single CR (carriage return) character, a single LF (linefeed)
291 character, the two-character sequence CRLF, any of the three preceding, or any
292 Unicode newline sequence. The Unicode newline sequences are the three just
293 mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
294 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
295 (paragraph separator, U+2029).
296 </P>
297 <P>
298 Each of the first three conventions is used by at least one operating system as
299 its standard newline sequence. When PCRE is built, a default can be specified.
300 The default default is LF, which is the Unix standard. When PCRE is run, the
301 default can be overridden, either when a pattern is compiled, or when it is
302 matched.
303 </P>
304 <P>
305 At compile time, the newline convention can be specified by the <i>options</i>
306 argument of <b>pcre_compile()</b>, or it can be specified by special text at the
307 start of the pattern itself; this overrides any other settings. See the
308 <a href="pcrepattern.html"><b>pcrepattern</b></a>
309 page for details of the special character sequences.
310 </P>
311 <P>
312 In the PCRE documentation the word "newline" is used to mean "the character or
313 pair of characters that indicate a line break". The choice of newline
314 convention affects the handling of the dot, circumflex, and dollar
315 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
316 recognized line ending sequence, the match position advancement for a
317 non-anchored pattern. There is more detail about this in the
318 <a href="#execoptions">section on <b>pcre_exec()</b> options</a>
319 below.
320 </P>
321 <P>
322 The choice of newline convention does not affect the interpretation of
323 the \n or \r escape sequences, nor does it affect what \R matches, which is
324 controlled in a similar way, but by separate options.
325 </P>
326 <br><a name="SEC6" href="#TOC1">MULTITHREADING</a><br>
327 <P>
328 The PCRE functions can be used in multi-threading applications, with the
329 proviso that the memory management functions pointed to by <b>pcre_malloc</b>,
330 <b>pcre_free</b>, <b>pcre_stack_malloc</b>, and <b>pcre_stack_free</b>, and the
331 callout function pointed to by <b>pcre_callout</b>, are shared by all threads.
332 </P>
333 <P>
334 The compiled form of a regular expression is not altered during matching, so
335 the same compiled pattern can safely be used by several threads at once.
336 </P>
337 <P>
338 If the just-in-time optimization feature is being used, it needs separate
339 memory stack areas for each thread. See the
340 <a href="pcrejit.html"><b>pcrejit</b></a>
341 documentation for more details.
342 </P>
343 <br><a name="SEC7" href="#TOC1">SAVING PRECOMPILED PATTERNS FOR LATER USE</a><br>
344 <P>
345 The compiled form of a regular expression can be saved and re-used at a later
346 time, possibly by a different program, and even on a host other than the one on
347 which it was compiled. Details are given in the
348 <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
349 documentation. However, compiling a regular expression with one version of PCRE
350 for use with a different version is not guaranteed to work and may cause
351 crashes.
352 </P>
353 <br><a name="SEC8" href="#TOC1">CHECKING BUILD-TIME OPTIONS</a><br>
354 <P>
355 <b>int pcre_config(int <i>what</i>, void *<i>where</i>);</b>
356 </P>
357 <P>
358 The function <b>pcre_config()</b> makes it possible for a PCRE client to
359 discover which optional features have been compiled into the PCRE library. The
360 <a href="pcrebuild.html"><b>pcrebuild</b></a>
361 documentation has more details about these optional features.
362 </P>
363 <P>
364 The first argument for <b>pcre_config()</b> is an integer, specifying which
365 information is required; the second argument is a pointer to a variable into
366 which the information is placed. The following information is available:
367 <pre>
369 </pre>
370 The output is an integer that is set to one if UTF-8 support is available;
371 otherwise it is set to zero.
372 <pre>
374 </pre>
375 The output is an integer that is set to one if support for Unicode character
376 properties is available; otherwise it is set to zero.
377 <pre>
379 </pre>
380 The output is an integer that is set to one if support for just-in-time
381 compiling is available; otherwise it is set to zero.
382 <pre>
384 </pre>
385 The output is an integer whose value specifies the default character sequence
386 that is recognized as meaning "newline". The four values that are supported
387 are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF, and -1 for ANY.
388 Though they are derived from ASCII, the same values are returned in EBCDIC
389 environments. The default should normally correspond to the standard sequence
390 for your operating system.
391 <pre>
393 </pre>
394 The output is an integer whose value indicates what character sequences the \R
395 escape sequence matches by default. A value of 0 means that \R matches any
396 Unicode line ending sequence; a value of 1 means that \R matches only CR, LF,
397 or CRLF. The default can be overridden when a pattern is compiled or matched.
398 <pre>
400 </pre>
401 The output is an integer that contains the number of bytes used for internal
402 linkage in compiled regular expressions. The value is 2, 3, or 4. Larger values
403 allow larger regular expressions to be compiled, at the expense of slower
404 matching. The default value of 2 is sufficient for all but the most massive
405 patterns, since it allows the compiled pattern to be up to 64K in size.
406 <pre>
408 </pre>
409 The output is an integer that contains the threshold above which the POSIX
410 interface uses <b>malloc()</b> for output vectors. Further details are given in
411 the
412 <a href="pcreposix.html"><b>pcreposix</b></a>
413 documentation.
414 <pre>
416 </pre>
417 The output is a long integer that gives the default limit for the number of
418 internal matching function calls in a <b>pcre_exec()</b> execution. Further
419 details are given with <b>pcre_exec()</b> below.
420 <pre>
422 </pre>
423 The output is a long integer that gives the default limit for the depth of
424 recursion when calling the internal matching function in a <b>pcre_exec()</b>
425 execution. Further details are given with <b>pcre_exec()</b> below.
426 <pre>
428 </pre>
429 The output is an integer that is set to one if internal recursion when running
430 <b>pcre_exec()</b> is implemented by recursive function calls that use the stack
431 to remember their state. This is the usual way that PCRE is compiled. The
432 output is zero if PCRE was compiled to use blocks of data on the heap instead
433 of recursive function calls. In this case, <b>pcre_stack_malloc</b> and
434 <b>pcre_stack_free</b> are called to manage memory blocks on the heap, thus
435 avoiding the use of the stack.
436 </P>
437 <br><a name="SEC9" href="#TOC1">COMPILING A PATTERN</a><br>
438 <P>
439 <b>pcre *pcre_compile(const char *<i>pattern</i>, int <i>options</i>,</b>
440 <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
441 <b>const unsigned char *<i>tableptr</i>);</b>
442 <b>pcre *pcre_compile2(const char *<i>pattern</i>, int <i>options</i>,</b>
443 <b>int *<i>errorcodeptr</i>,</b>
444 <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
445 <b>const unsigned char *<i>tableptr</i>);</b>
446 </P>
447 <P>
448 Either of the functions <b>pcre_compile()</b> or <b>pcre_compile2()</b> can be
449 called to compile a pattern into an internal form. The only difference between
450 the two interfaces is that <b>pcre_compile2()</b> has an additional argument,
451 <i>errorcodeptr</i>, via which a numerical error code can be returned. To avoid
452 too much repetition, we refer just to <b>pcre_compile()</b> below, but the
453 information applies equally to <b>pcre_compile2()</b>.
454 </P>
455 <P>
456 The pattern is a C string terminated by a binary zero, and is passed in the
457 <i>pattern</i> argument. A pointer to a single block of memory that is obtained
458 via <b>pcre_malloc</b> is returned. This contains the compiled code and related
459 data. The <b>pcre</b> type is defined for the returned block; this is a typedef
460 for a structure whose contents are not externally defined. It is up to the
461 caller to free the memory (via <b>pcre_free</b>) when it is no longer required.
462 </P>
463 <P>
464 Although the compiled code of a PCRE regex is relocatable, that is, it does not
465 depend on memory location, the complete <b>pcre</b> data block is not
466 fully relocatable, because it may contain a copy of the <i>tableptr</i>
467 argument, which is an address (see below).
468 </P>
469 <P>
470 The <i>options</i> argument contains various bit settings that affect the
471 compilation. It should be zero if no options are required. The available
472 options are described below. Some of them (in particular, those that are
473 compatible with Perl, but some others as well) can also be set and unset from
474 within the pattern (see the detailed description in the
475 <a href="pcrepattern.html"><b>pcrepattern</b></a>
476 documentation). For those options that can be different in different parts of
477 the pattern, the contents of the <i>options</i> argument specifies their
478 settings at the start of compilation and execution. The PCRE_ANCHORED,
479 PCRE_BSR_<i>xxx</i>, PCRE_NEWLINE_<i>xxx</i>, PCRE_NO_UTF8_CHECK, and
480 PCRE_NO_START_OPT options can be set at the time of matching as well as at
481 compile time.
482 </P>
483 <P>
484 If <i>errptr</i> is NULL, <b>pcre_compile()</b> returns NULL immediately.
485 Otherwise, if compilation of a pattern fails, <b>pcre_compile()</b> returns
486 NULL, and sets the variable pointed to by <i>errptr</i> to point to a textual
487 error message. This is a static string that is part of the library. You must
488 not try to free it. Normally, the offset from the start of the pattern to the
489 byte that was being processed when the error was discovered is placed in the
490 variable pointed to by <i>erroffset</i>, which must not be NULL (if it is, an
491 immediate error is given). However, for an invalid UTF-8 string, the offset is
492 that of the first byte of the failing character. Also, some errors are not
493 detected until checks are carried out when the whole pattern has been scanned;
494 in these cases the offset passed back is the length of the pattern.
495 </P>
496 <P>
497 Note that the offset is in bytes, not characters, even in UTF-8 mode. It may
498 sometimes point into the middle of a UTF-8 character.
499 </P>
500 <P>
501 If <b>pcre_compile2()</b> is used instead of <b>pcre_compile()</b>, and the
502 <i>errorcodeptr</i> argument is not NULL, a non-zero error code number is
503 returned via this argument in the event of an error. This is in addition to the
504 textual error message. Error codes and messages are listed below.
505 </P>
506 <P>
507 If the final argument, <i>tableptr</i>, is NULL, PCRE uses a default set of
508 character tables that are built when PCRE is compiled, using the default C
509 locale. Otherwise, <i>tableptr</i> must be an address that is the result of a
510 call to <b>pcre_maketables()</b>. This value is stored with the compiled
511 pattern, and used again by <b>pcre_exec()</b>, unless another table pointer is
512 passed to it. For more discussion, see the section on locale support below.
513 </P>
514 <P>
515 This code fragment shows a typical straightforward call to <b>pcre_compile()</b>:
516 <pre>
517 pcre *re;
518 const char *error;
519 int erroffset;
520 re = pcre_compile(
521 "^A.*Z", /* the pattern */
522 0, /* default options */
523 &error, /* for error message */
524 &erroffset, /* for error offset */
525 NULL); /* use default character tables */
526 </pre>
527 The following names for option bits are defined in the <b>pcre.h</b> header
528 file:
529 <pre>
531 </pre>
532 If this bit is set, the pattern is forced to be "anchored", that is, it is
533 constrained to match only at the first matching point in the string that is
534 being searched (the "subject string"). This effect can also be achieved by
535 appropriate constructs in the pattern itself, which is the only way to do it in
536 Perl.
537 <pre>
539 </pre>
540 If this bit is set, <b>pcre_compile()</b> automatically inserts callout items,
541 all with number 255, before each pattern item. For discussion of the callout
542 facility, see the
543 <a href="pcrecallout.html"><b>pcrecallout</b></a>
544 documentation.
545 <pre>
548 </pre>
549 These options (which are mutually exclusive) control what the \R escape
550 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
551 match any Unicode newline sequence. The default is specified when PCRE is
552 built. It can be overridden from within the pattern, or by setting an option
553 when a compiled pattern is matched.
554 <pre>
556 </pre>
557 If this bit is set, letters in the pattern match both upper and lower case
558 letters. It is equivalent to Perl's /i option, and it can be changed within a
559 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
560 concept of case for characters whose values are less than 128, so caseless
561 matching is always possible. For characters with higher values, the concept of
562 case is supported if PCRE is compiled with Unicode property support, but not
563 otherwise. If you want to use caseless matching for characters 128 and above,
564 you must ensure that PCRE is compiled with Unicode property support as well as
565 with UTF-8 support.
566 <pre>
568 </pre>
569 If this bit is set, a dollar metacharacter in the pattern matches only at the
570 end of the subject string. Without this option, a dollar also matches
571 immediately before a newline at the end of the string (but not before any other
572 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
573 There is no equivalent to this option in Perl, and no way to set it within a
574 pattern.
575 <pre>
577 </pre>
578 If this bit is set, a dot metacharacter in the pattern matches a character of
579 any value, including one that indicates a newline. However, it only ever
580 matches one character, even if newlines are coded as CRLF. Without this option,
581 a dot does not match when the current position is at a newline. This option is
582 equivalent to Perl's /s option, and it can be changed within a pattern by a
583 (?s) option setting. A negative class such as [^a] always matches newline
584 characters, independent of the setting of this option.
585 <pre>
587 </pre>
588 If this bit is set, names used to identify capturing subpatterns need not be
589 unique. This can be helpful for certain types of pattern when it is known that
590 only one instance of the named subpattern can ever be matched. There are more
591 details of named subpatterns below; see also the
592 <a href="pcrepattern.html"><b>pcrepattern</b></a>
593 documentation.
594 <pre>
596 </pre>
597 If this bit is set, whitespace data characters in the pattern are totally
598 ignored except when escaped or inside a character class. Whitespace does not
599 include the VT character (code 11). In addition, characters between an
600 unescaped # outside a character class and the next newline, inclusive, are also
601 ignored. This is equivalent to Perl's /x option, and it can be changed within a
602 pattern by a (?x) option setting.
603 </P>
604 <P>
605 Which characters are interpreted as newlines is controlled by the options
606 passed to <b>pcre_compile()</b> or by a special sequence at the start of the
607 pattern, as described in the section entitled
608 <a href="pcrepattern.html#newlines">"Newline conventions"</a>
609 in the <b>pcrepattern</b> documentation. Note that the end of this type of
610 comment is a literal newline sequence in the pattern; escape sequences that
611 happen to represent a newline do not count.
612 </P>
613 <P>
614 This option makes it possible to include comments inside complicated patterns.
615 Note, however, that this applies only to data characters. Whitespace characters
616 may never appear within special character sequences in a pattern, for example
617 within the sequence (?( that introduces a conditional subpattern.
618 <pre>
620 </pre>
621 This option was invented in order to turn on additional functionality of PCRE
622 that is incompatible with Perl, but it is currently of very little use. When
623 set, any backslash in a pattern that is followed by a letter that has no
624 special meaning causes an error, thus reserving these combinations for future
625 expansion. By default, as in Perl, a backslash followed by a letter with no
626 special meaning is treated as a literal. (Perl can, however, be persuaded to
627 give an error for this, by running it with the -w option.) There are at present
628 no other features controlled by this option. It can also be set by a (?X)
629 option setting within a pattern.
630 <pre>
632 </pre>
633 If this option is set, an unanchored pattern is required to match before or at
634 the first newline in the subject string, though the matched text may continue
635 over the newline.
636 <pre>
638 </pre>
639 If this option is set, PCRE's behaviour is changed in some ways so that it is
640 compatible with JavaScript rather than Perl. The changes are as follows:
641 </P>
642 <P>
643 (1) A lone closing square bracket in a pattern causes a compile-time error,
644 because this is illegal in JavaScript (by default it is treated as a data
645 character). Thus, the pattern AB]CD becomes illegal when this option is set.
646 </P>
647 <P>
648 (2) At run time, a back reference to an unset subpattern group matches an empty
649 string (by default this causes the current matching alternative to fail). A
650 pattern such as (\1)(a) succeeds when this option is set (assuming it can find
651 an "a" in the subject), whereas it fails by default, for Perl compatibility.
652 <pre>
654 </pre>
655 By default, PCRE treats the subject string as consisting of a single line of
656 characters (even if it actually contains newlines). The "start of line"
657 metacharacter (^) matches only at the start of the string, while the "end of
658 line" metacharacter ($) matches only at the end of the string, or before a
659 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
660 Perl.
661 </P>
662 <P>
663 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
664 match immediately following or immediately before internal newlines in the
665 subject string, respectively, as well as at the very start and end. This is
666 equivalent to Perl's /m option, and it can be changed within a pattern by a
667 (?m) option setting. If there are no newlines in a subject string, or no
668 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
669 <pre>
675 </pre>
676 These options override the default newline definition that was chosen when PCRE
677 was built. Setting the first or the second specifies that a newline is
678 indicated by a single character (CR or LF, respectively). Setting
679 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
680 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
681 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
682 that any Unicode newline sequence should be recognized. The Unicode newline
683 sequences are the three just mentioned, plus the single characters VT (vertical
684 tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
685 separator, U+2028), and PS (paragraph separator, U+2029). The last two are
686 recognized only in UTF-8 mode.
687 </P>
688 <P>
689 The newline setting in the options word uses three bits that are treated
690 as a number, giving eight possibilities. Currently only six are used (default
691 plus the five values above). This means that if you set more than one newline
692 option, the combination may or may not be sensible. For example,
694 other combinations may yield unused numbers and cause an error.
695 </P>
696 <P>
697 The only time that a line break in a pattern is specially recognized when
698 compiling is when PCRE_EXTENDED is set. CR and LF are whitespace characters,
699 and so are ignored in this mode. Also, an unescaped # outside a character class
700 indicates a comment that lasts until after the next line break sequence. In
701 other circumstances, line break sequences in patterns are treated as literal
702 data.
703 </P>
704 <P>
705 The newline option that is set at compile time becomes the default that is used
706 for <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, but it can be overridden.
707 <pre>
709 </pre>
710 If this option is set, it disables the use of numbered capturing parentheses in
711 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
712 were followed by ?: but named parentheses can still be used for capturing (and
713 they acquire numbers in the usual way). There is no equivalent of this option
714 in Perl.
715 <pre>
717 </pre>
718 This is an option that acts at matching time; that is, it is really an option
719 for <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b>. If it is set at compile time,
720 it is remembered with the compiled pattern and assumed at matching time. For
721 details see the discussion of PCRE_NO_START_OPTIMIZE
722 <a href="#execoptions">below.</a>
723 <pre>
725 </pre>
726 This option changes the way PCRE processes \B, \b, \D, \d, \S, \s, \W,
727 \w, and some of the POSIX character classes. By default, only ASCII characters
728 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
729 classify characters. More details are given in the section on
730 <a href="pcre.html#genericchartypes">generic character types</a>
731 in the
732 <a href="pcrepattern.html"><b>pcrepattern</b></a>
733 page. If you set PCRE_UCP, matching one of the items it affects takes much
734 longer. The option is available only if PCRE has been compiled with Unicode
735 property support.
736 <pre>
738 </pre>
739 This option inverts the "greediness" of the quantifiers so that they are not
740 greedy by default, but become greedy if followed by "?". It is not compatible
741 with Perl. It can also be set by a (?U) option setting within the pattern.
742 <pre>
744 </pre>
745 This option causes PCRE to regard both the pattern and the subject as strings
746 of UTF-8 characters instead of single-byte character strings. However, it is
747 available only when PCRE is built to include UTF-8 support. If not, the use
748 of this option provokes an error. Details of how this option changes the
749 behaviour of PCRE are given in the
750 <a href="pcreunicode.html"><b>pcreunicode</b></a>
751 page.
752 <pre>
754 </pre>
755 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
756 automatically checked. There is a discussion about the
757 <a href="pcre.html#utf8strings">validity of UTF-8 strings</a>
758 in the main
759 <a href="pcre.html"><b>pcre</b></a>
760 page. If an invalid UTF-8 sequence of bytes is found, <b>pcre_compile()</b>
761 returns an error. If you already know that your pattern is valid, and you want
762 to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
763 option. When it is set, the effect of passing an invalid UTF-8 string as a
764 pattern is undefined. It may cause your program to crash. Note that this option
765 can also be passed to <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, to suppress
766 the UTF-8 validity checking of subject strings.
767 </P>
768 <br><a name="SEC10" href="#TOC1">COMPILATION ERROR CODES</a><br>
769 <P>
770 The following table lists the error codes than may be returned by
771 <b>pcre_compile2()</b>, along with the error messages that may be returned by
772 both compiling functions. As PCRE has developed, some error codes have fallen
773 out of use. To avoid confusion, they have not been re-used.
774 <pre>
775 0 no error
776 1 \ at end of pattern
777 2 \c at end of pattern
778 3 unrecognized character follows \
779 4 numbers out of order in {} quantifier
780 5 number too big in {} quantifier
781 6 missing terminating ] for character class
782 7 invalid escape sequence in character class
783 8 range out of order in character class
784 9 nothing to repeat
785 10 [this code is not in use]
786 11 internal error: unexpected repeat
787 12 unrecognized character after (? or (?-
788 13 POSIX named classes are supported only within a class
789 14 missing )
790 15 reference to non-existent subpattern
791 16 erroffset passed as NULL
792 17 unknown option bit(s) set
793 18 missing ) after comment
794 19 [this code is not in use]
795 20 regular expression is too large
796 21 failed to get memory
797 22 unmatched parentheses
798 23 internal error: code overflow
799 24 unrecognized character after (?&#60;
800 25 lookbehind assertion is not fixed length
801 26 malformed number or name after (?(
802 27 conditional group contains more than two branches
803 28 assertion expected after (?(
804 29 (?R or (?[+-]digits must be followed by )
805 30 unknown POSIX class name
806 31 POSIX collating elements are not supported
807 32 this version of PCRE is not compiled with PCRE_UTF8 support
808 33 [this code is not in use]
809 34 character value in \x{...} sequence is too large
810 35 invalid condition (?(0)
811 36 \C not allowed in lookbehind assertion
812 37 PCRE does not support \L, \l, \N{name}, \U, or \u
813 38 number after (?C is &#62; 255
814 39 closing ) for (?C expected
815 40 recursive call could loop indefinitely
816 41 unrecognized character after (?P
817 42 syntax error in subpattern name (missing terminator)
818 43 two named subpatterns have the same name
819 44 invalid UTF-8 string
820 45 support for \P, \p, and \X has not been compiled
821 46 malformed \P or \p sequence
822 47 unknown property name after \P or \p
823 48 subpattern name is too long (maximum 32 characters)
824 49 too many named subpatterns (maximum 10000)
825 50 [this code is not in use]
826 51 octal value is greater than \377 (not in UTF-8 mode)
827 52 internal error: overran compiling workspace
828 53 internal error: previously-checked referenced subpattern
829 not found
830 54 DEFINE group contains more than one branch
831 55 repeating a DEFINE group is not allowed
832 56 inconsistent NEWLINE options
833 57 \g is not followed by a braced, angle-bracketed, or quoted
834 name/number or by a plain number
835 58 a numbered reference must not be zero
836 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
837 60 (*VERB) not recognized
838 61 number is too big
839 62 subpattern name expected
840 63 digit expected after (?+
841 64 ] is an invalid data character in JavaScript compatibility mode
842 65 different names for subpatterns of the same number are
843 not allowed
844 66 (*MARK) must have an argument
845 67 this version of PCRE is not compiled with PCRE_UCP support
846 68 \c must be followed by an ASCII character
847 69 \k is not followed by a braced, angle-bracketed, or quoted name
848 </pre>
849 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
850 be used if the limits were changed when PCRE was built.
851 <a name="studyingapattern"></a></P>
852 <br><a name="SEC11" href="#TOC1">STUDYING A PATTERN</a><br>
853 <P>
854 <b>pcre_extra *pcre_study(const pcre *<i>code</i>, int <i>options</i></b>
855 <b>const char **<i>errptr</i>);</b>
856 </P>
857 <P>
858 If a compiled pattern is going to be used several times, it is worth spending
859 more time analyzing it in order to speed up the time taken for matching. The
860 function <b>pcre_study()</b> takes a pointer to a compiled pattern as its first
861 argument. If studying the pattern produces additional information that will
862 help speed up matching, <b>pcre_study()</b> returns a pointer to a
863 <b>pcre_extra</b> block, in which the <i>study_data</i> field points to the
864 results of the study.
865 </P>
866 <P>
867 The returned value from <b>pcre_study()</b> can be passed directly to
868 <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b>. However, a <b>pcre_extra</b> block
869 also contains other fields that can be set by the caller before the block is
870 passed; these are described
871 <a href="#extradata">below</a>
872 in the section on matching a pattern.
873 </P>
874 <P>
875 If studying the pattern does not produce any useful information,
876 <b>pcre_study()</b> returns NULL. In that circumstance, if the calling program
877 wants to pass any of the other fields to <b>pcre_exec()</b> or
878 <b>pcre_dfa_exec()</b>, it must set up its own <b>pcre_extra</b> block.
879 </P>
880 <P>
881 The second argument of <b>pcre_study()</b> contains option bits. There is only
882 one option: PCRE_STUDY_JIT_COMPILE. If this is set, and the just-in-time
883 compiler is available, the pattern is further compiled into machine code that
884 executes much faster than the <b>pcre_exec()</b> matching function. If
885 the just-in-time compiler is not available, this option is ignored. All other
886 bits in the <i>options</i> argument must be zero.
887 </P>
888 <P>
889 JIT compilation is a heavyweight optimization. It can take some time for
890 patterns to be analyzed, and for one-off matches and simple patterns the
891 benefit of faster execution might be offset by a much slower study time.
892 Not all patterns can be optimized by the JIT compiler. For those that cannot be
893 handled, matching automatically falls back to the <b>pcre_exec()</b>
894 interpreter. For more details, see the
895 <a href="pcrejit.html"><b>pcrejit</b></a>
896 documentation.
897 </P>
898 <P>
899 The third argument for <b>pcre_study()</b> is a pointer for an error message. If
900 studying succeeds (even if no data is returned), the variable it points to is
901 set to NULL. Otherwise it is set to point to a textual error message. This is a
902 static string that is part of the library. You must not try to free it. You
903 should test the error pointer for NULL after calling <b>pcre_study()</b>, to be
904 sure that it has run successfully.
905 </P>
906 <P>
907 When you are finished with a pattern, you can free the memory used for the
908 study data by calling <b>pcre_free_study()</b>. This function was added to the
909 API for release 8.20. For earlier versions, the memory could be freed with
910 <b>pcre_free()</b>, just like the pattern itself. This will still work in cases
911 where PCRE_STUDY_JIT_COMPILE is not used, but it is advisable to change to the
912 new function when convenient.
913 </P>
914 <P>
915 This is a typical way in which <b>pcre_study</b>() is used (except that in a
916 real application there should be tests for errors):
917 <pre>
918 int rc;
919 pcre *re;
920 pcre_extra *sd;
921 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
922 sd = pcre_study(
923 re, /* result of pcre_compile() */
924 0, /* no options */
925 &error); /* set to NULL or points to a message */
926 rc = pcre_exec( /* see below for details of pcre_exec() options */
927 re, sd, "subject", 7, 0, 0, ovector, 30);
928 ...
929 pcre_free_study(sd);
930 pcre_free(re);
931 </pre>
932 Studying a pattern does two things: first, a lower bound for the length of
933 subject string that is needed to match the pattern is computed. This does not
934 mean that there are any strings of that length that match, but it does
935 guarantee that no shorter strings match. The value is used by
936 <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b> to avoid wasting time by trying to
937 match strings that are shorter than the lower bound. You can find out the value
938 in a calling program via the <b>pcre_fullinfo()</b> function.
939 </P>
940 <P>
941 Studying a pattern is also useful for non-anchored patterns that do not have a
942 single fixed starting character. A bitmap of possible starting bytes is
943 created. This speeds up finding a position in the subject at which to start
944 matching.
945 </P>
946 <P>
947 These two optimizations apply to both <b>pcre_exec()</b> and
948 <b>pcre_dfa_exec()</b>. However, they are not used by <b>pcre_exec()</b> if
949 <b>pcre_study()</b> is called with the PCRE_STUDY_JIT_COMPILE option, and
950 just-in-time compiling is successful. The optimizations can be disabled by
951 setting the PCRE_NO_START_OPTIMIZE option when calling <b>pcre_exec()</b> or
952 <b>pcre_dfa_exec()</b>. You might want to do this if your pattern contains
953 callouts or (*MARK) (which cannot be handled by the JIT compiler), and you want
954 to make use of these facilities in cases where matching fails. See the
955 discussion of PCRE_NO_START_OPTIMIZE
956 <a href="#execoptions">below.</a>
957 <a name="localesupport"></a></P>
958 <br><a name="SEC12" href="#TOC1">LOCALE SUPPORT</a><br>
959 <P>
960 PCRE handles caseless matching, and determines whether characters are letters,
961 digits, or whatever, by reference to a set of tables, indexed by character
962 value. When running in UTF-8 mode, this applies only to characters with codes
963 less than 128. By default, higher-valued codes never match escapes such as \w
964 or \d, but they can be tested with \p if PCRE is built with Unicode character
965 property support. Alternatively, the PCRE_UCP option can be set at compile
966 time; this causes \w and friends to use Unicode property support instead of
967 built-in tables. The use of locales with Unicode is discouraged. If you are
968 handling characters with codes greater than 128, you should either use UTF-8
969 and Unicode, or use locales, but not try to mix the two.
970 </P>
971 <P>
972 PCRE contains an internal set of tables that are used when the final argument
973 of <b>pcre_compile()</b> is NULL. These are sufficient for many applications.
974 Normally, the internal tables recognize only ASCII characters. However, when
975 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
976 default "C" locale of the local system, which may cause them to be different.
977 </P>
978 <P>
979 The internal tables can always be overridden by tables supplied by the
980 application that calls PCRE. These may be created in a different locale from
981 the default. As more and more applications change to using Unicode, the need
982 for this locale support is expected to die away.
983 </P>
984 <P>
985 External tables are built by calling the <b>pcre_maketables()</b> function,
986 which has no arguments, in the relevant locale. The result can then be passed
987 to <b>pcre_compile()</b> or <b>pcre_exec()</b> as often as necessary. For
988 example, to build and use tables that are appropriate for the French locale
989 (where accented characters with values greater than 128 are treated as letters),
990 the following code could be used:
991 <pre>
992 setlocale(LC_CTYPE, "fr_FR");
993 tables = pcre_maketables();
994 re = pcre_compile(..., tables);
995 </pre>
996 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
997 are using Windows, the name for the French locale is "french".
998 </P>
999 <P>
1000 When <b>pcre_maketables()</b> runs, the tables are built in memory that is
1001 obtained via <b>pcre_malloc</b>. It is the caller's responsibility to ensure
1002 that the memory containing the tables remains available for as long as it is
1003 needed.
1004 </P>
1005 <P>
1006 The pointer that is passed to <b>pcre_compile()</b> is saved with the compiled
1007 pattern, and the same tables are used via this pointer by <b>pcre_study()</b>
1008 and normally also by <b>pcre_exec()</b>. Thus, by default, for any single
1009 pattern, compilation, studying and matching all happen in the same locale, but
1010 different patterns can be compiled in different locales.
1011 </P>
1012 <P>
1013 It is possible to pass a table pointer or NULL (indicating the use of the
1014 internal tables) to <b>pcre_exec()</b>. Although not intended for this purpose,
1015 this facility could be used to match a pattern in a different locale from the
1016 one in which it was compiled. Passing table pointers at run time is discussed
1017 below in the section on matching a pattern.
1018 <a name="infoaboutpattern"></a></P>
1019 <br><a name="SEC13" href="#TOC1">INFORMATION ABOUT A PATTERN</a><br>
1020 <P>
1021 <b>int pcre_fullinfo(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
1022 <b>int <i>what</i>, void *<i>where</i>);</b>
1023 </P>
1024 <P>
1025 The <b>pcre_fullinfo()</b> function returns information about a compiled
1026 pattern. It replaces the obsolete <b>pcre_info()</b> function, which is
1027 nevertheless retained for backwards compability (and is documented below).
1028 </P>
1029 <P>
1030 The first argument for <b>pcre_fullinfo()</b> is a pointer to the compiled
1031 pattern. The second argument is the result of <b>pcre_study()</b>, or NULL if
1032 the pattern was not studied. The third argument specifies which piece of
1033 information is required, and the fourth argument is a pointer to a variable
1034 to receive the data. The yield of the function is zero for success, or one of
1035 the following negative numbers:
1036 <pre>
1037 PCRE_ERROR_NULL the argument <i>code</i> was NULL
1038 the argument <i>where</i> was NULL
1039 PCRE_ERROR_BADMAGIC the "magic number" was not found
1040 PCRE_ERROR_BADOPTION the value of <i>what</i> was invalid
1041 </pre>
1042 The "magic number" is placed at the start of each compiled pattern as an simple
1043 check against passing an arbitrary memory pointer. Here is a typical call of
1044 <b>pcre_fullinfo()</b>, to obtain the length of the compiled pattern:
1045 <pre>
1046 int rc;
1047 size_t length;
1048 rc = pcre_fullinfo(
1049 re, /* result of pcre_compile() */
1050 sd, /* result of pcre_study(), or NULL */
1051 PCRE_INFO_SIZE, /* what is required */
1052 &length); /* where to put the data */
1053 </pre>
1054 The possible values for the third argument are defined in <b>pcre.h</b>, and are
1055 as follows:
1056 <pre>
1058 </pre>
1059 Return the number of the highest back reference in the pattern. The fourth
1060 argument should point to an <b>int</b> variable. Zero is returned if there are
1061 no back references.
1062 <pre>
1064 </pre>
1065 Return the number of capturing subpatterns in the pattern. The fourth argument
1066 should point to an <b>int</b> variable.
1067 <pre>
1069 </pre>
1070 Return a pointer to the internal default character tables within PCRE. The
1071 fourth argument should point to an <b>unsigned char *</b> variable. This
1072 information call is provided for internal use by the <b>pcre_study()</b>
1073 function. External callers can cause PCRE to use its internal tables by passing
1074 a NULL table pointer.
1075 <pre>
1077 </pre>
1078 Return information about the first byte of any matched string, for a
1079 non-anchored pattern. The fourth argument should point to an <b>int</b>
1080 variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
1081 still recognized for backwards compatibility.)
1082 </P>
1083 <P>
1084 If there is a fixed first byte, for example, from a pattern such as
1085 (cat|cow|coyote), its value is returned. Otherwise, if either
1086 <br>
1087 <br>
1088 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1089 starts with "^", or
1090 <br>
1091 <br>
1092 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1093 (if it were set, the pattern would be anchored),
1094 <br>
1095 <br>
1096 -1 is returned, indicating that the pattern matches only at the start of a
1097 subject string or after any newline within the string. Otherwise -2 is
1098 returned. For anchored patterns, -2 is returned.
1099 <pre>
1101 </pre>
1102 If the pattern was studied, and this resulted in the construction of a 256-bit
1103 table indicating a fixed set of bytes for the first byte in any matching
1104 string, a pointer to the table is returned. Otherwise NULL is returned. The
1105 fourth argument should point to an <b>unsigned char *</b> variable.
1106 <pre>
1108 </pre>
1109 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1110 otherwise 0. The fourth argument should point to an <b>int</b> variable. An
1111 explicit match is either a literal CR or LF character, or \r or \n.
1112 <pre>
1114 </pre>
1115 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1116 0. The fourth argument should point to an <b>int</b> variable. (?J) and
1117 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1118 <pre>
1120 </pre>
1121 Return 1 if the pattern was studied with the PCRE_STUDY_JIT_COMPILE option, and
1122 just-in-time compiling was successful. The fourth argument should point to an
1123 <b>int</b> variable. A return value of 0 means that JIT support is not available
1124 in this version of PCRE, or that the pattern was not studied with the
1125 PCRE_STUDY_JIT_COMPILE option, or that the JIT compiler could not handle this
1126 particular pattern. See the
1127 <a href="pcrejit.html"><b>pcrejit</b></a>
1128 documentation for details of what can and cannot be handled.
1129 <pre>
1131 </pre>
1132 Return the value of the rightmost literal byte that must exist in any matched
1133 string, other than at its start, if such a byte has been recorded. The fourth
1134 argument should point to an <b>int</b> variable. If there is no such byte, -1 is
1135 returned. For anchored patterns, a last literal byte is recorded only if it
1136 follows something of variable length. For example, for the pattern
1137 /^a\d+z\d+/ the returned value is "z", but for /^a\dz\d/ the returned value
1138 is -1.
1139 <pre>
1141 </pre>
1142 If the pattern was studied and a minimum length for matching subject strings
1143 was computed, its value is returned. Otherwise the returned value is -1. The
1144 value is a number of characters, not bytes (this may be relevant in UTF-8
1145 mode). The fourth argument should point to an <b>int</b> variable. A
1146 non-negative value is a lower bound to the length of any matching string. There
1147 may not be any strings of that length that do actually match, but every string
1148 that does match is at least that long.
1149 <pre>
1153 </pre>
1154 PCRE supports the use of named as well as numbered capturing parentheses. The
1155 names are just an additional way of identifying the parentheses, which still
1156 acquire numbers. Several convenience functions such as
1157 <b>pcre_get_named_substring()</b> are provided for extracting captured
1158 substrings by name. It is also possible to extract the data directly, by first
1159 converting the name to a number in order to access the correct pointers in the
1160 output vector (described with <b>pcre_exec()</b> below). To do the conversion,
1161 you need to use the name-to-number map, which is described by these three
1162 values.
1163 </P>
1164 <P>
1165 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1166 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1167 entry; both of these return an <b>int</b> value. The entry size depends on the
1168 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1169 entry of the table (a pointer to <b>char</b>). The first two bytes of each entry
1170 are the number of the capturing parenthesis, most significant byte first. The
1171 rest of the entry is the corresponding name, zero terminated.
1172 </P>
1173 <P>
1174 The names are in alphabetical order. Duplicate names may appear if (?| is used
1175 to create multiple groups with the same number, as described in the
1176 <a href="pcrepattern.html#dupsubpatternnumber">section on duplicate subpattern numbers</a>
1177 in the
1178 <a href="pcrepattern.html"><b>pcrepattern</b></a>
1179 page. Duplicate names for subpatterns with different numbers are permitted only
1180 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1181 table in the order in which they were found in the pattern. In the absence of
1182 (?| this is the order of increasing number; when (?| is used this is not
1183 necessarily the case because later subpatterns may have lower numbers.
1184 </P>
1185 <P>
1186 As a simple example of the name/number table, consider the following pattern
1187 (assume PCRE_EXTENDED is set, so white space - including newlines - is
1188 ignored):
1189 <pre>
1190 (?&#60;date&#62; (?&#60;year&#62;(\d\d)?\d\d) - (?&#60;month&#62;\d\d) - (?&#60;day&#62;\d\d) )
1191 </pre>
1192 There are four named subpatterns, so the table has four entries, and each entry
1193 in the table is eight bytes long. The table is as follows, with non-printing
1194 bytes shows in hexadecimal, and undefined bytes shown as ??:
1195 <pre>
1196 00 01 d a t e 00 ??
1197 00 05 d a y 00 ?? ??
1198 00 04 m o n t h 00
1199 00 02 y e a r 00 ??
1200 </pre>
1201 When writing code to extract data from named subpatterns using the
1202 name-to-number map, remember that the length of the entries is likely to be
1203 different for each compiled pattern.
1204 <pre>
1206 </pre>
1207 Return 1 if the pattern can be used for partial matching with
1208 <b>pcre_exec()</b>, otherwise 0. The fourth argument should point to an
1209 <b>int</b> variable. From release 8.00, this always returns 1, because the
1210 restrictions that previously applied to partial matching have been lifted. The
1211 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1212 documentation gives details of partial matching.
1213 <pre>
1215 </pre>
1216 Return a copy of the options with which the pattern was compiled. The fourth
1217 argument should point to an <b>unsigned long int</b> variable. These option bits
1218 are those specified in the call to <b>pcre_compile()</b>, modified by any
1219 top-level option settings at the start of the pattern itself. In other words,
1220 they are the options that will be in force when matching starts. For example,
1221 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1223 </P>
1224 <P>
1225 A pattern is automatically anchored by PCRE if all of its top-level
1226 alternatives begin with one of the following:
1227 <pre>
1228 ^ unless PCRE_MULTILINE is set
1229 \A always
1230 \G always
1231 .* if PCRE_DOTALL is set and there are no back references to the subpattern in which .* appears
1232 </pre>
1233 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1234 <b>pcre_fullinfo()</b>.
1235 <pre>
1237 </pre>
1238 Return the size of the compiled pattern, that is, the value that was passed as
1239 the argument to <b>pcre_malloc()</b> when PCRE was getting memory in which to
1240 place the compiled data. The fourth argument should point to a <b>size_t</b>
1241 variable.
1242 <pre>
1244 </pre>
1245 Return the size of the data block pointed to by the <i>study_data</i> field in a
1246 <b>pcre_extra</b> block. If <b>pcre_extra</b> is NULL, or there is no study data,
1247 zero is returned. The fourth argument should point to a <b>size_t</b> variable.
1248 The <i>study_data</i> field is set by <b>pcre_study()</b> to record information
1249 that will speed up matching (see the section entitled
1250 <a href="#studyingapattern">"Studying a pattern"</a>
1251 above). The format of the <i>study_data</i> block is private, but its length
1252 is made available via this option so that it can be saved and restored (see the
1253 <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1254 documentation for details).
1255 </P>
1256 <br><a name="SEC14" href="#TOC1">OBSOLETE INFO FUNCTION</a><br>
1257 <P>
1258 <b>int pcre_info(const pcre *<i>code</i>, int *<i>optptr</i>, int</b>
1259 <b>*<i>firstcharptr</i>);</b>
1260 </P>
1261 <P>
1262 The <b>pcre_info()</b> function is now obsolete because its interface is too
1263 restrictive to return all the available data about a compiled pattern. New
1264 programs should use <b>pcre_fullinfo()</b> instead. The yield of
1265 <b>pcre_info()</b> is the number of capturing subpatterns, or one of the
1266 following negative numbers:
1267 <pre>
1268 PCRE_ERROR_NULL the argument <i>code</i> was NULL
1269 PCRE_ERROR_BADMAGIC the "magic number" was not found
1270 </pre>
1271 If the <i>optptr</i> argument is not NULL, a copy of the options with which the
1272 pattern was compiled is placed in the integer it points to (see
1273 PCRE_INFO_OPTIONS above).
1274 </P>
1275 <P>
1276 If the pattern is not anchored and the <i>firstcharptr</i> argument is not NULL,
1277 it is used to pass back information about the first character of any matched
1278 string (see PCRE_INFO_FIRSTBYTE above).
1279 </P>
1280 <br><a name="SEC15" href="#TOC1">REFERENCE COUNTS</a><br>
1281 <P>
1282 <b>int pcre_refcount(pcre *<i>code</i>, int <i>adjust</i>);</b>
1283 </P>
1284 <P>
1285 The <b>pcre_refcount()</b> function is used to maintain a reference count in the
1286 data block that contains a compiled pattern. It is provided for the benefit of
1287 applications that operate in an object-oriented manner, where different parts
1288 of the application may be using the same compiled pattern, but you want to free
1289 the block when they are all done.
1290 </P>
1291 <P>
1292 When a pattern is compiled, the reference count field is initialized to zero.
1293 It is changed only by calling this function, whose action is to add the
1294 <i>adjust</i> value (which may be positive or negative) to it. The yield of the
1295 function is the new value. However, the value of the count is constrained to
1296 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1297 it is forced to the appropriate limit value.
1298 </P>
1299 <P>
1300 Except when it is zero, the reference count is not correctly preserved if a
1301 pattern is compiled on one host and then transferred to a host whose byte-order
1302 is different. (This seems a highly unlikely scenario.)
1303 </P>
1304 <br><a name="SEC16" href="#TOC1">MATCHING A PATTERN: THE TRADITIONAL FUNCTION</a><br>
1305 <P>
1306 <b>int pcre_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
1307 <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
1308 <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>);</b>
1309 </P>
1310 <P>
1311 The function <b>pcre_exec()</b> is called to match a subject string against a
1312 compiled pattern, which is passed in the <i>code</i> argument. If the
1313 pattern was studied, the result of the study should be passed in the
1314 <i>extra</i> argument. This function is the main matching facility of the
1315 library, and it operates in a Perl-like manner. For specialist use there is
1316 also an alternative matching function, which is described
1317 <a href="#dfamatch">below</a>
1318 in the section about the <b>pcre_dfa_exec()</b> function.
1319 </P>
1320 <P>
1321 In most applications, the pattern will have been compiled (and optionally
1322 studied) in the same process that calls <b>pcre_exec()</b>. However, it is
1323 possible to save compiled patterns and study data, and then use them later
1324 in different processes, possibly even on different hosts. For a discussion
1325 about this, see the
1326 <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1327 documentation.
1328 </P>
1329 <P>
1330 Here is an example of a simple call to <b>pcre_exec()</b>:
1331 <pre>
1332 int rc;
1333 int ovector[30];
1334 rc = pcre_exec(
1335 re, /* result of pcre_compile() */
1336 NULL, /* we didn't study the pattern */
1337 "some string", /* the subject string */
1338 11, /* the length of the subject string */
1339 0, /* start at offset 0 in the subject */
1340 0, /* default options */
1341 ovector, /* vector of integers for substring information */
1342 30); /* number of elements (NOT size in bytes) */
1343 <a name="extradata"></a></PRE>
1344 </P>
1345 <br><b>
1346 Extra data for <b>pcre_exec()</b>
1347 </b><br>
1348 <P>
1349 If the <i>extra</i> argument is not NULL, it must point to a <b>pcre_extra</b>
1350 data block. The <b>pcre_study()</b> function returns such a block (when it
1351 doesn't return NULL), but you can also create one for yourself, and pass
1352 additional information in it. The <b>pcre_extra</b> block contains the following
1353 fields (not necessarily in this order):
1354 <pre>
1355 unsigned long int <i>flags</i>;
1356 void *<i>study_data</i>;
1357 void *<i>executable_jit</i>;
1358 unsigned long int <i>match_limit</i>;
1359 unsigned long int <i>match_limit_recursion</i>;
1360 void *<i>callout_data</i>;
1361 const unsigned char *<i>tables</i>;
1362 unsigned char **<i>mark</i>;
1363 </pre>
1364 The <i>flags</i> field is a bitmap that specifies which of the other fields
1365 are set. The flag bits are:
1366 <pre>
1374 </pre>
1375 Other flag bits should be set to zero. The <i>study_data</i> field and sometimes
1376 the <i>executable_jit</i> field are set in the <b>pcre_extra</b> block that is
1377 returned by <b>pcre_study()</b>, together with the appropriate flag bits. You
1378 should not set these yourself, but you may add to the block by setting the
1379 other fields and their corresponding flag bits.
1380 </P>
1381 <P>
1382 The <i>match_limit</i> field provides a means of preventing PCRE from using up a
1383 vast amount of resources when running patterns that are not going to match,
1384 but which have a very large number of possibilities in their search trees. The
1385 classic example is a pattern that uses nested unlimited repeats.
1386 </P>
1387 <P>
1388 Internally, <b>pcre_exec()</b> uses a function called <b>match()</b>, which it
1389 calls repeatedly (sometimes recursively). The limit set by <i>match_limit</i> is
1390 imposed on the number of times this function is called during a match, which
1391 has the effect of limiting the amount of backtracking that can take place. For
1392 patterns that are not anchored, the count restarts from zero for each position
1393 in the subject string.
1394 </P>
1395 <P>
1396 When <b>pcre_exec()</b> is called with a pattern that was successfully studied
1397 with the PCRE_STUDY_JIT_COMPILE option, the way that the matching is executed
1398 is entirely different. However, there is still the possibility of runaway
1399 matching that goes on for a very long time, and so the <i>match_limit</i> value
1400 is also used in this case (but in a different way) to limit how long the
1401 matching can continue.
1402 </P>
1403 <P>
1404 The default value for the limit can be set when PCRE is built; the default
1405 default is 10 million, which handles all but the most extreme cases. You can
1406 override the default by suppling <b>pcre_exec()</b> with a <b>pcre_extra</b>
1407 block in which <i>match_limit</i> is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1408 the <i>flags</i> field. If the limit is exceeded, <b>pcre_exec()</b> returns
1410 </P>
1411 <P>
1412 The <i>match_limit_recursion</i> field is similar to <i>match_limit</i>, but
1413 instead of limiting the total number of times that <b>match()</b> is called, it
1414 limits the depth of recursion. The recursion depth is a smaller number than the
1415 total number of calls, because not all calls to <b>match()</b> are recursive.
1416 This limit is of use only if it is set smaller than <i>match_limit</i>.
1417 </P>
1418 <P>
1419 Limiting the recursion depth limits the amount of machine stack that can be
1420 used, or, when PCRE has been compiled to use memory on the heap instead of the
1421 stack, the amount of heap memory that can be used. This limit is not relevant,
1422 and is ignored, if the pattern was successfully studied with
1424 </P>
1425 <P>
1426 The default value for <i>match_limit_recursion</i> can be set when PCRE is
1427 built; the default default is the same value as the default for
1428 <i>match_limit</i>. You can override the default by suppling <b>pcre_exec()</b>
1429 with a <b>pcre_extra</b> block in which <i>match_limit_recursion</i> is set, and
1430 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the <i>flags</i> field. If the limit
1431 is exceeded, <b>pcre_exec()</b> returns PCRE_ERROR_RECURSIONLIMIT.
1432 </P>
1433 <P>
1434 The <i>callout_data</i> field is used in conjunction with the "callout" feature,
1435 and is described in the
1436 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1437 documentation.
1438 </P>
1439 <P>
1440 The <i>tables</i> field is used to pass a character tables pointer to
1441 <b>pcre_exec()</b>; this overrides the value that is stored with the compiled
1442 pattern. A non-NULL value is stored with the compiled pattern only if custom
1443 tables were supplied to <b>pcre_compile()</b> via its <i>tableptr</i> argument.
1444 If NULL is passed to <b>pcre_exec()</b> using this mechanism, it forces PCRE's
1445 internal tables to be used. This facility is helpful when re-using patterns
1446 that have been saved after compiling with an external set of tables, because
1447 the external tables might be at a different address when <b>pcre_exec()</b> is
1448 called. See the
1449 <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1450 documentation for a discussion of saving compiled patterns for later use.
1451 </P>
1452 <P>
1453 If PCRE_EXTRA_MARK is set in the <i>flags</i> field, the <i>mark</i> field must
1454 be set to point to a <b>char *</b> variable. If the pattern contains any
1455 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1456 a name to pass back, a pointer to the name string (zero terminated) is placed
1457 in the variable pointed to by the <i>mark</i> field. The names are within the
1458 compiled pattern; if you wish to retain such a name you must copy it before
1459 freeing the memory of a compiled pattern. If there is no name to pass back, the
1460 variable pointed to by the <i>mark</i> field set to NULL. For details of the
1461 backtracking control verbs, see the section entitled
1462 <a href="pcrepattern#backtrackcontrol">"Backtracking control"</a>
1463 in the
1464 <a href="pcrepattern.html"><b>pcrepattern</b></a>
1465 documentation.
1466 <a name="execoptions"></a></P>
1467 <br><b>
1468 Option bits for <b>pcre_exec()</b>
1469 </b><br>
1470 <P>
1471 The unused bits of the <i>options</i> argument for <b>pcre_exec()</b> must be
1472 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
1476 </P>
1477 <P>
1478 If the pattern was successfully studied with the PCRE_STUDY_JIT_COMPILE option,
1479 the only supported options for JIT execution are PCRE_NO_UTF8_CHECK,
1481 particular that partial matching is not supported. If an unsupported option is
1482 used, JIT execution is disabled and the normal interpretive code in
1483 <b>pcre_exec()</b> is run.
1484 <pre>
1486 </pre>
1487 The PCRE_ANCHORED option limits <b>pcre_exec()</b> to matching at the first
1488 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1489 to be anchored by virtue of its contents, it cannot be made unachored at
1490 matching time.
1491 <pre>
1494 </pre>
1495 These options (which are mutually exclusive) control what the \R escape
1496 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1497 match any Unicode newline sequence. These options override the choice that was
1498 made or defaulted when the pattern was compiled.
1499 <pre>
1505 </pre>
1506 These options override the newline definition that was chosen or defaulted when
1507 the pattern was compiled. For details, see the description of
1508 <b>pcre_compile()</b> above. During matching, the newline choice affects the
1509 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1510 the way the match position is advanced after a match failure for an unanchored
1511 pattern.
1512 </P>
1513 <P>
1515 match attempt for an unanchored pattern fails when the current position is at a
1516 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1517 characters, the match position is advanced by two characters instead of one, in
1518 other words, to after the CRLF.
1519 </P>
1520 <P>
1521 The above rule is a compromise that makes the most common cases work as
1522 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1523 set), it does not match the string "\r\nA" because, after failing at the
1524 start, it skips both the CR and the LF before retrying. However, the pattern
1525 [\r\n]A does match that string, because it contains an explicit CR or LF
1526 reference, and so advances only by one character after the first failure.
1527 </P>
1528 <P>
1529 An explicit match for CR of LF is either a literal appearance of one of those
1530 characters, or one of the \r or \n escape sequences. Implicit matches such as
1531 [^X] do not count, nor does \s (which includes CR and LF in the characters
1532 that it matches).
1533 </P>
1534 <P>
1535 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1536 valid newline sequence and explicit \r or \n escapes appear in the pattern.
1537 <pre>
1539 </pre>
1540 This option specifies that first character of the subject string is not the
1541 beginning of a line, so the circumflex metacharacter should not match before
1542 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1543 never to match. This option affects only the behaviour of the circumflex
1544 metacharacter. It does not affect \A.
1545 <pre>
1547 </pre>
1548 This option specifies that the end of the subject string is not the end of a
1549 line, so the dollar metacharacter should not match it nor (except in multiline
1550 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1551 compile time) causes dollar never to match. This option affects only the
1552 behaviour of the dollar metacharacter. It does not affect \Z or \z.
1553 <pre>
1555 </pre>
1556 An empty string is not considered to be a valid match if this option is set. If
1557 there are alternatives in the pattern, they are tried. If all the alternatives
1558 match the empty string, the entire match fails. For example, if the pattern
1559 <pre>
1560 a?b?
1561 </pre>
1562 is applied to a string not beginning with "a" or "b", it matches an empty
1563 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1564 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1565 <pre>
1567 </pre>
1568 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1569 the start of the subject is permitted. If the pattern is anchored, such a match
1570 can occur only if the pattern contains \K.
1571 </P>
1572 <P>
1573 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1574 does make a special case of a pattern match of the empty string within its
1575 <b>split()</b> function, and when using the /g modifier. It is possible to
1576 emulate Perl's behaviour after matching a null string by first trying the match
1577 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1578 if that fails, by advancing the starting offset (see below) and trying an
1579 ordinary match again. There is some code that demonstrates how to do this in
1580 the
1581 <a href="pcredemo.html"><b>pcredemo</b></a>
1582 sample program. In the most general case, you have to check to see if the
1583 newline convention recognizes CRLF as a newline, and if so, and the current
1584 character is CR followed by LF, advance the starting offset by two characters
1585 instead of one.
1586 <pre>
1588 </pre>
1589 There are a number of optimizations that <b>pcre_exec()</b> uses at the start of
1590 a match, in order to speed up the process. For example, if it is known that an
1591 unanchored match must start with a specific character, it searches the subject
1592 for that character, and fails immediately if it cannot find it, without
1593 actually running the main matching function. This means that a special item
1594 such as (*COMMIT) at the start of a pattern is not considered until after a
1595 suitable starting point for the match has been found. When callouts or (*MARK)
1596 items are in use, these "start-up" optimizations can cause them to be skipped
1597 if the pattern is never actually used. The start-up optimizations are in effect
1598 a pre-scan of the subject that takes place before the pattern is run.
1599 </P>
1600 <P>
1601 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1602 causing performance to suffer, but ensuring that in cases where the result is
1603 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1604 are considered at every possible starting position in the subject string. If
1605 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1606 time.
1607 </P>
1608 <P>
1609 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1610 Consider the pattern
1611 <pre>
1613 </pre>
1614 When this is compiled, PCRE records the fact that a match must start with the
1615 character "A". Suppose the subject string is "DEFABC". The start-up
1616 optimization scans along the subject, finds "A" and runs the first match
1617 attempt from there. The (*COMMIT) item means that the pattern must match the
1618 current starting position, which in this case, it does. However, if the same
1619 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1620 subject string does not happen. The first match attempt is run starting from
1621 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1622 the overall result is "no match". If the pattern is studied, more start-up
1623 optimizations may be used. For example, a minimum length for the subject may be
1624 recorded. Consider the pattern
1625 <pre>
1626 (*MARK:A)(X|Y)
1627 </pre>
1628 The minimum length for a match is one character. If the subject is "ABC", there
1629 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1630 If the pattern is studied, the final attempt does not take place, because PCRE
1631 knows that the subject is too short, and so the (*MARK) is never encountered.
1632 In this case, studying the pattern does not affect the overall match result,
1633 which is still "no match", but it does affect the auxiliary information that is
1634 returned.
1635 <pre>
1637 </pre>
1638 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1639 string is automatically checked when <b>pcre_exec()</b> is subsequently called.
1640 The value of <i>startoffset</i> is also checked to ensure that it points to the
1641 start of a UTF-8 character. There is a discussion about the validity of UTF-8
1642 strings in the
1643 <a href="pcre.html#utf8strings">section on UTF-8 support</a>
1644 in the main
1645 <a href="pcre.html"><b>pcre</b></a>
1646 page. If an invalid UTF-8 sequence of bytes is found, <b>pcre_exec()</b> returns
1647 the error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is
1648 a truncated UTF-8 character at the end of the subject, PCRE_ERROR_SHORTUTF8. In
1649 both cases, information about the precise nature of the error may also be
1650 returned (see the descriptions of these errors in the section entitled \fIError
1651 return values from\fP <b>pcre_exec()</b>
1652 <a href="#errorlist">below).</a>
1653 If <i>startoffset</i> contains a value that does not point to the start of a
1654 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1655 returned.
1656 </P>
1657 <P>
1658 If you already know that your subject is valid, and you want to skip these
1659 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1660 calling <b>pcre_exec()</b>. You might want to do this for the second and
1661 subsequent calls to <b>pcre_exec()</b> if you are making repeated calls to find
1662 all the matches in a single subject string. However, you should be sure that
1663 the value of <i>startoffset</i> points to the start of a UTF-8 character (or the
1664 end of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1665 invalid UTF-8 string as a subject or an invalid value of <i>startoffset</i> is
1666 undefined. Your program may crash.
1667 <pre>
1670 </pre>
1671 These options turn on the partial matching feature. For backwards
1672 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1673 occurs if the end of the subject string is reached successfully, but there are
1674 not enough subject characters to complete the match. If this happens when
1675 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1676 testing any remaining alternatives. Only if no complete match can be found is
1677 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1678 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1679 but only if no complete match can be found.
1680 </P>
1681 <P>
1682 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1683 partial match is found, <b>pcre_exec()</b> immediately returns
1684 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1685 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1686 important that an alternative complete match.
1687 </P>
1688 <P>
1689 In both cases, the portion of the string that was inspected when the partial
1690 match was found is set as the first matching string. There is a more detailed
1691 discussion of partial and multi-segment matching, with examples, in the
1692 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1693 documentation.
1694 </P>
1695 <br><b>
1696 The string to be matched by <b>pcre_exec()</b>
1697 </b><br>
1698 <P>
1699 The subject string is passed to <b>pcre_exec()</b> as a pointer in
1700 <i>subject</i>, a length (in bytes) in <i>length</i>, and a starting byte offset
1701 in <i>startoffset</i>. If this is negative or greater than the length of the
1702 subject, <b>pcre_exec()</b> returns PCRE_ERROR_BADOFFSET. When the starting
1703 offset is zero, the search for a match starts at the beginning of the subject,
1704 and this is by far the most common case. In UTF-8 mode, the byte offset must
1705 point to the start of a UTF-8 character (or the end of the subject). Unlike the
1706 pattern string, the subject may contain binary zero bytes.
1707 </P>
1708 <P>
1709 A non-zero starting offset is useful when searching for another match in the
1710 same subject by calling <b>pcre_exec()</b> again after a previous success.
1711 Setting <i>startoffset</i> differs from just passing over a shortened string and
1712 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1713 lookbehind. For example, consider the pattern
1714 <pre>
1715 \Biss\B
1716 </pre>
1717 which finds occurrences of "iss" in the middle of words. (\B matches only if
1718 the current position in the subject is not a word boundary.) When applied to
1719 the string "Mississipi" the first call to <b>pcre_exec()</b> finds the first
1720 occurrence. If <b>pcre_exec()</b> is called again with just the remainder of the
1721 subject, namely "issipi", it does not match, because \B is always false at the
1722 start of the subject, which is deemed to be a word boundary. However, if
1723 <b>pcre_exec()</b> is passed the entire string again, but with <i>startoffset</i>
1724 set to 4, it finds the second occurrence of "iss" because it is able to look
1725 behind the starting point to discover that it is preceded by a letter.
1726 </P>
1727 <P>
1728 Finding all the matches in a subject is tricky when the pattern can match an
1729 empty string. It is possible to emulate Perl's /g behaviour by first trying the
1730 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1731 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1732 and trying an ordinary match again. There is some code that demonstrates how to
1733 do this in the
1734 <a href="pcredemo.html"><b>pcredemo</b></a>
1735 sample program. In the most general case, you have to check to see if the
1736 newline convention recognizes CRLF as a newline, and if so, and the current
1737 character is CR followed by LF, advance the starting offset by two characters
1738 instead of one.
1739 </P>
1740 <P>
1741 If a non-zero starting offset is passed when the pattern is anchored, one
1742 attempt to match at the given offset is made. This can only succeed if the
1743 pattern does not require the match to be at the start of the subject.
1744 </P>
1745 <br><b>
1746 How <b>pcre_exec()</b> returns captured substrings
1747 </b><br>
1748 <P>
1749 In general, a pattern matches a certain portion of the subject, and in
1750 addition, further substrings from the subject may be picked out by parts of the
1751 pattern. Following the usage in Jeffrey Friedl's book, this is called
1752 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1753 a fragment of a pattern that picks out a substring. PCRE supports several other
1754 kinds of parenthesized subpattern that do not cause substrings to be captured.
1755 </P>
1756 <P>
1757 Captured substrings are returned to the caller via a vector of integers whose
1758 address is passed in <i>ovector</i>. The number of elements in the vector is
1759 passed in <i>ovecsize</i>, which must be a non-negative number. <b>Note</b>: this
1760 argument is NOT the size of <i>ovector</i> in bytes.
1761 </P>
1762 <P>
1763 The first two-thirds of the vector is used to pass back captured substrings,
1764 each substring using a pair of integers. The remaining third of the vector is
1765 used as workspace by <b>pcre_exec()</b> while matching capturing subpatterns,
1766 and is not available for passing back information. The number passed in
1767 <i>ovecsize</i> should always be a multiple of three. If it is not, it is
1768 rounded down.
1769 </P>
1770 <P>
1771 When a match is successful, information about captured substrings is returned
1772 in pairs of integers, starting at the beginning of <i>ovector</i>, and
1773 continuing up to two-thirds of its length at the most. The first element of
1774 each pair is set to the byte offset of the first character in a substring, and
1775 the second is set to the byte offset of the first character after the end of a
1776 substring. <b>Note</b>: these values are always byte offsets, even in UTF-8
1777 mode. They are not character counts.
1778 </P>
1779 <P>
1780 The first pair of integers, <i>ovector[0]</i> and <i>ovector[1]</i>, identify the
1781 portion of the subject string matched by the entire pattern. The next pair is
1782 used for the first capturing subpattern, and so on. The value returned by
1783 <b>pcre_exec()</b> is one more than the highest numbered pair that has been set.
1784 For example, if two substrings have been captured, the returned value is 3. If
1785 there are no capturing subpatterns, the return value from a successful match is
1786 1, indicating that just the first pair of offsets has been set.
1787 </P>
1788 <P>
1789 If a capturing subpattern is matched repeatedly, it is the last portion of the
1790 string that it matched that is returned.
1791 </P>
1792 <P>
1793 If the vector is too small to hold all the captured substring offsets, it is
1794 used as far as possible (up to two-thirds of its length), and the function
1795 returns a value of zero. If neither the actual string matched not any captured
1796 substrings are of interest, <b>pcre_exec()</b> may be called with <i>ovector</i>
1797 passed as NULL and <i>ovecsize</i> as zero. However, if the pattern contains
1798 back references and the <i>ovector</i> is not big enough to remember the related
1799 substrings, PCRE has to get additional memory for use during matching. Thus it
1800 is usually advisable to supply an <i>ovector</i> of reasonable size.
1801 </P>
1802 <P>
1803 There are some cases where zero is returned (indicating vector overflow) when
1804 in fact the vector is exactly the right size for the final match. For example,
1805 consider the pattern
1806 <pre>
1807 (a)(?:(b)c|bd)
1808 </pre>
1809 If a vector of 6 elements (allowing for only 1 captured substring) is given
1810 with subject string "abd", <b>pcre_exec()</b> will try to set the second
1811 captured string, thereby recording a vector overflow, before failing to match
1812 "c" and backing up to try the second alternative. The zero return, however,
1813 does correctly indicate that the maximum number of slots (namely 2) have been
1814 filled. In similar cases where there is temporary overflow, but the final
1815 number of used slots is actually less than the maximum, a non-zero value is
1816 returned.
1817 </P>
1818 <P>
1819 The <b>pcre_fullinfo()</b> function can be used to find out how many capturing
1820 subpatterns there are in a compiled pattern. The smallest size for
1821 <i>ovector</i> that will allow for <i>n</i> captured substrings, in addition to
1822 the offsets of the substring matched by the whole pattern, is (<i>n</i>+1)*3.
1823 </P>
1824 <P>
1825 It is possible for capturing subpattern number <i>n+1</i> to match some part of
1826 the subject when subpattern <i>n</i> has not been used at all. For example, if
1827 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1828 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1829 happens, both values in the offset pairs corresponding to unused subpatterns
1830 are set to -1.
1831 </P>
1832 <P>
1833 Offset values that correspond to unused subpatterns at the end of the
1834 expression are also set to -1. For example, if the string "abc" is matched
1835 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1836 return from the function is 2, because the highest used capturing subpattern
1837 number is 1, and the offsets for for the second and third capturing subpatterns
1838 (assuming the vector is large enough, of course) are set to -1.
1839 </P>
1840 <P>
1841 <b>Note</b>: Elements in the first two-thirds of <i>ovector</i> that do not
1842 correspond to capturing parentheses in the pattern are never changed. That is,
1843 if a pattern contains <i>n</i> capturing parentheses, no more than
1844 <i>ovector[0]</i> to <i>ovector[2n+1]</i> are set by <b>pcre_exec()</b>. The other
1845 elements (in the first two-thirds) retain whatever values they previously had.
1846 </P>
1847 <P>
1848 Some convenience functions are provided for extracting the captured substrings
1849 as separate strings. These are described below.
1850 <a name="errorlist"></a></P>
1851 <br><b>
1852 Error return values from <b>pcre_exec()</b>
1853 </b><br>
1854 <P>
1855 If <b>pcre_exec()</b> fails, it returns a negative number. The following are
1856 defined in the header file:
1857 <pre>
1859 </pre>
1860 The subject string did not match the pattern.
1861 <pre>
1863 </pre>
1864 Either <i>code</i> or <i>subject</i> was passed as NULL, or <i>ovector</i> was
1865 NULL and <i>ovecsize</i> was not zero.
1866 <pre>
1868 </pre>
1869 An unrecognized bit was set in the <i>options</i> argument.
1870 <pre>
1872 </pre>
1873 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1874 the case when it is passed a junk pointer and to detect when a pattern that was
1875 compiled in an environment of one endianness is run in an environment with the
1876 other endianness. This is the error that PCRE gives when the magic number is
1877 not present.
1878 <pre>
1880 </pre>
1881 While running the pattern match, an unknown item was encountered in the
1882 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1883 of the compiled pattern.
1884 <pre>
1886 </pre>
1887 If a pattern contains back references, but the <i>ovector</i> that is passed to
1888 <b>pcre_exec()</b> is not big enough to remember the referenced substrings, PCRE
1889 gets a block of memory at the start of matching to use for this purpose. If the
1890 call via <b>pcre_malloc()</b> fails, this error is given. The memory is
1891 automatically freed at the end of matching.
1892 </P>
1893 <P>
1894 This error is also given if <b>pcre_stack_malloc()</b> fails in
1895 <b>pcre_exec()</b>. This can happen only when PCRE has been compiled with
1896 <b>--disable-stack-for-recursion</b>.
1897 <pre>
1899 </pre>
1900 This error is used by the <b>pcre_copy_substring()</b>,
1901 <b>pcre_get_substring()</b>, and <b>pcre_get_substring_list()</b> functions (see
1902 below). It is never returned by <b>pcre_exec()</b>.
1903 <pre>
1905 </pre>
1906 The backtracking limit, as specified by the <i>match_limit</i> field in a
1907 <b>pcre_extra</b> structure (or defaulted) was reached. See the description
1908 above.
1909 <pre>
1911 </pre>
1912 This error is never generated by <b>pcre_exec()</b> itself. It is provided for
1913 use by callout functions that want to yield a distinctive error code. See the
1914 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1915 documentation for details.
1916 <pre>
1918 </pre>
1919 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
1920 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
1921 (<i>ovecsize</i>) is at least 2, the byte offset to the start of the the invalid
1922 UTF-8 character is placed in the first element, and a reason code is placed in
1923 the second element. The reason codes are listed in the
1924 <a href="#badutf8reasons">following section.</a>
1925 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
1926 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
1927 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
1928 <pre>
1930 </pre>
1931 The UTF-8 byte sequence that was passed as a subject was checked and found to
1932 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
1933 <i>startoffset</i> did not point to the beginning of a UTF-8 character or the
1934 end of the subject.
1935 <pre>
1937 </pre>
1938 The subject string did not match, but it did match partially. See the
1939 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1940 documentation for details of partial matching.
1941 <pre>
1943 </pre>
1944 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1945 option was used with a compiled pattern containing items that were not
1946 supported for partial matching. From release 8.00 onwards, there are no
1947 restrictions on partial matching.
1948 <pre>
1950 </pre>
1951 An unexpected internal error has occurred. This error could be caused by a bug
1952 in PCRE or by overwriting of the compiled pattern.
1953 <pre>
1955 </pre>
1956 This error is given if the value of the <i>ovecsize</i> argument is negative.
1957 <pre>
1959 </pre>
1960 The internal recursion limit, as specified by the <i>match_limit_recursion</i>
1961 field in a <b>pcre_extra</b> structure (or defaulted) was reached. See the
1962 description above.
1963 <pre>
1965 </pre>
1966 An invalid combination of PCRE_NEWLINE_<i>xxx</i> options was given.
1967 <pre>
1969 </pre>
1970 The value of <i>startoffset</i> was negative or greater than the length of the
1971 subject, that is, the value in <i>length</i>.
1972 <pre>
1974 </pre>
1975 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
1976 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
1977 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
1978 fact sufficient to detect this case, but this special error code for
1979 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
1980 retained for backwards compatibility.
1981 <pre>
1983 </pre>
1984 This error is returned when <b>pcre_exec()</b> detects a recursion loop within
1985 the pattern. Specifically, it means that either the whole pattern or a
1986 subpattern has been called recursively for the second time at the same position
1987 in the subject string. Some simple patterns that might do this are detected and
1988 faulted at compile time, but more complicated cases, in particular mutual
1989 recursions between two different subpatterns, cannot be detected until run
1990 time.
1991 <pre>
1993 </pre>
1994 This error is returned when a pattern that was successfully studied using the
1995 PCRE_STUDY_JIT_COMPILE option is being matched, but the memory available for
1996 the just-in-time processing stack is not large enough. See the
1997 <a href="pcrejit.html"><b>pcrejit</b></a>
1998 documentation for more details.
1999 </P>
2000 <P>
2001 Error numbers -16 to -20 and -22 are not used by <b>pcre_exec()</b>.
2002 <a name="badutf8reasons"></a></P>
2003 <br><b>
2004 Reason codes for invalid UTF-8 strings
2005 </b><br>
2006 <P>
2007 When <b>pcre_exec()</b> returns either PCRE_ERROR_BADUTF8 or
2008 PCRE_ERROR_SHORTUTF8, and the size of the output vector (<i>ovecsize</i>) is at
2009 least 2, the offset of the start of the invalid UTF-8 character is placed in
2010 the first output vector element (<i>ovector[0]</i>) and a reason code is placed
2011 in the second element (<i>ovector[1]</i>). The reason codes are given names in
2012 the <b>pcre.h</b> header file:
2013 <pre>
2019 </pre>
2020 The string ends with a truncated UTF-8 character; the code specifies how many
2021 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2022 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2023 allows for up to 6 bytes, and this is checked first; hence the possibility of
2024 4 or 5 missing bytes.
2025 <pre>
2030 PCRE_UTF8_ERR10
2031 </pre>
2032 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2033 character do not have the binary value 0b10 (that is, either the most
2034 significant bit is 0, or the next bit is 1).
2035 <pre>
2036 PCRE_UTF8_ERR11
2037 PCRE_UTF8_ERR12
2038 </pre>
2039 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2040 these code points are excluded by RFC 3629.
2041 <pre>
2042 PCRE_UTF8_ERR13
2043 </pre>
2044 A 4-byte character has a value greater than 0x10fff; these code points are
2045 excluded by RFC 3629.
2046 <pre>
2047 PCRE_UTF8_ERR14
2048 </pre>
2049 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2050 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2051 from UTF-8.
2052 <pre>
2053 PCRE_UTF8_ERR15
2054 PCRE_UTF8_ERR16
2055 PCRE_UTF8_ERR17
2056 PCRE_UTF8_ERR18
2057 PCRE_UTF8_ERR19
2058 </pre>
2059 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2060 value that can be represented by fewer bytes, which is invalid. For example,
2061 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2062 one byte.
2063 <pre>
2064 PCRE_UTF8_ERR20
2065 </pre>
2066 The two most significant bits of the first byte of a character have the binary
2067 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2068 byte can only validly occur as the second or subsequent byte of a multi-byte
2069 character.
2070 <pre>
2071 PCRE_UTF8_ERR21
2072 </pre>
2073 The first byte of a character has the value 0xfe or 0xff. These values can
2074 never occur in a valid UTF-8 string.
2075 </P>
2076 <br><a name="SEC17" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NUMBER</a><br>
2077 <P>
2078 <b>int pcre_copy_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
2079 <b>int <i>stringcount</i>, int <i>stringnumber</i>, char *<i>buffer</i>,</b>
2080 <b>int <i>buffersize</i>);</b>
2081 </P>
2082 <P>
2083 <b>int pcre_get_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
2084 <b>int <i>stringcount</i>, int <i>stringnumber</i>,</b>
2085 <b>const char **<i>stringptr</i>);</b>
2086 </P>
2087 <P>
2088 <b>int pcre_get_substring_list(const char *<i>subject</i>,</b>
2089 <b>int *<i>ovector</i>, int <i>stringcount</i>, const char ***<i>listptr</i>);</b>
2090 </P>
2091 <P>
2092 Captured substrings can be accessed directly by using the offsets returned by
2093 <b>pcre_exec()</b> in <i>ovector</i>. For convenience, the functions
2094 <b>pcre_copy_substring()</b>, <b>pcre_get_substring()</b>, and
2095 <b>pcre_get_substring_list()</b> are provided for extracting captured substrings
2096 as new, separate, zero-terminated strings. These functions identify substrings
2097 by number. The next section describes functions for extracting named
2098 substrings.
2099 </P>
2100 <P>
2101 A substring that contains a binary zero is correctly extracted and has a
2102 further zero added on the end, but the result is not, of course, a C string.
2103 However, you can process such a string by referring to the length that is
2104 returned by <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>.
2105 Unfortunately, the interface to <b>pcre_get_substring_list()</b> is not adequate
2106 for handling strings containing binary zeros, because the end of the final
2107 string is not independently indicated.
2108 </P>
2109 <P>
2110 The first three arguments are the same for all three of these functions:
2111 <i>subject</i> is the subject string that has just been successfully matched,
2112 <i>ovector</i> is a pointer to the vector of integer offsets that was passed to
2113 <b>pcre_exec()</b>, and <i>stringcount</i> is the number of substrings that were
2114 captured by the match, including the substring that matched the entire regular
2115 expression. This is the value returned by <b>pcre_exec()</b> if it is greater
2116 than zero. If <b>pcre_exec()</b> returned zero, indicating that it ran out of
2117 space in <i>ovector</i>, the value passed as <i>stringcount</i> should be the
2118 number of elements in the vector divided by three.
2119 </P>
2120 <P>
2121 The functions <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>
2122 extract a single substring, whose number is given as <i>stringnumber</i>. A
2123 value of zero extracts the substring that matched the entire pattern, whereas
2124 higher values extract the captured substrings. For <b>pcre_copy_substring()</b>,
2125 the string is placed in <i>buffer</i>, whose length is given by
2126 <i>buffersize</i>, while for <b>pcre_get_substring()</b> a new block of memory is
2127 obtained via <b>pcre_malloc</b>, and its address is returned via
2128 <i>stringptr</i>. The yield of the function is the length of the string, not
2129 including the terminating zero, or one of these error codes:
2130 <pre>
2132 </pre>
2133 The buffer was too small for <b>pcre_copy_substring()</b>, or the attempt to get
2134 memory failed for <b>pcre_get_substring()</b>.
2135 <pre>
2137 </pre>
2138 There is no substring whose number is <i>stringnumber</i>.
2139 </P>
2140 <P>
2141 The <b>pcre_get_substring_list()</b> function extracts all available substrings
2142 and builds a list of pointers to them. All this is done in a single block of
2143 memory that is obtained via <b>pcre_malloc</b>. The address of the memory block
2144 is returned via <i>listptr</i>, which is also the start of the list of string
2145 pointers. The end of the list is marked by a NULL pointer. The yield of the
2146 function is zero if all went well, or the error code
2147 <pre>
2149 </pre>
2150 if the attempt to get the memory block failed.
2151 </P>
2152 <P>
2153 When any of these functions encounter a substring that is unset, which can
2154 happen when capturing subpattern number <i>n+1</i> matches some part of the
2155 subject, but subpattern <i>n</i> has not been used at all, they return an empty
2156 string. This can be distinguished from a genuine zero-length substring by
2157 inspecting the appropriate offset in <i>ovector</i>, which is negative for unset
2158 substrings.
2159 </P>
2160 <P>
2161 The two convenience functions <b>pcre_free_substring()</b> and
2162 <b>pcre_free_substring_list()</b> can be used to free the memory returned by
2163 a previous call of <b>pcre_get_substring()</b> or
2164 <b>pcre_get_substring_list()</b>, respectively. They do nothing more than call
2165 the function pointed to by <b>pcre_free</b>, which of course could be called
2166 directly from a C program. However, PCRE is used in some situations where it is
2167 linked via a special interface to another programming language that cannot use
2168 <b>pcre_free</b> directly; it is for these cases that the functions are
2169 provided.
2170 </P>
2171 <br><a name="SEC18" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NAME</a><br>
2172 <P>
2173 <b>int pcre_get_stringnumber(const pcre *<i>code</i>,</b>
2174 <b>const char *<i>name</i>);</b>
2175 </P>
2176 <P>
2177 <b>int pcre_copy_named_substring(const pcre *<i>code</i>,</b>
2178 <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
2179 <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
2180 <b>char *<i>buffer</i>, int <i>buffersize</i>);</b>
2181 </P>
2182 <P>
2183 <b>int pcre_get_named_substring(const pcre *<i>code</i>,</b>
2184 <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
2185 <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
2186 <b>const char **<i>stringptr</i>);</b>
2187 </P>
2188 <P>
2189 To extract a substring by name, you first have to find associated number.
2190 For example, for this pattern
2191 <pre>
2192 (a+)b(?&#60;xxx&#62;\d+)...
2193 </pre>
2194 the number of the subpattern called "xxx" is 2. If the name is known to be
2195 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2196 calling <b>pcre_get_stringnumber()</b>. The first argument is the compiled
2197 pattern, and the second is the name. The yield of the function is the
2198 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2199 that name.
2200 </P>
2201 <P>
2202 Given the number, you can extract the substring directly, or use one of the
2203 functions described in the previous section. For convenience, there are also
2204 two functions that do the whole job.
2205 </P>
2206 <P>
2207 Most of the arguments of <b>pcre_copy_named_substring()</b> and
2208 <b>pcre_get_named_substring()</b> are the same as those for the similarly named
2209 functions that extract by number. As these are described in the previous
2210 section, they are not re-described here. There are just two differences:
2211 </P>
2212 <P>
2213 First, instead of a substring number, a substring name is given. Second, there
2214 is an extra argument, given at the start, which is a pointer to the compiled
2215 pattern. This is needed in order to gain access to the name-to-number
2216 translation table.
2217 </P>
2218 <P>
2219 These functions call <b>pcre_get_stringnumber()</b>, and if it succeeds, they
2220 then call <b>pcre_copy_substring()</b> or <b>pcre_get_substring()</b>, as
2221 appropriate. <b>NOTE:</b> If PCRE_DUPNAMES is set and there are duplicate names,
2222 the behaviour may not be what you want (see the next section).
2223 </P>
2224 <P>
2225 <b>Warning:</b> If the pattern uses the (?| feature to set up multiple
2226 subpatterns with the same number, as described in the
2227 <a href="pcrepattern.html#dupsubpatternnumber">section on duplicate subpattern numbers</a>
2228 in the
2229 <a href="pcrepattern.html"><b>pcrepattern</b></a>
2230 page, you cannot use names to distinguish the different subpatterns, because
2231 names are not included in the compiled code. The matching process uses only
2232 numbers. For this reason, the use of different names for subpatterns of the
2233 same number causes an error at compile time.
2234 </P>
2235 <br><a name="SEC19" href="#TOC1">DUPLICATE SUBPATTERN NAMES</a><br>
2236 <P>
2237 <b>int pcre_get_stringtable_entries(const pcre *<i>code</i>,</b>
2238 <b>const char *<i>name</i>, char **<i>first</i>, char **<i>last</i>);</b>
2239 </P>
2240 <P>
2241 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2242 are not required to be unique. (Duplicate names are always allowed for
2243 subpatterns with the same number, created by using the (?| feature. Indeed, if
2244 such subpatterns are named, they are required to use the same names.)
2245 </P>
2246 <P>
2247 Normally, patterns with duplicate names are such that in any one match, only
2248 one of the named subpatterns participates. An example is shown in the
2249 <a href="pcrepattern.html"><b>pcrepattern</b></a>
2250 documentation.
2251 </P>
2252 <P>
2253 When duplicates are present, <b>pcre_copy_named_substring()</b> and
2254 <b>pcre_get_named_substring()</b> return the first substring corresponding to
2255 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2256 returned; no data is returned. The <b>pcre_get_stringnumber()</b> function
2257 returns one of the numbers that are associated with the name, but it is not
2258 defined which it is.
2259 </P>
2260 <P>
2261 If you want to get full details of all captured substrings for a given name,
2262 you must use the <b>pcre_get_stringtable_entries()</b> function. The first
2263 argument is the compiled pattern, and the second is the name. The third and
2264 fourth are pointers to variables which are updated by the function. After it
2265 has run, they point to the first and last entries in the name-to-number table
2266 for the given name. The function itself returns the length of each entry, or
2267 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2268 described above in the section entitled <i>Information about a pattern</i>
2269 <a href="#infoaboutpattern">above.</a>
2270 Given all the relevant entries for the name, you can extract each of their
2271 numbers, and hence the captured data, if any.
2272 </P>
2273 <br><a name="SEC20" href="#TOC1">FINDING ALL POSSIBLE MATCHES</a><br>
2274 <P>
2275 The traditional matching function uses a similar algorithm to Perl, which stops
2276 when it finds the first match, starting at a given point in the subject. If you
2277 want to find all possible matches, or the longest possible match, consider
2278 using the alternative matching function (see below) instead. If you cannot use
2279 the alternative function, but still need to find all possible matches, you
2280 can kludge it up by making use of the callout facility, which is described in
2281 the
2282 <a href="pcrecallout.html"><b>pcrecallout</b></a>
2283 documentation.
2284 </P>
2285 <P>
2286 What you have to do is to insert a callout right at the end of the pattern.
2287 When your callout function is called, extract and save the current matched
2288 substring. Then return 1, which forces <b>pcre_exec()</b> to backtrack and try
2289 other alternatives. Ultimately, when it runs out of matches, <b>pcre_exec()</b>
2290 will yield PCRE_ERROR_NOMATCH.
2291 <a name="dfamatch"></a></P>
2292 <br><a name="SEC21" href="#TOC1">MATCHING A PATTERN: THE ALTERNATIVE FUNCTION</a><br>
2293 <P>
2294 <b>int pcre_dfa_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
2295 <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
2296 <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>,</b>
2297 <b>int *<i>workspace</i>, int <i>wscount</i>);</b>
2298 </P>
2299 <P>
2300 The function <b>pcre_dfa_exec()</b> is called to match a subject string against
2301 a compiled pattern, using a matching algorithm that scans the subject string
2302 just once, and does not backtrack. This has different characteristics to the
2303 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2304 patterns are not supported. Nevertheless, there are times when this kind of
2305 matching can be useful. For a discussion of the two matching algorithms, and a
2306 list of features that <b>pcre_dfa_exec()</b> does not support, see the
2307 <a href="pcrematching.html"><b>pcrematching</b></a>
2308 documentation.
2309 </P>
2310 <P>
2311 The arguments for the <b>pcre_dfa_exec()</b> function are the same as for
2312 <b>pcre_exec()</b>, plus two extras. The <i>ovector</i> argument is used in a
2313 different way, and this is described below. The other common arguments are used
2314 in the same way as for <b>pcre_exec()</b>, so their description is not repeated
2315 here.
2316 </P>
2317 <P>
2318 The two additional arguments provide workspace for the function. The workspace
2319 vector should contain at least 20 elements. It is used for keeping track of
2320 multiple paths through the pattern tree. More workspace will be needed for
2321 patterns and subjects where there are a lot of potential matches.
2322 </P>
2323 <P>
2324 Here is an example of a simple call to <b>pcre_dfa_exec()</b>:
2325 <pre>
2326 int rc;
2327 int ovector[10];
2328 int wspace[20];
2329 rc = pcre_dfa_exec(
2330 re, /* result of pcre_compile() */
2331 NULL, /* we didn't study the pattern */
2332 "some string", /* the subject string */
2333 11, /* the length of the subject string */
2334 0, /* start at offset 0 in the subject */
2335 0, /* default options */
2336 ovector, /* vector of integers for substring information */
2337 10, /* number of elements (NOT size in bytes) */
2338 wspace, /* working space vector */
2339 20); /* number of elements (NOT size in bytes) */
2340 </PRE>
2341 </P>
2342 <br><b>
2343 Option bits for <b>pcre_dfa_exec()</b>
2344 </b><br>
2345 <P>
2346 The unused bits of the <i>options</i> argument for <b>pcre_dfa_exec()</b> must be
2347 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
2351 All but the last four of these are exactly the same as for <b>pcre_exec()</b>,
2352 so their description is not repeated here.
2353 <pre>
2356 </pre>
2357 These have the same general effect as they do for <b>pcre_exec()</b>, but the
2358 details are slightly different. When PCRE_PARTIAL_HARD is set for
2359 <b>pcre_dfa_exec()</b>, it returns PCRE_ERROR_PARTIAL if the end of the subject
2360 is reached and there is still at least one matching possibility that requires
2361 additional characters. This happens even if some complete matches have also
2362 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2363 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2364 there have been no complete matches, but there is still at least one matching
2365 possibility. The portion of the string that was inspected when the longest
2366 partial match was found is set as the first matching string in both cases.
2367 There is a more detailed discussion of partial and multi-segment matching, with
2368 examples, in the
2369 <a href="pcrepartial.html"><b>pcrepartial</b></a>
2370 documentation.
2371 <pre>
2373 </pre>
2374 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2375 soon as it has found one match. Because of the way the alternative algorithm
2376 works, this is necessarily the shortest possible match at the first possible
2377 matching point in the subject string.
2378 <pre>
2380 </pre>
2381 When <b>pcre_dfa_exec()</b> returns a partial match, it is possible to call it
2382 again, with additional subject characters, and have it continue with the same
2383 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2384 <i>workspace</i> and <i>wscount</i> options must reference the same vector as
2385 before because data about the match so far is left in them after a partial
2386 match. There is more discussion of this facility in the
2387 <a href="pcrepartial.html"><b>pcrepartial</b></a>
2388 documentation.
2389 </P>
2390 <br><b>
2391 Successful returns from <b>pcre_dfa_exec()</b>
2392 </b><br>
2393 <P>
2394 When <b>pcre_dfa_exec()</b> succeeds, it may have matched more than one
2395 substring in the subject. Note, however, that all the matches from one run of
2396 the function start at the same point in the subject. The shorter matches are
2397 all initial substrings of the longer matches. For example, if the pattern
2398 <pre>
2399 &#60;.*&#62;
2400 </pre>
2401 is matched against the string
2402 <pre>
2403 This is &#60;something&#62; &#60;something else&#62; &#60;something further&#62; no more
2404 </pre>
2405 the three matched strings are
2406 <pre>
2407 &#60;something&#62;
2408 &#60;something&#62; &#60;something else&#62;
2409 &#60;something&#62; &#60;something else&#62; &#60;something further&#62;
2410 </pre>
2411 On success, the yield of the function is a number greater than zero, which is
2412 the number of matched substrings. The substrings themselves are returned in
2413 <i>ovector</i>. Each string uses two elements; the first is the offset to the
2414 start, and the second is the offset to the end. In fact, all the strings have
2415 the same start offset. (Space could have been saved by giving this only once,
2416 but it was decided to retain some compatibility with the way <b>pcre_exec()</b>
2417 returns data, even though the meaning of the strings is different.)
2418 </P>
2419 <P>
2420 The strings are returned in reverse order of length; that is, the longest
2421 matching string is given first. If there were too many matches to fit into
2422 <i>ovector</i>, the yield of the function is zero, and the vector is filled with
2423 the longest matches. Unlike <b>pcre_exec()</b>, <b>pcre_dfa_exec()</b> can use
2424 the entire <i>ovector</i> for returning matched strings.
2425 </P>
2426 <br><b>
2427 Error returns from <b>pcre_dfa_exec()</b>
2428 </b><br>
2429 <P>
2430 The <b>pcre_dfa_exec()</b> function returns a negative number when it fails.
2431 Many of the errors are the same as for <b>pcre_exec()</b>, and these are
2432 described
2433 <a href="#errorlist">above.</a>
2434 There are in addition the following errors that are specific to
2435 <b>pcre_dfa_exec()</b>:
2436 <pre>
2438 </pre>
2439 This return is given if <b>pcre_dfa_exec()</b> encounters an item in the pattern
2440 that it does not support, for instance, the use of \C or a back reference.
2441 <pre>
2443 </pre>
2444 This return is given if <b>pcre_dfa_exec()</b> encounters a condition item that
2445 uses a back reference for the condition, or a test for recursion in a specific
2446 group. These are not supported.
2447 <pre>
2449 </pre>
2450 This return is given if <b>pcre_dfa_exec()</b> is called with an <i>extra</i>
2451 block that contains a setting of the <i>match_limit</i> or
2452 <i>match_limit_recursion</i> fields. This is not supported (these fields are
2453 meaningless for DFA matching).
2454 <pre>
2456 </pre>
2457 This return is given if <b>pcre_dfa_exec()</b> runs out of space in the
2458 <i>workspace</i> vector.
2459 <pre>
2461 </pre>
2462 When a recursive subpattern is processed, the matching function calls itself
2463 recursively, using private vectors for <i>ovector</i> and <i>workspace</i>. This
2464 error is given if the output vector is not large enough. This should be
2465 extremely rare, as a vector of size 1000 is used.
2466 </P>
2467 <br><a name="SEC22" href="#TOC1">SEE ALSO</a><br>
2468 <P>
2469 <b>pcrebuild</b>(3), <b>pcrecallout</b>(3), <b>pcrecpp(3)</b>(3),
2470 <b>pcrematching</b>(3), <b>pcrepartial</b>(3), <b>pcreposix</b>(3),
2471 <b>pcreprecompile</b>(3), <b>pcresample</b>(3), <b>pcrestack</b>(3).
2472 </P>
2473 <br><a name="SEC23" href="#TOC1">AUTHOR</a><br>
2474 <P>
2475 Philip Hazel
2476 <br>
2477 University Computing Service
2478 <br>
2479 Cambridge CB2 3QH, England.
2480 <br>
2481 </P>
2482 <br><a name="SEC24" href="#TOC1">REVISION</a><br>
2483 <P>
2484 Last updated: 06 September 2011
2485 <br>
2486 Copyright &copy; 1997-2011 University of Cambridge.
2487 <br>
2488 <p>
2489 Return to the <a href="index.html">PCRE index page</a>.
2490 </p>


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

  ViewVC Help
Powered by ViewVC 1.1.5