ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log

Revision 1320 - (show annotations)
Wed May 1 16:39:35 2013 UTC (6 years, 4 months ago) by ph10
File MIME type: text/html
File size: 130978 byte(s)
Error occurred while calculating annotation data.
Source tidies (trails spaces, html updates) for 8.33-RC1.
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>
18 <li><a name="TOC3" href="#SEC3">PCRE NATIVE API AUXILIARY FUNCTIONS</a>
20 <li><a name="TOC5" href="#SEC5">PCRE 8-BIT, 16-BIT, AND 32-BIT LIBRARIES</a>
21 <li><a name="TOC6" href="#SEC6">PCRE API OVERVIEW</a>
22 <li><a name="TOC7" href="#SEC7">NEWLINES</a>
23 <li><a name="TOC8" href="#SEC8">MULTITHREADING</a>
25 <li><a name="TOC10" href="#SEC10">CHECKING BUILD-TIME OPTIONS</a>
26 <li><a name="TOC11" href="#SEC11">COMPILING A PATTERN</a>
27 <li><a name="TOC12" href="#SEC12">COMPILATION ERROR CODES</a>
28 <li><a name="TOC13" href="#SEC13">STUDYING A PATTERN</a>
29 <li><a name="TOC14" href="#SEC14">LOCALE SUPPORT</a>
30 <li><a name="TOC15" href="#SEC15">INFORMATION ABOUT A PATTERN</a>
31 <li><a name="TOC16" href="#SEC16">REFERENCE COUNTS</a>
35 <li><a name="TOC20" href="#SEC20">DUPLICATE SUBPATTERN NAMES</a>
36 <li><a name="TOC21" href="#SEC21">FINDING ALL POSSIBLE MATCHES</a>
37 <li><a name="TOC22" href="#SEC22">OBTAINING AN ESTIMATE OF STACK USAGE</a>
39 <li><a name="TOC24" href="#SEC24">SEE ALSO</a>
40 <li><a name="TOC25" href="#SEC25">AUTHOR</a>
41 <li><a name="TOC26" href="#SEC26">REVISION</a>
42 </ul>
43 <P>
44 <b>#include &#60;pcre.h&#62;</b>
45 </P>
46 <br><a name="SEC1" href="#TOC1">PCRE NATIVE API BASIC FUNCTIONS</a><br>
47 <P>
48 <b>pcre *pcre_compile(const char *<i>pattern</i>, int <i>options</i>,</b>
49 <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
50 <b>const unsigned char *<i>tableptr</i>);</b>
51 </P>
52 <P>
53 <b>pcre *pcre_compile2(const char *<i>pattern</i>, int <i>options</i>,</b>
54 <b>int *<i>errorcodeptr</i>,</b>
55 <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
56 <b>const unsigned char *<i>tableptr</i>);</b>
57 </P>
58 <P>
59 <b>pcre_extra *pcre_study(const pcre *<i>code</i>, int <i>options</i>,</b>
60 <b>const char **<i>errptr</i>);</b>
61 </P>
62 <P>
63 <b>void pcre_free_study(pcre_extra *<i>extra</i>);</b>
64 </P>
65 <P>
66 <b>int pcre_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
67 <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
68 <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>);</b>
69 </P>
70 <P>
71 <b>int pcre_dfa_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
72 <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
73 <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>,</b>
74 <b>int *<i>workspace</i>, int <i>wscount</i>);</b>
75 </P>
77 <P>
78 <b>int pcre_copy_named_substring(const pcre *<i>code</i>,</b>
79 <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
80 <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
81 <b>char *<i>buffer</i>, int <i>buffersize</i>);</b>
82 </P>
83 <P>
84 <b>int pcre_copy_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
85 <b>int <i>stringcount</i>, int <i>stringnumber</i>, char *<i>buffer</i>,</b>
86 <b>int <i>buffersize</i>);</b>
87 </P>
88 <P>
89 <b>int pcre_get_named_substring(const pcre *<i>code</i>,</b>
90 <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
91 <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
92 <b>const char **<i>stringptr</i>);</b>
93 </P>
94 <P>
95 <b>int pcre_get_stringnumber(const pcre *<i>code</i>,</b>
96 <b>const char *<i>name</i>);</b>
97 </P>
98 <P>
99 <b>int pcre_get_stringtable_entries(const pcre *<i>code</i>,</b>
100 <b>const char *<i>name</i>, char **<i>first</i>, char **<i>last</i>);</b>
101 </P>
102 <P>
103 <b>int pcre_get_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
104 <b>int <i>stringcount</i>, int <i>stringnumber</i>,</b>
105 <b>const char **<i>stringptr</i>);</b>
106 </P>
107 <P>
108 <b>int pcre_get_substring_list(const char *<i>subject</i>,</b>
109 <b>int *<i>ovector</i>, int <i>stringcount</i>, const char ***<i>listptr</i>);</b>
110 </P>
111 <P>
112 <b>void pcre_free_substring(const char *<i>stringptr</i>);</b>
113 </P>
114 <P>
115 <b>void pcre_free_substring_list(const char **<i>stringptr</i>);</b>
116 </P>
117 <br><a name="SEC3" href="#TOC1">PCRE NATIVE API AUXILIARY FUNCTIONS</a><br>
118 <P>
119 <b>int pcre_jit_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
120 <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
121 <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>,</b>
122 <b>pcre_jit_stack *<i>jstack</i>);</b>
123 </P>
124 <P>
125 <b>pcre_jit_stack *pcre_jit_stack_alloc(int <i>startsize</i>, int <i>maxsize</i>);</b>
126 </P>
127 <P>
128 <b>void pcre_jit_stack_free(pcre_jit_stack *<i>stack</i>);</b>
129 </P>
130 <P>
131 <b>void pcre_assign_jit_stack(pcre_extra *<i>extra</i>,</b>
132 <b>pcre_jit_callback <i>callback</i>, void *<i>data</i>);</b>
133 </P>
134 <P>
135 <b>const unsigned char *pcre_maketables(void);</b>
136 </P>
137 <P>
138 <b>int pcre_fullinfo(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
139 <b>int <i>what</i>, void *<i>where</i>);</b>
140 </P>
141 <P>
142 <b>int pcre_refcount(pcre *<i>code</i>, int <i>adjust</i>);</b>
143 </P>
144 <P>
145 <b>int pcre_config(int <i>what</i>, void *<i>where</i>);</b>
146 </P>
147 <P>
148 <b>const char *pcre_version(void);</b>
149 </P>
150 <P>
151 <b>int pcre_pattern_to_host_byte_order(pcre *<i>code</i>,</b>
152 <b>pcre_extra *<i>extra</i>, const unsigned char *<i>tables</i>);</b>
153 </P>
154 <br><a name="SEC4" href="#TOC1">PCRE NATIVE API INDIRECTED FUNCTIONS</a><br>
155 <P>
156 <b>void *(*pcre_malloc)(size_t);</b>
157 </P>
158 <P>
159 <b>void (*pcre_free)(void *);</b>
160 </P>
161 <P>
162 <b>void *(*pcre_stack_malloc)(size_t);</b>
163 </P>
164 <P>
165 <b>void (*pcre_stack_free)(void *);</b>
166 </P>
167 <P>
168 <b>int (*pcre_callout)(pcre_callout_block *);</b>
169 </P>
170 <br><a name="SEC5" href="#TOC1">PCRE 8-BIT, 16-BIT, AND 32-BIT LIBRARIES</a><br>
171 <P>
172 As well as support for 8-bit character strings, PCRE also supports 16-bit
173 strings (from release 8.30) and 32-bit strings (from release 8.32), by means of
174 two additional libraries. They can be built as well as, or instead of, the
175 8-bit library. To avoid too much complication, this document describes the
176 8-bit versions of the functions, with only occasional references to the 16-bit
177 and 32-bit libraries.
178 </P>
179 <P>
180 The 16-bit and 32-bit functions operate in the same way as their 8-bit
181 counterparts; they just use different data types for their arguments and
182 results, and their names start with <b>pcre16_</b> or <b>pcre32_</b> instead of
183 <b>pcre_</b>. For every option that has UTF8 in its name (for example,
184 PCRE_UTF8), there are corresponding 16-bit and 32-bit names with UTF8 replaced
185 by UTF16 or UTF32, respectively. This facility is in fact just cosmetic; the
186 16-bit and 32-bit option names define the same bit values.
187 </P>
188 <P>
189 References to bytes and UTF-8 in this document should be read as references to
190 16-bit data quantities and UTF-16 when using the 16-bit library, or 32-bit data
191 quantities and UTF-32 when using the 32-bit library, unless specified
192 otherwise. More details of the specific differences for the 16-bit and 32-bit
193 libraries are given in the
194 <a href="pcre16.html"><b>pcre16</b></a>
195 and
196 <a href="pcre32.html"><b>pcre32</b></a>
197 pages.
198 </P>
199 <br><a name="SEC6" href="#TOC1">PCRE API OVERVIEW</a><br>
200 <P>
201 PCRE has its own native API, which is described in this document. There are
202 also some wrapper functions (for the 8-bit library only) that correspond to the
203 POSIX regular expression API, but they do not give access to all the
204 functionality. They are described in the
205 <a href="pcreposix.html"><b>pcreposix</b></a>
206 documentation. Both of these APIs define a set of C function calls. A C++
207 wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
208 documented in the
209 <a href="pcrecpp.html"><b>pcrecpp</b></a>
210 page.
211 </P>
212 <P>
213 The native API C function prototypes are defined in the header file
214 <b>pcre.h</b>, and on Unix-like systems the (8-bit) library itself is called
215 <b>libpcre</b>. It can normally be accessed by adding <b>-lpcre</b> to the
216 command for linking an application that uses PCRE. The header file defines the
217 macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
218 for the library. Applications can use these to include support for different
219 releases of PCRE.
220 </P>
221 <P>
222 In a Windows environment, if you want to statically link an application program
223 against a non-dll <b>pcre.a</b> file, you must define PCRE_STATIC before
224 including <b>pcre.h</b> or <b>pcrecpp.h</b>, because otherwise the
225 <b>pcre_malloc()</b> and <b>pcre_free()</b> exported functions will be declared
226 <b>__declspec(dllimport)</b>, with unwanted results.
227 </P>
228 <P>
229 The functions <b>pcre_compile()</b>, <b>pcre_compile2()</b>, <b>pcre_study()</b>,
230 and <b>pcre_exec()</b> are used for compiling and matching regular expressions
231 in a Perl-compatible manner. A sample program that demonstrates the simplest
232 way of using them is provided in the file called <i>pcredemo.c</i> in the PCRE
233 source distribution. A listing of this program is given in the
234 <a href="pcredemo.html"><b>pcredemo</b></a>
235 documentation, and the
236 <a href="pcresample.html"><b>pcresample</b></a>
237 documentation describes how to compile and run it.
238 </P>
239 <P>
240 Just-in-time compiler support is an optional feature of PCRE that can be built
241 in appropriate hardware environments. It greatly speeds up the matching
242 performance of many patterns. Simple programs can easily request that it be
243 used if available, by setting an option that is ignored when it is not
244 relevant. More complicated programs might need to make use of the functions
245 <b>pcre_jit_stack_alloc()</b>, <b>pcre_jit_stack_free()</b>, and
246 <b>pcre_assign_jit_stack()</b> in order to control the JIT code's memory usage.
247 </P>
248 <P>
249 From release 8.32 there is also a direct interface for JIT execution, which
250 gives improved performance. The JIT-specific functions are discussed in the
251 <a href="pcrejit.html"><b>pcrejit</b></a>
252 documentation.
253 </P>
254 <P>
255 A second matching function, <b>pcre_dfa_exec()</b>, which is not
256 Perl-compatible, is also provided. This uses a different algorithm for the
257 matching. The alternative algorithm finds all possible matches (at a given
258 point in the subject), and scans the subject just once (unless there are
259 lookbehind assertions). However, this algorithm does not return captured
260 substrings. A description of the two matching algorithms and their advantages
261 and disadvantages is given in the
262 <a href="pcrematching.html"><b>pcrematching</b></a>
263 documentation.
264 </P>
265 <P>
266 In addition to the main compiling and matching functions, there are convenience
267 functions for extracting captured substrings from a subject string that is
268 matched by <b>pcre_exec()</b>. They are:
269 <pre>
270 <b>pcre_copy_substring()</b>
271 <b>pcre_copy_named_substring()</b>
272 <b>pcre_get_substring()</b>
273 <b>pcre_get_named_substring()</b>
274 <b>pcre_get_substring_list()</b>
275 <b>pcre_get_stringnumber()</b>
276 <b>pcre_get_stringtable_entries()</b>
277 </pre>
278 <b>pcre_free_substring()</b> and <b>pcre_free_substring_list()</b> are also
279 provided, to free the memory used for extracted strings.
280 </P>
281 <P>
282 The function <b>pcre_maketables()</b> is used to build a set of character tables
283 in the current locale for passing to <b>pcre_compile()</b>, <b>pcre_exec()</b>,
284 or <b>pcre_dfa_exec()</b>. This is an optional facility that is provided for
285 specialist use. Most commonly, no special tables are passed, in which case
286 internal tables that are generated when PCRE is built are used.
287 </P>
288 <P>
289 The function <b>pcre_fullinfo()</b> is used to find out information about a
290 compiled pattern. The function <b>pcre_version()</b> returns a pointer to a
291 string containing the version of PCRE and its date of release.
292 </P>
293 <P>
294 The function <b>pcre_refcount()</b> maintains a reference count in a data block
295 containing a compiled pattern. This is provided for the benefit of
296 object-oriented applications.
297 </P>
298 <P>
299 The global variables <b>pcre_malloc</b> and <b>pcre_free</b> initially contain
300 the entry points of the standard <b>malloc()</b> and <b>free()</b> functions,
301 respectively. PCRE calls the memory management functions via these variables,
302 so a calling program can replace them if it wishes to intercept the calls. This
303 should be done before calling any PCRE functions.
304 </P>
305 <P>
306 The global variables <b>pcre_stack_malloc</b> and <b>pcre_stack_free</b> are also
307 indirections to memory management functions. These special functions are used
308 only when PCRE is compiled to use the heap for remembering data, instead of
309 recursive function calls, when running the <b>pcre_exec()</b> function. See the
310 <a href="pcrebuild.html"><b>pcrebuild</b></a>
311 documentation for details of how to do this. It is a non-standard way of
312 building PCRE, for use in environments that have limited stacks. Because of the
313 greater use of memory management, it runs more slowly. Separate functions are
314 provided so that special-purpose external code can be used for this case. When
315 used, these functions are always called in a stack-like manner (last obtained,
316 first freed), and always for memory blocks of the same size. There is a
317 discussion about PCRE's stack usage in the
318 <a href="pcrestack.html"><b>pcrestack</b></a>
319 documentation.
320 </P>
321 <P>
322 The global variable <b>pcre_callout</b> initially contains NULL. It can be set
323 by the caller to a "callout" function, which PCRE will then call at specified
324 points during a matching operation. Details are given in the
325 <a href="pcrecallout.html"><b>pcrecallout</b></a>
326 documentation.
327 <a name="newlines"></a></P>
328 <br><a name="SEC7" href="#TOC1">NEWLINES</a><br>
329 <P>
330 PCRE supports five different conventions for indicating line breaks in
331 strings: a single CR (carriage return) character, a single LF (linefeed)
332 character, the two-character sequence CRLF, any of the three preceding, or any
333 Unicode newline sequence. The Unicode newline sequences are the three just
334 mentioned, plus the single characters VT (vertical tab, U+000B), FF (form feed,
335 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
336 (paragraph separator, U+2029).
337 </P>
338 <P>
339 Each of the first three conventions is used by at least one operating system as
340 its standard newline sequence. When PCRE is built, a default can be specified.
341 The default default is LF, which is the Unix standard. When PCRE is run, the
342 default can be overridden, either when a pattern is compiled, or when it is
343 matched.
344 </P>
345 <P>
346 At compile time, the newline convention can be specified by the <i>options</i>
347 argument of <b>pcre_compile()</b>, or it can be specified by special text at the
348 start of the pattern itself; this overrides any other settings. See the
349 <a href="pcrepattern.html"><b>pcrepattern</b></a>
350 page for details of the special character sequences.
351 </P>
352 <P>
353 In the PCRE documentation the word "newline" is used to mean "the character or
354 pair of characters that indicate a line break". The choice of newline
355 convention affects the handling of the dot, circumflex, and dollar
356 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
357 recognized line ending sequence, the match position advancement for a
358 non-anchored pattern. There is more detail about this in the
359 <a href="#execoptions">section on <b>pcre_exec()</b> options</a>
360 below.
361 </P>
362 <P>
363 The choice of newline convention does not affect the interpretation of
364 the \n or \r escape sequences, nor does it affect what \R matches, which is
365 controlled in a similar way, but by separate options.
366 </P>
367 <br><a name="SEC8" href="#TOC1">MULTITHREADING</a><br>
368 <P>
369 The PCRE functions can be used in multi-threading applications, with the
370 proviso that the memory management functions pointed to by <b>pcre_malloc</b>,
371 <b>pcre_free</b>, <b>pcre_stack_malloc</b>, and <b>pcre_stack_free</b>, and the
372 callout function pointed to by <b>pcre_callout</b>, are shared by all threads.
373 </P>
374 <P>
375 The compiled form of a regular expression is not altered during matching, so
376 the same compiled pattern can safely be used by several threads at once.
377 </P>
378 <P>
379 If the just-in-time optimization feature is being used, it needs separate
380 memory stack areas for each thread. See the
381 <a href="pcrejit.html"><b>pcrejit</b></a>
382 documentation for more details.
383 </P>
384 <br><a name="SEC9" href="#TOC1">SAVING PRECOMPILED PATTERNS FOR LATER USE</a><br>
385 <P>
386 The compiled form of a regular expression can be saved and re-used at a later
387 time, possibly by a different program, and even on a host other than the one on
388 which it was compiled. Details are given in the
389 <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
390 documentation, which includes a description of the
391 <b>pcre_pattern_to_host_byte_order()</b> function. However, compiling a regular
392 expression with one version of PCRE for use with a different version is not
393 guaranteed to work and may cause crashes.
394 </P>
395 <br><a name="SEC10" href="#TOC1">CHECKING BUILD-TIME OPTIONS</a><br>
396 <P>
397 <b>int pcre_config(int <i>what</i>, void *<i>where</i>);</b>
398 </P>
399 <P>
400 The function <b>pcre_config()</b> makes it possible for a PCRE client to
401 discover which optional features have been compiled into the PCRE library. The
402 <a href="pcrebuild.html"><b>pcrebuild</b></a>
403 documentation has more details about these optional features.
404 </P>
405 <P>
406 The first argument for <b>pcre_config()</b> is an integer, specifying which
407 information is required; the second argument is a pointer to a variable into
408 which the information is placed. The returned value is zero on success, or the
409 negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
410 not recognized. The following information is available:
411 <pre>
413 </pre>
414 The output is an integer that is set to one if UTF-8 support is available;
415 otherwise it is set to zero. This value should normally be given to the 8-bit
416 version of this function, <b>pcre_config()</b>. If it is given to the 16-bit
417 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
418 <pre>
420 </pre>
421 The output is an integer that is set to one if UTF-16 support is available;
422 otherwise it is set to zero. This value should normally be given to the 16-bit
423 version of this function, <b>pcre16_config()</b>. If it is given to the 8-bit
424 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
425 <pre>
427 </pre>
428 The output is an integer that is set to one if UTF-32 support is available;
429 otherwise it is set to zero. This value should normally be given to the 32-bit
430 version of this function, <b>pcre32_config()</b>. If it is given to the 8-bit
431 or 16-bit version of this function, the result is PCRE_ERROR_BADOPTION.
432 <pre>
434 </pre>
435 The output is an integer that is set to one if support for Unicode character
436 properties is available; otherwise it is set to zero.
437 <pre>
439 </pre>
440 The output is an integer that is set to one if support for just-in-time
441 compiling is available; otherwise it is set to zero.
442 <pre>
444 </pre>
445 The output is a pointer to a zero-terminated "const char *" string. If JIT
446 support is available, the string contains the name of the architecture for
447 which the JIT compiler is configured, for example "x86 32bit (little endian +
448 unaligned)". If JIT support is not available, the result is NULL.
449 <pre>
451 </pre>
452 The output is an integer whose value specifies the default character sequence
453 that is recognized as meaning "newline". The values that are supported in
454 ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for
455 ANYCRLF, and -1 for ANY. In EBCDIC environments, CR, ANYCRLF, and ANY yield the
456 same values. However, the value for LF is normally 21, though some EBCDIC
457 environments use 37. The corresponding values for CRLF are 3349 and 3365. The
458 default should normally correspond to the standard sequence for your operating
459 system.
460 <pre>
462 </pre>
463 The output is an integer whose value indicates what character sequences the \R
464 escape sequence matches by default. A value of 0 means that \R matches any
465 Unicode line ending sequence; a value of 1 means that \R matches only CR, LF,
466 or CRLF. The default can be overridden when a pattern is compiled or matched.
467 <pre>
469 </pre>
470 The output is an integer that contains the number of bytes used for internal
471 linkage in compiled regular expressions. For the 8-bit library, the value can
472 be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
473 a number of bytes. For the 32-bit library, the value is either 2 or 4 and is
474 still a number of bytes. The default value of 2 is sufficient for all but the
475 most massive patterns, since it allows the compiled pattern to be up to 64K in
476 size. Larger values allow larger regular expressions to be compiled, at the
477 expense of slower matching.
478 <pre>
480 </pre>
481 The output is an integer that contains the threshold above which the POSIX
482 interface uses <b>malloc()</b> for output vectors. Further details are given in
483 the
484 <a href="pcreposix.html"><b>pcreposix</b></a>
485 documentation.
486 <pre>
488 </pre>
489 The output is a long integer that gives the default limit for the number of
490 internal matching function calls in a <b>pcre_exec()</b> execution. Further
491 details are given with <b>pcre_exec()</b> below.
492 <pre>
494 </pre>
495 The output is a long integer that gives the default limit for the depth of
496 recursion when calling the internal matching function in a <b>pcre_exec()</b>
497 execution. Further details are given with <b>pcre_exec()</b> below.
498 <pre>
500 </pre>
501 The output is an integer that is set to one if internal recursion when running
502 <b>pcre_exec()</b> is implemented by recursive function calls that use the stack
503 to remember their state. This is the usual way that PCRE is compiled. The
504 output is zero if PCRE was compiled to use blocks of data on the heap instead
505 of recursive function calls. In this case, <b>pcre_stack_malloc</b> and
506 <b>pcre_stack_free</b> are called to manage memory blocks on the heap, thus
507 avoiding the use of the stack.
508 </P>
509 <br><a name="SEC11" href="#TOC1">COMPILING A PATTERN</a><br>
510 <P>
511 <b>pcre *pcre_compile(const char *<i>pattern</i>, int <i>options</i>,</b>
512 <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
513 <b>const unsigned char *<i>tableptr</i>);</b>
514 <b>pcre *pcre_compile2(const char *<i>pattern</i>, int <i>options</i>,</b>
515 <b>int *<i>errorcodeptr</i>,</b>
516 <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
517 <b>const unsigned char *<i>tableptr</i>);</b>
518 </P>
519 <P>
520 Either of the functions <b>pcre_compile()</b> or <b>pcre_compile2()</b> can be
521 called to compile a pattern into an internal form. The only difference between
522 the two interfaces is that <b>pcre_compile2()</b> has an additional argument,
523 <i>errorcodeptr</i>, via which a numerical error code can be returned. To avoid
524 too much repetition, we refer just to <b>pcre_compile()</b> below, but the
525 information applies equally to <b>pcre_compile2()</b>.
526 </P>
527 <P>
528 The pattern is a C string terminated by a binary zero, and is passed in the
529 <i>pattern</i> argument. A pointer to a single block of memory that is obtained
530 via <b>pcre_malloc</b> is returned. This contains the compiled code and related
531 data. The <b>pcre</b> type is defined for the returned block; this is a typedef
532 for a structure whose contents are not externally defined. It is up to the
533 caller to free the memory (via <b>pcre_free</b>) when it is no longer required.
534 </P>
535 <P>
536 Although the compiled code of a PCRE regex is relocatable, that is, it does not
537 depend on memory location, the complete <b>pcre</b> data block is not
538 fully relocatable, because it may contain a copy of the <i>tableptr</i>
539 argument, which is an address (see below).
540 </P>
541 <P>
542 The <i>options</i> argument contains various bit settings that affect the
543 compilation. It should be zero if no options are required. The available
544 options are described below. Some of them (in particular, those that are
545 compatible with Perl, but some others as well) can also be set and unset from
546 within the pattern (see the detailed description in the
547 <a href="pcrepattern.html"><b>pcrepattern</b></a>
548 documentation). For those options that can be different in different parts of
549 the pattern, the contents of the <i>options</i> argument specifies their
550 settings at the start of compilation and execution. The PCRE_ANCHORED,
551 PCRE_BSR_<i>xxx</i>, PCRE_NEWLINE_<i>xxx</i>, PCRE_NO_UTF8_CHECK, and
552 PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
553 compile time.
554 </P>
555 <P>
556 If <i>errptr</i> is NULL, <b>pcre_compile()</b> returns NULL immediately.
557 Otherwise, if compilation of a pattern fails, <b>pcre_compile()</b> returns
558 NULL, and sets the variable pointed to by <i>errptr</i> to point to a textual
559 error message. This is a static string that is part of the library. You must
560 not try to free it. Normally, the offset from the start of the pattern to the
561 byte that was being processed when the error was discovered is placed in the
562 variable pointed to by <i>erroffset</i>, which must not be NULL (if it is, an
563 immediate error is given). However, for an invalid UTF-8 string, the offset is
564 that of the first byte of the failing character.
565 </P>
566 <P>
567 Some errors are not detected until the whole pattern has been scanned; in these
568 cases, the offset passed back is the length of the pattern. Note that the
569 offset is in bytes, not characters, even in UTF-8 mode. It may sometimes point
570 into the middle of a UTF-8 character.
571 </P>
572 <P>
573 If <b>pcre_compile2()</b> is used instead of <b>pcre_compile()</b>, and the
574 <i>errorcodeptr</i> argument is not NULL, a non-zero error code number is
575 returned via this argument in the event of an error. This is in addition to the
576 textual error message. Error codes and messages are listed below.
577 </P>
578 <P>
579 If the final argument, <i>tableptr</i>, is NULL, PCRE uses a default set of
580 character tables that are built when PCRE is compiled, using the default C
581 locale. Otherwise, <i>tableptr</i> must be an address that is the result of a
582 call to <b>pcre_maketables()</b>. This value is stored with the compiled
583 pattern, and used again by <b>pcre_exec()</b>, unless another table pointer is
584 passed to it. For more discussion, see the section on locale support below.
585 </P>
586 <P>
587 This code fragment shows a typical straightforward call to <b>pcre_compile()</b>:
588 <pre>
589 pcre *re;
590 const char *error;
591 int erroffset;
592 re = pcre_compile(
593 "^A.*Z", /* the pattern */
594 0, /* default options */
595 &error, /* for error message */
596 &erroffset, /* for error offset */
597 NULL); /* use default character tables */
598 </pre>
599 The following names for option bits are defined in the <b>pcre.h</b> header
600 file:
601 <pre>
603 </pre>
604 If this bit is set, the pattern is forced to be "anchored", that is, it is
605 constrained to match only at the first matching point in the string that is
606 being searched (the "subject string"). This effect can also be achieved by
607 appropriate constructs in the pattern itself, which is the only way to do it in
608 Perl.
609 <pre>
611 </pre>
612 If this bit is set, <b>pcre_compile()</b> automatically inserts callout items,
613 all with number 255, before each pattern item. For discussion of the callout
614 facility, see the
615 <a href="pcrecallout.html"><b>pcrecallout</b></a>
616 documentation.
617 <pre>
620 </pre>
621 These options (which are mutually exclusive) control what the \R escape
622 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
623 match any Unicode newline sequence. The default is specified when PCRE is
624 built. It can be overridden from within the pattern, or by setting an option
625 when a compiled pattern is matched.
626 <pre>
628 </pre>
629 If this bit is set, letters in the pattern match both upper and lower case
630 letters. It is equivalent to Perl's /i option, and it can be changed within a
631 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
632 concept of case for characters whose values are less than 128, so caseless
633 matching is always possible. For characters with higher values, the concept of
634 case is supported if PCRE is compiled with Unicode property support, but not
635 otherwise. If you want to use caseless matching for characters 128 and above,
636 you must ensure that PCRE is compiled with Unicode property support as well as
637 with UTF-8 support.
638 <pre>
640 </pre>
641 If this bit is set, a dollar metacharacter in the pattern matches only at the
642 end of the subject string. Without this option, a dollar also matches
643 immediately before a newline at the end of the string (but not before any other
644 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
645 There is no equivalent to this option in Perl, and no way to set it within a
646 pattern.
647 <pre>
649 </pre>
650 If this bit is set, a dot metacharacter in the pattern matches a character of
651 any value, including one that indicates a newline. However, it only ever
652 matches one character, even if newlines are coded as CRLF. Without this option,
653 a dot does not match when the current position is at a newline. This option is
654 equivalent to Perl's /s option, and it can be changed within a pattern by a
655 (?s) option setting. A negative class such as [^a] always matches newline
656 characters, independent of the setting of this option.
657 <pre>
659 </pre>
660 If this bit is set, names used to identify capturing subpatterns need not be
661 unique. This can be helpful for certain types of pattern when it is known that
662 only one instance of the named subpattern can ever be matched. There are more
663 details of named subpatterns below; see also the
664 <a href="pcrepattern.html"><b>pcrepattern</b></a>
665 documentation.
666 <pre>
668 </pre>
669 If this bit is set, white space data characters in the pattern are totally
670 ignored except when escaped or inside a character class. White space does not
671 include the VT character (code 11). In addition, characters between an
672 unescaped # outside a character class and the next newline, inclusive, are also
673 ignored. This is equivalent to Perl's /x option, and it can be changed within a
674 pattern by a (?x) option setting.
675 </P>
676 <P>
677 Which characters are interpreted as newlines is controlled by the options
678 passed to <b>pcre_compile()</b> or by a special sequence at the start of the
679 pattern, as described in the section entitled
680 <a href="pcrepattern.html#newlines">"Newline conventions"</a>
681 in the <b>pcrepattern</b> documentation. Note that the end of this type of
682 comment is a literal newline sequence in the pattern; escape sequences that
683 happen to represent a newline do not count.
684 </P>
685 <P>
686 This option makes it possible to include comments inside complicated patterns.
687 Note, however, that this applies only to data characters. White space characters
688 may never appear within special character sequences in a pattern, for example
689 within the sequence (?( that introduces a conditional subpattern.
690 <pre>
692 </pre>
693 This option was invented in order to turn on additional functionality of PCRE
694 that is incompatible with Perl, but it is currently of very little use. When
695 set, any backslash in a pattern that is followed by a letter that has no
696 special meaning causes an error, thus reserving these combinations for future
697 expansion. By default, as in Perl, a backslash followed by a letter with no
698 special meaning is treated as a literal. (Perl can, however, be persuaded to
699 give an error for this, by running it with the -w option.) There are at present
700 no other features controlled by this option. It can also be set by a (?X)
701 option setting within a pattern.
702 <pre>
704 </pre>
705 If this option is set, an unanchored pattern is required to match before or at
706 the first newline in the subject string, though the matched text may continue
707 over the newline.
708 <pre>
710 </pre>
711 If this option is set, PCRE's behaviour is changed in some ways so that it is
712 compatible with JavaScript rather than Perl. The changes are as follows:
713 </P>
714 <P>
715 (1) A lone closing square bracket in a pattern causes a compile-time error,
716 because this is illegal in JavaScript (by default it is treated as a data
717 character). Thus, the pattern AB]CD becomes illegal when this option is set.
718 </P>
719 <P>
720 (2) At run time, a back reference to an unset subpattern group matches an empty
721 string (by default this causes the current matching alternative to fail). A
722 pattern such as (\1)(a) succeeds when this option is set (assuming it can find
723 an "a" in the subject), whereas it fails by default, for Perl compatibility.
724 </P>
725 <P>
726 (3) \U matches an upper case "U" character; by default \U causes a compile
727 time error (Perl uses \U to upper case subsequent characters).
728 </P>
729 <P>
730 (4) \u matches a lower case "u" character unless it is followed by four
731 hexadecimal digits, in which case the hexadecimal number defines the code point
732 to match. By default, \u causes a compile time error (Perl uses it to upper
733 case the following character).
734 </P>
735 <P>
736 (5) \x matches a lower case "x" character unless it is followed by two
737 hexadecimal digits, in which case the hexadecimal number defines the code point
738 to match. By default, as in Perl, a hexadecimal number is always expected after
739 \x, but it may have zero, one, or two digits (so, for example, \xz matches a
740 binary zero character followed by z).
741 <pre>
743 </pre>
744 By default, PCRE treats the subject string as consisting of a single line of
745 characters (even if it actually contains newlines). The "start of line"
746 metacharacter (^) matches only at the start of the string, while the "end of
747 line" metacharacter ($) matches only at the end of the string, or before a
748 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
749 Perl.
750 </P>
751 <P>
752 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
753 match immediately following or immediately before internal newlines in the
754 subject string, respectively, as well as at the very start and end. This is
755 equivalent to Perl's /m option, and it can be changed within a pattern by a
756 (?m) option setting. If there are no newlines in a subject string, or no
757 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
758 <pre>
760 </pre>
761 This option locks out interpretation of the pattern as UTF-8 (or UTF-16 or
762 UTF-32 in the 16-bit and 32-bit libraries). In particular, it prevents the
763 creator of the pattern from switching to UTF interpretation by starting the
764 pattern with (*UTF). This may be useful in applications that process patterns
765 from external sources. The combination of PCRE_UTF8 and PCRE_NEVER_UTF also
766 causes an error.
767 <pre>
773 </pre>
774 These options override the default newline definition that was chosen when PCRE
775 was built. Setting the first or the second specifies that a newline is
776 indicated by a single character (CR or LF, respectively). Setting
777 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
778 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
779 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
780 that any Unicode newline sequence should be recognized.
781 </P>
782 <P>
783 In an ASCII/Unicode environment, the Unicode newline sequences are the three
784 just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
785 feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
786 (paragraph separator, U+2029). For the 8-bit library, the last two are
787 recognized only in UTF-8 mode.
788 </P>
789 <P>
790 When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
791 CR is 0x0d, the same as ASCII. However, the character code for LF is normally
792 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
793 not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
794 less than 256. For more details, see the
795 <a href="pcrebuild.html"><b>pcrebuild</b></a>
796 documentation.
797 </P>
798 <P>
799 The newline setting in the options word uses three bits that are treated
800 as a number, giving eight possibilities. Currently only six are used (default
801 plus the five values above). This means that if you set more than one newline
802 option, the combination may or may not be sensible. For example,
804 other combinations may yield unused numbers and cause an error.
805 </P>
806 <P>
807 The only time that a line break in a pattern is specially recognized when
808 compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
809 and so are ignored in this mode. Also, an unescaped # outside a character class
810 indicates a comment that lasts until after the next line break sequence. In
811 other circumstances, line break sequences in patterns are treated as literal
812 data.
813 </P>
814 <P>
815 The newline option that is set at compile time becomes the default that is used
816 for <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, but it can be overridden.
817 <pre>
819 </pre>
820 If this option is set, it disables the use of numbered capturing parentheses in
821 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
822 were followed by ?: but named parentheses can still be used for capturing (and
823 they acquire numbers in the usual way). There is no equivalent of this option
824 in Perl.
825 <pre>
827 </pre>
828 This is an option that acts at matching time; that is, it is really an option
829 for <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b>. If it is set at compile time,
830 it is remembered with the compiled pattern and assumed at matching time. This
831 is necessary if you want to use JIT execution, because the JIT compiler needs
832 to know whether or not this option is set. For details see the discussion of
834 <a href="#execoptions">below.</a>
835 <pre>
837 </pre>
838 This option changes the way PCRE processes \B, \b, \D, \d, \S, \s, \W,
839 \w, and some of the POSIX character classes. By default, only ASCII characters
840 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
841 classify characters. More details are given in the section on
842 <a href="pcre.html#genericchartypes">generic character types</a>
843 in the
844 <a href="pcrepattern.html"><b>pcrepattern</b></a>
845 page. If you set PCRE_UCP, matching one of the items it affects takes much
846 longer. The option is available only if PCRE has been compiled with Unicode
847 property support.
848 <pre>
850 </pre>
851 This option inverts the "greediness" of the quantifiers so that they are not
852 greedy by default, but become greedy if followed by "?". It is not compatible
853 with Perl. It can also be set by a (?U) option setting within the pattern.
854 <pre>
856 </pre>
857 This option causes PCRE to regard both the pattern and the subject as strings
858 of UTF-8 characters instead of single-byte strings. However, it is available
859 only when PCRE is built to include UTF support. If not, the use of this option
860 provokes an error. Details of how this option changes the behaviour of PCRE are
861 given in the
862 <a href="pcreunicode.html"><b>pcreunicode</b></a>
863 page.
864 <pre>
866 </pre>
867 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
868 automatically checked. There is a discussion about the
869 <a href="pcreunicode.html#utf8strings">validity of UTF-8 strings</a>
870 in the
871 <a href="pcreunicode.html"><b>pcreunicode</b></a>
872 page. If an invalid UTF-8 sequence is found, <b>pcre_compile()</b> returns an
873 error. If you already know that your pattern is valid, and you want to skip
874 this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
875 When it is set, the effect of passing an invalid UTF-8 string as a pattern is
876 undefined. It may cause your program to crash. Note that this option can also
877 be passed to <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, to suppress the
878 validity checking of subject strings only. If the same string is being matched
879 many times, the option can be safely set for the second and subsequent
880 matchings to improve performance.
881 </P>
882 <br><a name="SEC12" href="#TOC1">COMPILATION ERROR CODES</a><br>
883 <P>
884 The following table lists the error codes than may be returned by
885 <b>pcre_compile2()</b>, along with the error messages that may be returned by
886 both compiling functions. Note that error messages are always 8-bit ASCII
887 strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
888 have fallen out of use. To avoid confusion, they have not been re-used.
889 <pre>
890 0 no error
891 1 \ at end of pattern
892 2 \c at end of pattern
893 3 unrecognized character follows \
894 4 numbers out of order in {} quantifier
895 5 number too big in {} quantifier
896 6 missing terminating ] for character class
897 7 invalid escape sequence in character class
898 8 range out of order in character class
899 9 nothing to repeat
900 10 [this code is not in use]
901 11 internal error: unexpected repeat
902 12 unrecognized character after (? or (?-
903 13 POSIX named classes are supported only within a class
904 14 missing )
905 15 reference to non-existent subpattern
906 16 erroffset passed as NULL
907 17 unknown option bit(s) set
908 18 missing ) after comment
909 19 [this code is not in use]
910 20 regular expression is too large
911 21 failed to get memory
912 22 unmatched parentheses
913 23 internal error: code overflow
914 24 unrecognized character after (?&#60;
915 25 lookbehind assertion is not fixed length
916 26 malformed number or name after (?(
917 27 conditional group contains more than two branches
918 28 assertion expected after (?(
919 29 (?R or (?[+-]digits must be followed by )
920 30 unknown POSIX class name
921 31 POSIX collating elements are not supported
922 32 this version of PCRE is compiled without UTF support
923 33 [this code is not in use]
924 34 character value in \x{...} sequence is too large
925 35 invalid condition (?(0)
926 36 \C not allowed in lookbehind assertion
927 37 PCRE does not support \L, \l, \N{name}, \U, or \u
928 38 number after (?C is &#62; 255
929 39 closing ) for (?C expected
930 40 recursive call could loop indefinitely
931 41 unrecognized character after (?P
932 42 syntax error in subpattern name (missing terminator)
933 43 two named subpatterns have the same name
934 44 invalid UTF-8 string (specifically UTF-8)
935 45 support for \P, \p, and \X has not been compiled
936 46 malformed \P or \p sequence
937 47 unknown property name after \P or \p
938 48 subpattern name is too long (maximum 32 characters)
939 49 too many named subpatterns (maximum 10000)
940 50 [this code is not in use]
941 51 octal value is greater than \377 in 8-bit non-UTF-8 mode
942 52 internal error: overran compiling workspace
943 53 internal error: previously-checked referenced subpattern
944 not found
945 54 DEFINE group contains more than one branch
946 55 repeating a DEFINE group is not allowed
947 56 inconsistent NEWLINE options
948 57 \g is not followed by a braced, angle-bracketed, or quoted
949 name/number or by a plain number
950 58 a numbered reference must not be zero
951 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
952 60 (*VERB) not recognized or malformed
953 61 number is too big
954 62 subpattern name expected
955 63 digit expected after (?+
956 64 ] is an invalid data character in JavaScript compatibility mode
957 65 different names for subpatterns of the same number are
958 not allowed
959 66 (*MARK) must have an argument
960 67 this version of PCRE is not compiled with Unicode property
961 support
962 68 \c must be followed by an ASCII character
963 69 \k is not followed by a braced, angle-bracketed, or quoted name
964 70 internal error: unknown opcode in find_fixedlength()
965 71 \N is not supported in a class
966 72 too many forward references
967 73 disallowed Unicode code point (&#62;= 0xd800 && &#60;= 0xdfff)
968 74 invalid UTF-16 string (specifically UTF-16)
969 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
970 76 character value in \u.... sequence is too large
971 77 invalid UTF-32 string (specifically UTF-32)
972 </pre>
973 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
974 be used if the limits were changed when PCRE was built.
975 <a name="studyingapattern"></a></P>
976 <br><a name="SEC13" href="#TOC1">STUDYING A PATTERN</a><br>
977 <P>
978 <b>pcre_extra *pcre_study(const pcre *<i>code</i>, int <i>options</i></b>
979 <b>const char **<i>errptr</i>);</b>
980 </P>
981 <P>
982 If a compiled pattern is going to be used several times, it is worth spending
983 more time analyzing it in order to speed up the time taken for matching. The
984 function <b>pcre_study()</b> takes a pointer to a compiled pattern as its first
985 argument. If studying the pattern produces additional information that will
986 help speed up matching, <b>pcre_study()</b> returns a pointer to a
987 <b>pcre_extra</b> block, in which the <i>study_data</i> field points to the
988 results of the study.
989 </P>
990 <P>
991 The returned value from <b>pcre_study()</b> can be passed directly to
992 <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b>. However, a <b>pcre_extra</b> block
993 also contains other fields that can be set by the caller before the block is
994 passed; these are described
995 <a href="#extradata">below</a>
996 in the section on matching a pattern.
997 </P>
998 <P>
999 If studying the pattern does not produce any useful information,
1000 <b>pcre_study()</b> returns NULL by default. In that circumstance, if the
1001 calling program wants to pass any of the other fields to <b>pcre_exec()</b> or
1002 <b>pcre_dfa_exec()</b>, it must set up its own <b>pcre_extra</b> block. However,
1003 if <b>pcre_study()</b> is called with the PCRE_STUDY_EXTRA_NEEDED option, it
1004 returns a <b>pcre_extra</b> block even if studying did not find any additional
1005 information. It may still return NULL, however, if an error occurs in
1006 <b>pcre_study()</b>.
1007 </P>
1008 <P>
1009 The second argument of <b>pcre_study()</b> contains option bits. There are three
1010 further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1011 <pre>
1015 </pre>
1016 If any of these are set, and the just-in-time compiler is available, the
1017 pattern is further compiled into machine code that executes much faster than
1018 the <b>pcre_exec()</b> interpretive matching function. If the just-in-time
1019 compiler is not available, these options are ignored. All undefined bits in the
1020 <i>options</i> argument must be zero.
1021 </P>
1022 <P>
1023 JIT compilation is a heavyweight optimization. It can take some time for
1024 patterns to be analyzed, and for one-off matches and simple patterns the
1025 benefit of faster execution might be offset by a much slower study time.
1026 Not all patterns can be optimized by the JIT compiler. For those that cannot be
1027 handled, matching automatically falls back to the <b>pcre_exec()</b>
1028 interpreter. For more details, see the
1029 <a href="pcrejit.html"><b>pcrejit</b></a>
1030 documentation.
1031 </P>
1032 <P>
1033 The third argument for <b>pcre_study()</b> is a pointer for an error message. If
1034 studying succeeds (even if no data is returned), the variable it points to is
1035 set to NULL. Otherwise it is set to point to a textual error message. This is a
1036 static string that is part of the library. You must not try to free it. You
1037 should test the error pointer for NULL after calling <b>pcre_study()</b>, to be
1038 sure that it has run successfully.
1039 </P>
1040 <P>
1041 When you are finished with a pattern, you can free the memory used for the
1042 study data by calling <b>pcre_free_study()</b>. This function was added to the
1043 API for release 8.20. For earlier versions, the memory could be freed with
1044 <b>pcre_free()</b>, just like the pattern itself. This will still work in cases
1045 where JIT optimization is not used, but it is advisable to change to the new
1046 function when convenient.
1047 </P>
1048 <P>
1049 This is a typical way in which <b>pcre_study</b>() is used (except that in a
1050 real application there should be tests for errors):
1051 <pre>
1052 int rc;
1053 pcre *re;
1054 pcre_extra *sd;
1055 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1056 sd = pcre_study(
1057 re, /* result of pcre_compile() */
1058 0, /* no options */
1059 &error); /* set to NULL or points to a message */
1060 rc = pcre_exec( /* see below for details of pcre_exec() options */
1061 re, sd, "subject", 7, 0, 0, ovector, 30);
1062 ...
1063 pcre_free_study(sd);
1064 pcre_free(re);
1065 </pre>
1066 Studying a pattern does two things: first, a lower bound for the length of
1067 subject string that is needed to match the pattern is computed. This does not
1068 mean that there are any strings of that length that match, but it does
1069 guarantee that no shorter strings match. The value is used to avoid wasting
1070 time by trying to match strings that are shorter than the lower bound. You can
1071 find out the value in a calling program via the <b>pcre_fullinfo()</b> function.
1072 </P>
1073 <P>
1074 Studying a pattern is also useful for non-anchored patterns that do not have a
1075 single fixed starting character. A bitmap of possible starting bytes is
1076 created. This speeds up finding a position in the subject at which to start
1077 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1078 In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1079 </P>
1080 <P>
1081 These two optimizations apply to both <b>pcre_exec()</b> and
1082 <b>pcre_dfa_exec()</b>, and the information is also used by the JIT compiler.
1083 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option.
1084 You might want to do this if your pattern contains callouts or (*MARK) and you
1085 want to make use of these facilities in cases where matching fails.
1086 </P>
1087 <P>
1088 PCRE_NO_START_OPTIMIZE can be specified at either compile time or execution
1089 time. However, if PCRE_NO_START_OPTIMIZE is passed to <b>pcre_exec()</b>, (that
1090 is, after any JIT compilation has happened) JIT execution is disabled. For JIT
1091 execution to work with PCRE_NO_START_OPTIMIZE, the option must be set at
1092 compile time.
1093 </P>
1094 <P>
1095 There is a longer discussion of PCRE_NO_START_OPTIMIZE
1096 <a href="#execoptions">below.</a>
1097 <a name="localesupport"></a></P>
1098 <br><a name="SEC14" href="#TOC1">LOCALE SUPPORT</a><br>
1099 <P>
1100 PCRE handles caseless matching, and determines whether characters are letters,
1101 digits, or whatever, by reference to a set of tables, indexed by character
1102 value. When running in UTF-8 mode, this applies only to characters
1103 with codes less than 128. By default, higher-valued codes never match escapes
1104 such as \w or \d, but they can be tested with \p if PCRE is built with
1105 Unicode character property support. Alternatively, the PCRE_UCP option can be
1106 set at compile time; this causes \w and friends to use Unicode property
1107 support instead of built-in tables. The use of locales with Unicode is
1108 discouraged. If you are handling characters with codes greater than 128, you
1109 should either use UTF-8 and Unicode, or use locales, but not try to mix the
1110 two.
1111 </P>
1112 <P>
1113 PCRE contains an internal set of tables that are used when the final argument
1114 of <b>pcre_compile()</b> is NULL. These are sufficient for many applications.
1115 Normally, the internal tables recognize only ASCII characters. However, when
1116 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1117 default "C" locale of the local system, which may cause them to be different.
1118 </P>
1119 <P>
1120 The internal tables can always be overridden by tables supplied by the
1121 application that calls PCRE. These may be created in a different locale from
1122 the default. As more and more applications change to using Unicode, the need
1123 for this locale support is expected to die away.
1124 </P>
1125 <P>
1126 External tables are built by calling the <b>pcre_maketables()</b> function,
1127 which has no arguments, in the relevant locale. The result can then be passed
1128 to <b>pcre_compile()</b> or <b>pcre_exec()</b> as often as necessary. For
1129 example, to build and use tables that are appropriate for the French locale
1130 (where accented characters with values greater than 128 are treated as letters),
1131 the following code could be used:
1132 <pre>
1133 setlocale(LC_CTYPE, "fr_FR");
1134 tables = pcre_maketables();
1135 re = pcre_compile(..., tables);
1136 </pre>
1137 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1138 are using Windows, the name for the French locale is "french".
1139 </P>
1140 <P>
1141 When <b>pcre_maketables()</b> runs, the tables are built in memory that is
1142 obtained via <b>pcre_malloc</b>. It is the caller's responsibility to ensure
1143 that the memory containing the tables remains available for as long as it is
1144 needed.
1145 </P>
1146 <P>
1147 The pointer that is passed to <b>pcre_compile()</b> is saved with the compiled
1148 pattern, and the same tables are used via this pointer by <b>pcre_study()</b>
1149 and normally also by <b>pcre_exec()</b>. Thus, by default, for any single
1150 pattern, compilation, studying and matching all happen in the same locale, but
1151 different patterns can be compiled in different locales.
1152 </P>
1153 <P>
1154 It is possible to pass a table pointer or NULL (indicating the use of the
1155 internal tables) to <b>pcre_exec()</b>. Although not intended for this purpose,
1156 this facility could be used to match a pattern in a different locale from the
1157 one in which it was compiled. Passing table pointers at run time is discussed
1158 below in the section on matching a pattern.
1159 <a name="infoaboutpattern"></a></P>
1160 <br><a name="SEC15" href="#TOC1">INFORMATION ABOUT A PATTERN</a><br>
1161 <P>
1162 <b>int pcre_fullinfo(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
1163 <b>int <i>what</i>, void *<i>where</i>);</b>
1164 </P>
1165 <P>
1166 The <b>pcre_fullinfo()</b> function returns information about a compiled
1167 pattern. It replaces the <b>pcre_info()</b> function, which was removed from the
1168 library at version 8.30, after more than 10 years of obsolescence.
1169 </P>
1170 <P>
1171 The first argument for <b>pcre_fullinfo()</b> is a pointer to the compiled
1172 pattern. The second argument is the result of <b>pcre_study()</b>, or NULL if
1173 the pattern was not studied. The third argument specifies which piece of
1174 information is required, and the fourth argument is a pointer to a variable
1175 to receive the data. The yield of the function is zero for success, or one of
1176 the following negative numbers:
1177 <pre>
1178 PCRE_ERROR_NULL the argument <i>code</i> was NULL
1179 the argument <i>where</i> was NULL
1180 PCRE_ERROR_BADMAGIC the "magic number" was not found
1181 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1182 endianness
1183 PCRE_ERROR_BADOPTION the value of <i>what</i> was invalid
1184 PCRE_ERROR_UNSET the requested field is not set
1185 </pre>
1186 The "magic number" is placed at the start of each compiled pattern as an simple
1187 check against passing an arbitrary memory pointer. The endianness error can
1188 occur if a compiled pattern is saved and reloaded on a different host. Here is
1189 a typical call of <b>pcre_fullinfo()</b>, to obtain the length of the compiled
1190 pattern:
1191 <pre>
1192 int rc;
1193 size_t length;
1194 rc = pcre_fullinfo(
1195 re, /* result of pcre_compile() */
1196 sd, /* result of pcre_study(), or NULL */
1197 PCRE_INFO_SIZE, /* what is required */
1198 &length); /* where to put the data */
1199 </pre>
1200 The possible values for the third argument are defined in <b>pcre.h</b>, and are
1201 as follows:
1202 <pre>
1204 </pre>
1205 Return the number of the highest back reference in the pattern. The fourth
1206 argument should point to an <b>int</b> variable. Zero is returned if there are
1207 no back references.
1208 <pre>
1210 </pre>
1211 Return the number of capturing subpatterns in the pattern. The fourth argument
1212 should point to an <b>int</b> variable.
1213 <pre>
1215 </pre>
1216 Return a pointer to the internal default character tables within PCRE. The
1217 fourth argument should point to an <b>unsigned char *</b> variable. This
1218 information call is provided for internal use by the <b>pcre_study()</b>
1219 function. External callers can cause PCRE to use its internal tables by passing
1220 a NULL table pointer.
1221 <pre>
1223 </pre>
1224 Return information about the first data unit of any matched string, for a
1225 non-anchored pattern. (The name of this option refers to the 8-bit library,
1226 where data units are bytes.) The fourth argument should point to an <b>int</b>
1227 variable.
1228 </P>
1229 <P>
1230 If there is a fixed first value, for example, the letter "c" from a pattern
1231 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1232 value is always less than 256. In the 16-bit library the value can be up to
1233 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1234 </P>
1235 <P>
1236 If there is no fixed first value, and if either
1237 <br>
1238 <br>
1239 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1240 starts with "^", or
1241 <br>
1242 <br>
1243 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1244 (if it were set, the pattern would be anchored),
1245 <br>
1246 <br>
1247 -1 is returned, indicating that the pattern matches only at the start of a
1248 subject string or after any newline within the string. Otherwise -2 is
1249 returned. For anchored patterns, -2 is returned.
1250 </P>
1251 <P>
1252 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1253 to return the full 32-bit range of the character, this value is deprecated;
1255 should be used.
1256 <pre>
1258 </pre>
1259 If the pattern was studied, and this resulted in the construction of a 256-bit
1260 table indicating a fixed set of values for the first data unit in any matching
1261 string, a pointer to the table is returned. Otherwise NULL is returned. The
1262 fourth argument should point to an <b>unsigned char *</b> variable.
1263 <pre>
1265 </pre>
1266 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1267 otherwise 0. The fourth argument should point to an <b>int</b> variable. An
1268 explicit match is either a literal CR or LF character, or \r or \n.
1269 <pre>
1271 </pre>
1272 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1273 0. The fourth argument should point to an <b>int</b> variable. (?J) and
1274 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1275 <pre>
1277 </pre>
1278 Return 1 if the pattern was studied with one of the JIT options, and
1279 just-in-time compiling was successful. The fourth argument should point to an
1280 <b>int</b> variable. A return value of 0 means that JIT support is not available
1281 in this version of PCRE, or that the pattern was not studied with a JIT option,
1282 or that the JIT compiler could not handle this particular pattern. See the
1283 <a href="pcrejit.html"><b>pcrejit</b></a>
1284 documentation for details of what can and cannot be handled.
1285 <pre>
1287 </pre>
1288 If the pattern was successfully studied with a JIT option, return the size of
1289 the JIT compiled code, otherwise return zero. The fourth argument should point
1290 to a <b>size_t</b> variable.
1291 <pre>
1293 </pre>
1294 Return the value of the rightmost literal data unit that must exist in any
1295 matched string, other than at its start, if such a value has been recorded. The
1296 fourth argument should point to an <b>int</b> variable. If there is no such
1297 value, -1 is returned. For anchored patterns, a last literal value is recorded
1298 only if it follows something of variable length. For example, for the pattern
1299 /^a\d+z\d+/ the returned value is "z", but for /^a\dz\d/ the returned value
1300 is -1.
1301 </P>
1302 <P>
1303 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1304 to return the full 32-bit range of the character, this value is deprecated;
1306 be used.
1307 <pre>
1309 </pre>
1310 If the pattern set a match limit by including an item of the form
1311 (*LIMIT_MATCH=nnnn) at the start, the value is returned. The fourth argument
1312 should point to an unsigned 32-bit integer. If no such value has been set, the
1313 call to <b>pcre_fullinfo()</b> returns the error PCRE_ERROR_UNSET.
1314 <pre>
1316 </pre>
1317 Return the number of characters (NB not bytes) in the longest lookbehind
1318 assertion in the pattern. This information is useful when doing multi-segment
1319 matching using the partial matching facilities. Note that the simple assertions
1320 \b and \B require a one-character lookbehind. \A also registers a
1321 one-character lookbehind, though it does not actually inspect the previous
1322 character. This is to ensure that at least one character from the old segment
1323 is retained when a new segment is processed. Otherwise, if there are no
1324 lookbehinds in the pattern, \A might match incorrectly at the start of a new
1325 segment.
1326 <pre>
1328 </pre>
1329 If the pattern was studied and a minimum length for matching subject strings
1330 was computed, its value is returned. Otherwise the returned value is -1. The
1331 value is a number of characters, which in UTF-8 mode may be different from the
1332 number of bytes. The fourth argument should point to an <b>int</b> variable. A
1333 non-negative value is a lower bound to the length of any matching string. There
1334 may not be any strings of that length that do actually match, but every string
1335 that does match is at least that long.
1336 <pre>
1340 </pre>
1341 PCRE supports the use of named as well as numbered capturing parentheses. The
1342 names are just an additional way of identifying the parentheses, which still
1343 acquire numbers. Several convenience functions such as
1344 <b>pcre_get_named_substring()</b> are provided for extracting captured
1345 substrings by name. It is also possible to extract the data directly, by first
1346 converting the name to a number in order to access the correct pointers in the
1347 output vector (described with <b>pcre_exec()</b> below). To do the conversion,
1348 you need to use the name-to-number map, which is described by these three
1349 values.
1350 </P>
1351 <P>
1352 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1353 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1354 entry; both of these return an <b>int</b> value. The entry size depends on the
1355 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1356 entry of the table. This is a pointer to <b>char</b> in the 8-bit library, where
1357 the first two bytes of each entry are the number of the capturing parenthesis,
1358 most significant byte first. In the 16-bit library, the pointer points to
1359 16-bit data units, the first of which contains the parenthesis number.
1360 In the 32-bit library, the pointer points to 32-bit data units, the first of
1361 which contains the parenthesis number. The rest
1362 of the entry is the corresponding name, zero terminated.
1363 </P>
1364 <P>
1365 The names are in alphabetical order. Duplicate names may appear if (?| is used
1366 to create multiple groups with the same number, as described in the
1367 <a href="pcrepattern.html#dupsubpatternnumber">section on duplicate subpattern numbers</a>
1368 in the
1369 <a href="pcrepattern.html"><b>pcrepattern</b></a>
1370 page. Duplicate names for subpatterns with different numbers are permitted only
1371 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1372 table in the order in which they were found in the pattern. In the absence of
1373 (?| this is the order of increasing number; when (?| is used this is not
1374 necessarily the case because later subpatterns may have lower numbers.
1375 </P>
1376 <P>
1377 As a simple example of the name/number table, consider the following pattern
1378 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1379 space - including newlines - is ignored):
1380 <pre>
1381 (?&#60;date&#62; (?&#60;year&#62;(\d\d)?\d\d) - (?&#60;month&#62;\d\d) - (?&#60;day&#62;\d\d) )
1382 </pre>
1383 There are four named subpatterns, so the table has four entries, and each entry
1384 in the table is eight bytes long. The table is as follows, with non-printing
1385 bytes shows in hexadecimal, and undefined bytes shown as ??:
1386 <pre>
1387 00 01 d a t e 00 ??
1388 00 05 d a y 00 ?? ??
1389 00 04 m o n t h 00
1390 00 02 y e a r 00 ??
1391 </pre>
1392 When writing code to extract data from named subpatterns using the
1393 name-to-number map, remember that the length of the entries is likely to be
1394 different for each compiled pattern.
1395 <pre>
1397 </pre>
1398 Return 1 if the pattern can be used for partial matching with
1399 <b>pcre_exec()</b>, otherwise 0. The fourth argument should point to an
1400 <b>int</b> variable. From release 8.00, this always returns 1, because the
1401 restrictions that previously applied to partial matching have been lifted. The
1402 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1403 documentation gives details of partial matching.
1404 <pre>
1406 </pre>
1407 Return a copy of the options with which the pattern was compiled. The fourth
1408 argument should point to an <b>unsigned long int</b> variable. These option bits
1409 are those specified in the call to <b>pcre_compile()</b>, modified by any
1410 top-level option settings at the start of the pattern itself. In other words,
1411 they are the options that will be in force when matching starts. For example,
1412 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1414 </P>
1415 <P>
1416 A pattern is automatically anchored by PCRE if all of its top-level
1417 alternatives begin with one of the following:
1418 <pre>
1419 ^ unless PCRE_MULTILINE is set
1420 \A always
1421 \G always
1422 .* if PCRE_DOTALL is set and there are no back references to the subpattern in which .* appears
1423 </pre>
1424 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1425 <b>pcre_fullinfo()</b>.
1426 <pre>
1428 </pre>
1429 If the pattern set a recursion limit by including an item of the form
1430 (*LIMIT_RECURSION=nnnn) at the start, the value is returned. The fourth
1431 argument should point to an unsigned 32-bit integer. If no such value has been
1432 set, the call to <b>pcre_fullinfo()</b> returns the error PCRE_ERROR_UNSET.
1433 <pre>
1435 </pre>
1436 Return the size of the compiled pattern in bytes (for both libraries). The
1437 fourth argument should point to a <b>size_t</b> variable. This value does not
1438 include the size of the <b>pcre</b> structure that is returned by
1439 <b>pcre_compile()</b>. The value that is passed as the argument to
1440 <b>pcre_malloc()</b> when <b>pcre_compile()</b> is getting memory in which to
1441 place the compiled data is the value returned by this option plus the size of
1442 the <b>pcre</b> structure. Studying a compiled pattern, with or without JIT,
1443 does not alter the value returned by this option.
1444 <pre>
1446 </pre>
1447 Return the size in bytes of the data block pointed to by the <i>study_data</i>
1448 field in a <b>pcre_extra</b> block. If <b>pcre_extra</b> is NULL, or there is no
1449 study data, zero is returned. The fourth argument should point to a
1450 <b>size_t</b> variable. The <i>study_data</i> field is set by <b>pcre_study()</b>
1451 to record information that will speed up matching (see the section entitled
1452 <a href="#studyingapattern">"Studying a pattern"</a>
1453 above). The format of the <i>study_data</i> block is private, but its length
1454 is made available via this option so that it can be saved and restored (see the
1455 <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1456 documentation for details).
1457 <pre>
1459 </pre>
1460 Return information about the first data unit of any matched string, for a
1461 non-anchored pattern. The fourth argument should point to an <b>int</b>
1462 variable.
1463 </P>
1464 <P>
1465 If there is a fixed first value, for example, the letter "c" from a pattern
1466 such as (cat|cow|coyote), 1 is returned, and the character value can be
1467 retrieved using PCRE_INFO_FIRSTCHARACTER.
1468 </P>
1469 <P>
1470 If there is no fixed first value, and if either
1471 <br>
1472 <br>
1473 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1474 starts with "^", or
1475 <br>
1476 <br>
1477 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1478 (if it were set, the pattern would be anchored),
1479 <br>
1480 <br>
1481 2 is returned, indicating that the pattern matches only at the start of a
1482 subject string or after any newline within the string. Otherwise 0 is
1483 returned. For anchored patterns, 0 is returned.
1484 <pre>
1486 </pre>
1487 Return the fixed first character value, if PCRE_INFO_FIRSTCHARACTERFLAGS
1488 returned 1; otherwise returns 0. The fourth argument should point to an
1489 <b>uint_t</b> variable.
1490 </P>
1491 <P>
1492 In the 8-bit library, the value is always less than 256. In the 16-bit library
1493 the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1494 can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1495 </P>
1496 <P>
1497 If there is no fixed first value, and if either
1498 <br>
1499 <br>
1500 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1501 starts with "^", or
1502 <br>
1503 <br>
1504 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1505 (if it were set, the pattern would be anchored),
1506 <br>
1507 <br>
1508 -1 is returned, indicating that the pattern matches only at the start of a
1509 subject string or after any newline within the string. Otherwise -2 is
1510 returned. For anchored patterns, -2 is returned.
1511 <pre>
1513 </pre>
1514 Returns 1 if there is a rightmost literal data unit that must exist in any
1515 matched string, other than at its start. The fourth argument should point to
1516 an <b>int</b> variable. If there is no such value, 0 is returned. If returning
1517 1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1518 </P>
1519 <P>
1520 For anchored patterns, a last literal value is recorded only if it follows
1521 something of variable length. For example, for the pattern /^a\d+z\d+/ the
1522 returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1523 /^a\dz\d/ the returned value is 0.
1524 <pre>
1526 </pre>
1527 Return the value of the rightmost literal data unit that must exist in any
1528 matched string, other than at its start, if such a value has been recorded. The
1529 fourth argument should point to an <b>uint32_t</b> variable. If there is no such
1530 value, 0 is returned.
1531 </P>
1532 <br><a name="SEC16" href="#TOC1">REFERENCE COUNTS</a><br>
1533 <P>
1534 <b>int pcre_refcount(pcre *<i>code</i>, int <i>adjust</i>);</b>
1535 </P>
1536 <P>
1537 The <b>pcre_refcount()</b> function is used to maintain a reference count in the
1538 data block that contains a compiled pattern. It is provided for the benefit of
1539 applications that operate in an object-oriented manner, where different parts
1540 of the application may be using the same compiled pattern, but you want to free
1541 the block when they are all done.
1542 </P>
1543 <P>
1544 When a pattern is compiled, the reference count field is initialized to zero.
1545 It is changed only by calling this function, whose action is to add the
1546 <i>adjust</i> value (which may be positive or negative) to it. The yield of the
1547 function is the new value. However, the value of the count is constrained to
1548 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1549 it is forced to the appropriate limit value.
1550 </P>
1551 <P>
1552 Except when it is zero, the reference count is not correctly preserved if a
1553 pattern is compiled on one host and then transferred to a host whose byte-order
1554 is different. (This seems a highly unlikely scenario.)
1555 </P>
1556 <br><a name="SEC17" href="#TOC1">MATCHING A PATTERN: THE TRADITIONAL FUNCTION</a><br>
1557 <P>
1558 <b>int pcre_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
1559 <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
1560 <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>);</b>
1561 </P>
1562 <P>
1563 The function <b>pcre_exec()</b> is called to match a subject string against a
1564 compiled pattern, which is passed in the <i>code</i> argument. If the
1565 pattern was studied, the result of the study should be passed in the
1566 <i>extra</i> argument. You can call <b>pcre_exec()</b> with the same <i>code</i>
1567 and <i>extra</i> arguments as many times as you like, in order to match
1568 different subject strings with the same pattern.
1569 </P>
1570 <P>
1571 This function is the main matching facility of the library, and it operates in
1572 a Perl-like manner. For specialist use there is also an alternative matching
1573 function, which is described
1574 <a href="#dfamatch">below</a>
1575 in the section about the <b>pcre_dfa_exec()</b> function.
1576 </P>
1577 <P>
1578 In most applications, the pattern will have been compiled (and optionally
1579 studied) in the same process that calls <b>pcre_exec()</b>. However, it is
1580 possible to save compiled patterns and study data, and then use them later
1581 in different processes, possibly even on different hosts. For a discussion
1582 about this, see the
1583 <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1584 documentation.
1585 </P>
1586 <P>
1587 Here is an example of a simple call to <b>pcre_exec()</b>:
1588 <pre>
1589 int rc;
1590 int ovector[30];
1591 rc = pcre_exec(
1592 re, /* result of pcre_compile() */
1593 NULL, /* we didn't study the pattern */
1594 "some string", /* the subject string */
1595 11, /* the length of the subject string */
1596 0, /* start at offset 0 in the subject */
1597 0, /* default options */
1598 ovector, /* vector of integers for substring information */
1599 30); /* number of elements (NOT size in bytes) */
1600 <a name="extradata"></a></PRE>
1601 </P>
1602 <br><b>
1603 Extra data for <b>pcre_exec()</b>
1604 </b><br>
1605 <P>
1606 If the <i>extra</i> argument is not NULL, it must point to a <b>pcre_extra</b>
1607 data block. The <b>pcre_study()</b> function returns such a block (when it
1608 doesn't return NULL), but you can also create one for yourself, and pass
1609 additional information in it. The <b>pcre_extra</b> block contains the following
1610 fields (not necessarily in this order):
1611 <pre>
1612 unsigned long int <i>flags</i>;
1613 void *<i>study_data</i>;
1614 void *<i>executable_jit</i>;
1615 unsigned long int <i>match_limit</i>;
1616 unsigned long int <i>match_limit_recursion</i>;
1617 void *<i>callout_data</i>;
1618 const unsigned char *<i>tables</i>;
1619 unsigned char **<i>mark</i>;
1620 </pre>
1621 In the 16-bit version of this structure, the <i>mark</i> field has type
1622 "PCRE_UCHAR16 **".
1623 <br>
1624 <br>
1625 In the 32-bit version of this structure, the <i>mark</i> field has type
1626 "PCRE_UCHAR32 **".
1627 </P>
1628 <P>
1629 The <i>flags</i> field is used to specify which of the other fields are set. The
1630 flag bits are:
1631 <pre>
1639 </pre>
1640 Other flag bits should be set to zero. The <i>study_data</i> field and sometimes
1641 the <i>executable_jit</i> field are set in the <b>pcre_extra</b> block that is
1642 returned by <b>pcre_study()</b>, together with the appropriate flag bits. You
1643 should not set these yourself, but you may add to the block by setting other
1644 fields and their corresponding flag bits.
1645 </P>
1646 <P>
1647 The <i>match_limit</i> field provides a means of preventing PCRE from using up a
1648 vast amount of resources when running patterns that are not going to match,
1649 but which have a very large number of possibilities in their search trees. The
1650 classic example is a pattern that uses nested unlimited repeats.
1651 </P>
1652 <P>
1653 Internally, <b>pcre_exec()</b> uses a function called <b>match()</b>, which it
1654 calls repeatedly (sometimes recursively). The limit set by <i>match_limit</i> is
1655 imposed on the number of times this function is called during a match, which
1656 has the effect of limiting the amount of backtracking that can take place. For
1657 patterns that are not anchored, the count restarts from zero for each position
1658 in the subject string.
1659 </P>
1660 <P>
1661 When <b>pcre_exec()</b> is called with a pattern that was successfully studied
1662 with a JIT option, the way that the matching is executed is entirely different.
1663 However, there is still the possibility of runaway matching that goes on for a
1664 very long time, and so the <i>match_limit</i> value is also used in this case
1665 (but in a different way) to limit how long the matching can continue.
1666 </P>
1667 <P>
1668 The default value for the limit can be set when PCRE is built; the default
1669 default is 10 million, which handles all but the most extreme cases. You can
1670 override the default by suppling <b>pcre_exec()</b> with a <b>pcre_extra</b>
1671 block in which <i>match_limit</i> is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1672 the <i>flags</i> field. If the limit is exceeded, <b>pcre_exec()</b> returns
1674 </P>
1675 <P>
1676 A value for the match limit may also be supplied by an item at the start of a
1677 pattern of the form
1678 <pre>
1679 (*LIMIT_MATCH=d)
1680 </pre>
1681 where d is a decimal number. However, such a setting is ignored unless d is
1682 less than the limit set by the caller of <b>pcre_exec()</b> or, if no such limit
1683 is set, less than the default.
1684 </P>
1685 <P>
1686 The <i>match_limit_recursion</i> field is similar to <i>match_limit</i>, but
1687 instead of limiting the total number of times that <b>match()</b> is called, it
1688 limits the depth of recursion. The recursion depth is a smaller number than the
1689 total number of calls, because not all calls to <b>match()</b> are recursive.
1690 This limit is of use only if it is set smaller than <i>match_limit</i>.
1691 </P>
1692 <P>
1693 Limiting the recursion depth limits the amount of machine stack that can be
1694 used, or, when PCRE has been compiled to use memory on the heap instead of the
1695 stack, the amount of heap memory that can be used. This limit is not relevant,
1696 and is ignored, when matching is done using JIT compiled code.
1697 </P>
1698 <P>
1699 The default value for <i>match_limit_recursion</i> can be set when PCRE is
1700 built; the default default is the same value as the default for
1701 <i>match_limit</i>. You can override the default by suppling <b>pcre_exec()</b>
1702 with a <b>pcre_extra</b> block in which <i>match_limit_recursion</i> is set, and
1703 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the <i>flags</i> field. If the limit
1704 is exceeded, <b>pcre_exec()</b> returns PCRE_ERROR_RECURSIONLIMIT.
1705 </P>
1706 <P>
1707 A value for the recursion limit may also be supplied by an item at the start of
1708 a pattern of the form
1709 <pre>
1711 </pre>
1712 where d is a decimal number. However, such a setting is ignored unless d is
1713 less than the limit set by the caller of <b>pcre_exec()</b> or, if no such limit
1714 is set, less than the default.
1715 </P>
1716 <P>
1717 The <i>callout_data</i> field is used in conjunction with the "callout" feature,
1718 and is described in the
1719 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1720 documentation.
1721 </P>
1722 <P>
1723 The <i>tables</i> field is used to pass a character tables pointer to
1724 <b>pcre_exec()</b>; this overrides the value that is stored with the compiled
1725 pattern. A non-NULL value is stored with the compiled pattern only if custom
1726 tables were supplied to <b>pcre_compile()</b> via its <i>tableptr</i> argument.
1727 If NULL is passed to <b>pcre_exec()</b> using this mechanism, it forces PCRE's
1728 internal tables to be used. This facility is helpful when re-using patterns
1729 that have been saved after compiling with an external set of tables, because
1730 the external tables might be at a different address when <b>pcre_exec()</b> is
1731 called. See the
1732 <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1733 documentation for a discussion of saving compiled patterns for later use.
1734 </P>
1735 <P>
1736 If PCRE_EXTRA_MARK is set in the <i>flags</i> field, the <i>mark</i> field must
1737 be set to point to a suitable variable. If the pattern contains any
1738 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1739 a name to pass back, a pointer to the name string (zero terminated) is placed
1740 in the variable pointed to by the <i>mark</i> field. The names are within the
1741 compiled pattern; if you wish to retain such a name you must copy it before
1742 freeing the memory of a compiled pattern. If there is no name to pass back, the
1743 variable pointed to by the <i>mark</i> field is set to NULL. For details of the
1744 backtracking control verbs, see the section entitled
1745 <a href="pcrepattern#backtrackcontrol">"Backtracking control"</a>
1746 in the
1747 <a href="pcrepattern.html"><b>pcrepattern</b></a>
1748 documentation.
1749 <a name="execoptions"></a></P>
1750 <br><b>
1751 Option bits for <b>pcre_exec()</b>
1752 </b><br>
1753 <P>
1754 The unused bits of the <i>options</i> argument for <b>pcre_exec()</b> must be
1755 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
1759 </P>
1760 <P>
1761 If the pattern was successfully studied with one of the just-in-time (JIT)
1762 compile options, the only supported options for JIT execution are
1765 unsupported option is used, JIT execution is disabled and the normal
1766 interpretive code in <b>pcre_exec()</b> is run.
1767 <pre>
1769 </pre>
1770 The PCRE_ANCHORED option limits <b>pcre_exec()</b> to matching at the first
1771 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1772 to be anchored by virtue of its contents, it cannot be made unachored at
1773 matching time.
1774 <pre>
1777 </pre>
1778 These options (which are mutually exclusive) control what the \R escape
1779 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1780 match any Unicode newline sequence. These options override the choice that was
1781 made or defaulted when the pattern was compiled.
1782 <pre>
1788 </pre>
1789 These options override the newline definition that was chosen or defaulted when
1790 the pattern was compiled. For details, see the description of
1791 <b>pcre_compile()</b> above. During matching, the newline choice affects the
1792 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1793 the way the match position is advanced after a match failure for an unanchored
1794 pattern.
1795 </P>
1796 <P>
1798 match attempt for an unanchored pattern fails when the current position is at a
1799 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1800 characters, the match position is advanced by two characters instead of one, in
1801 other words, to after the CRLF.
1802 </P>
1803 <P>
1804 The above rule is a compromise that makes the most common cases work as
1805 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1806 set), it does not match the string "\r\nA" because, after failing at the
1807 start, it skips both the CR and the LF before retrying. However, the pattern
1808 [\r\n]A does match that string, because it contains an explicit CR or LF
1809 reference, and so advances only by one character after the first failure.
1810 </P>
1811 <P>
1812 An explicit match for CR of LF is either a literal appearance of one of those
1813 characters, or one of the \r or \n escape sequences. Implicit matches such as
1814 [^X] do not count, nor does \s (which includes CR and LF in the characters
1815 that it matches).
1816 </P>
1817 <P>
1818 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1819 valid newline sequence and explicit \r or \n escapes appear in the pattern.
1820 <pre>
1822 </pre>
1823 This option specifies that first character of the subject string is not the
1824 beginning of a line, so the circumflex metacharacter should not match before
1825 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1826 never to match. This option affects only the behaviour of the circumflex
1827 metacharacter. It does not affect \A.
1828 <pre>
1830 </pre>
1831 This option specifies that the end of the subject string is not the end of a
1832 line, so the dollar metacharacter should not match it nor (except in multiline
1833 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1834 compile time) causes dollar never to match. This option affects only the
1835 behaviour of the dollar metacharacter. It does not affect \Z or \z.
1836 <pre>
1838 </pre>
1839 An empty string is not considered to be a valid match if this option is set. If
1840 there are alternatives in the pattern, they are tried. If all the alternatives
1841 match the empty string, the entire match fails. For example, if the pattern
1842 <pre>
1843 a?b?
1844 </pre>
1845 is applied to a string not beginning with "a" or "b", it matches an empty
1846 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1847 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1848 <pre>
1850 </pre>
1851 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1852 the start of the subject is permitted. If the pattern is anchored, such a match
1853 can occur only if the pattern contains \K.
1854 </P>
1855 <P>
1856 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1857 does make a special case of a pattern match of the empty string within its
1858 <b>split()</b> function, and when using the /g modifier. It is possible to
1859 emulate Perl's behaviour after matching a null string by first trying the match
1860 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1861 if that fails, by advancing the starting offset (see below) and trying an
1862 ordinary match again. There is some code that demonstrates how to do this in
1863 the
1864 <a href="pcredemo.html"><b>pcredemo</b></a>
1865 sample program. In the most general case, you have to check to see if the
1866 newline convention recognizes CRLF as a newline, and if so, and the current
1867 character is CR followed by LF, advance the starting offset by two characters
1868 instead of one.
1869 <pre>
1871 </pre>
1872 There are a number of optimizations that <b>pcre_exec()</b> uses at the start of
1873 a match, in order to speed up the process. For example, if it is known that an
1874 unanchored match must start with a specific character, it searches the subject
1875 for that character, and fails immediately if it cannot find it, without
1876 actually running the main matching function. This means that a special item
1877 such as (*COMMIT) at the start of a pattern is not considered until after a
1878 suitable starting point for the match has been found. Also, when callouts or
1879 (*MARK) items are in use, these "start-up" optimizations can cause them to be
1880 skipped if the pattern is never actually used. The start-up optimizations are
1881 in effect a pre-scan of the subject that takes place before the pattern is run.
1882 </P>
1883 <P>
1884 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1885 causing performance to suffer, but ensuring that in cases where the result is
1886 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1887 are considered at every possible starting position in the subject string. If
1888 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1889 time. The use of PCRE_NO_START_OPTIMIZE at matching time (that is, passing it
1890 to <b>pcre_exec()</b>) disables JIT execution; in this situation, matching is
1891 always done using interpretively.
1892 </P>
1893 <P>
1894 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1895 Consider the pattern
1896 <pre>
1898 </pre>
1899 When this is compiled, PCRE records the fact that a match must start with the
1900 character "A". Suppose the subject string is "DEFABC". The start-up
1901 optimization scans along the subject, finds "A" and runs the first match
1902 attempt from there. The (*COMMIT) item means that the pattern must match the
1903 current starting position, which in this case, it does. However, if the same
1904 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1905 subject string does not happen. The first match attempt is run starting from
1906 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1907 the overall result is "no match". If the pattern is studied, more start-up
1908 optimizations may be used. For example, a minimum length for the subject may be
1909 recorded. Consider the pattern
1910 <pre>
1911 (*MARK:A)(X|Y)
1912 </pre>
1913 The minimum length for a match is one character. If the subject is "ABC", there
1914 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1915 If the pattern is studied, the final attempt does not take place, because PCRE
1916 knows that the subject is too short, and so the (*MARK) is never encountered.
1917 In this case, studying the pattern does not affect the overall match result,
1918 which is still "no match", but it does affect the auxiliary information that is
1919 returned.
1920 <pre>
1922 </pre>
1923 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1924 string is automatically checked when <b>pcre_exec()</b> is subsequently called.
1925 The entire string is checked before any other processing takes place. The value
1926 of <i>startoffset</i> is also checked to ensure that it points to the start of a
1927 UTF-8 character. There is a discussion about the
1928 <a href="pcreunicode.html#utf8strings">validity of UTF-8 strings</a>
1929 in the
1930 <a href="pcreunicode.html"><b>pcreunicode</b></a>
1931 page. If an invalid sequence of bytes is found, <b>pcre_exec()</b> returns the
1932 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1933 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1934 cases, information about the precise nature of the error may also be returned
1935 (see the descriptions of these errors in the section entitled \fIError return
1936 values from\fP <b>pcre_exec()</b>
1937 <a href="#errorlist">below).</a>
1938 If <i>startoffset</i> contains a value that does not point to the start of a
1939 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1940 returned.
1941 </P>
1942 <P>
1943 If you already know that your subject is valid, and you want to skip these
1944 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1945 calling <b>pcre_exec()</b>. You might want to do this for the second and
1946 subsequent calls to <b>pcre_exec()</b> if you are making repeated calls to find
1947 all the matches in a single subject string. However, you should be sure that
1948 the value of <i>startoffset</i> points to the start of a character (or the end
1949 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1950 invalid string as a subject or an invalid value of <i>startoffset</i> is
1951 undefined. Your program may crash.
1952 <pre>
1955 </pre>
1956 These options turn on the partial matching feature. For backwards
1957 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1958 occurs if the end of the subject string is reached successfully, but there are
1959 not enough subject characters to complete the match. If this happens when
1960 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1961 testing any remaining alternatives. Only if no complete match can be found is
1962 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1963 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1964 but only if no complete match can be found.
1965 </P>
1966 <P>
1967 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1968 partial match is found, <b>pcre_exec()</b> immediately returns
1969 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1970 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1971 important that an alternative complete match.
1972 </P>
1973 <P>
1974 In both cases, the portion of the string that was inspected when the partial
1975 match was found is set as the first matching string. There is a more detailed
1976 discussion of partial and multi-segment matching, with examples, in the
1977 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1978 documentation.
1979 </P>
1980 <br><b>
1981 The string to be matched by <b>pcre_exec()</b>
1982 </b><br>
1983 <P>
1984 The subject string is passed to <b>pcre_exec()</b> as a pointer in
1985 <i>subject</i>, a length in bytes in <i>length</i>, and a starting byte offset
1986 in <i>startoffset</i>. If this is negative or greater than the length of the
1987 subject, <b>pcre_exec()</b> returns PCRE_ERROR_BADOFFSET. When the starting
1988 offset is zero, the search for a match starts at the beginning of the subject,
1989 and this is by far the most common case. In UTF-8 mode, the byte offset must
1990 point to the start of a UTF-8 character (or the end of the subject). Unlike the
1991 pattern string, the subject may contain binary zero bytes.
1992 </P>
1993 <P>
1994 A non-zero starting offset is useful when searching for another match in the
1995 same subject by calling <b>pcre_exec()</b> again after a previous success.
1996 Setting <i>startoffset</i> differs from just passing over a shortened string and
1997 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1998 lookbehind. For example, consider the pattern
1999 <pre>
2000 \Biss\B
2001 </pre>
2002 which finds occurrences of "iss" in the middle of words. (\B matches only if
2003 the current position in the subject is not a word boundary.) When applied to
2004 the string "Mississipi" the first call to <b>pcre_exec()</b> finds the first
2005 occurrence. If <b>pcre_exec()</b> is called again with just the remainder of the
2006 subject, namely "issipi", it does not match, because \B is always false at the
2007 start of the subject, which is deemed to be a word boundary. However, if
2008 <b>pcre_exec()</b> is passed the entire string again, but with <i>startoffset</i>
2009 set to 4, it finds the second occurrence of "iss" because it is able to look
2010 behind the starting point to discover that it is preceded by a letter.
2011 </P>
2012 <P>
2013 Finding all the matches in a subject is tricky when the pattern can match an
2014 empty string. It is possible to emulate Perl's /g behaviour by first trying the
2015 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
2016 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
2017 and trying an ordinary match again. There is some code that demonstrates how to
2018 do this in the
2019 <a href="pcredemo.html"><b>pcredemo</b></a>
2020 sample program. In the most general case, you have to check to see if the
2021 newline convention recognizes CRLF as a newline, and if so, and the current
2022 character is CR followed by LF, advance the starting offset by two characters
2023 instead of one.
2024 </P>
2025 <P>
2026 If a non-zero starting offset is passed when the pattern is anchored, one
2027 attempt to match at the given offset is made. This can only succeed if the
2028 pattern does not require the match to be at the start of the subject.
2029 </P>
2030 <br><b>
2031 How <b>pcre_exec()</b> returns captured substrings
2032 </b><br>
2033 <P>
2034 In general, a pattern matches a certain portion of the subject, and in
2035 addition, further substrings from the subject may be picked out by parts of the
2036 pattern. Following the usage in Jeffrey Friedl's book, this is called
2037 "capturing" in what follows, and the phrase "capturing subpattern" is used for
2038 a fragment of a pattern that picks out a substring. PCRE supports several other
2039 kinds of parenthesized subpattern that do not cause substrings to be captured.
2040 </P>
2041 <P>
2042 Captured substrings are returned to the caller via a vector of integers whose
2043 address is passed in <i>ovector</i>. The number of elements in the vector is
2044 passed in <i>ovecsize</i>, which must be a non-negative number. <b>Note</b>: this
2045 argument is NOT the size of <i>ovector</i> in bytes.
2046 </P>
2047 <P>
2048 The first two-thirds of the vector is used to pass back captured substrings,
2049 each substring using a pair of integers. The remaining third of the vector is
2050 used as workspace by <b>pcre_exec()</b> while matching capturing subpatterns,
2051 and is not available for passing back information. The number passed in
2052 <i>ovecsize</i> should always be a multiple of three. If it is not, it is
2053 rounded down.
2054 </P>
2055 <P>
2056 When a match is successful, information about captured substrings is returned
2057 in pairs of integers, starting at the beginning of <i>ovector</i>, and
2058 continuing up to two-thirds of its length at the most. The first element of
2059 each pair is set to the byte offset of the first character in a substring, and
2060 the second is set to the byte offset of the first character after the end of a
2061 substring. <b>Note</b>: these values are always byte offsets, even in UTF-8
2062 mode. They are not character counts.
2063 </P>
2064 <P>
2065 The first pair of integers, <i>ovector[0]</i> and <i>ovector[1]</i>, identify the
2066 portion of the subject string matched by the entire pattern. The next pair is
2067 used for the first capturing subpattern, and so on. The value returned by
2068 <b>pcre_exec()</b> is one more than the highest numbered pair that has been set.
2069 For example, if two substrings have been captured, the returned value is 3. If
2070 there are no capturing subpatterns, the return value from a successful match is
2071 1, indicating that just the first pair of offsets has been set.
2072 </P>
2073 <P>
2074 If a capturing subpattern is matched repeatedly, it is the last portion of the
2075 string that it matched that is returned.
2076 </P>
2077 <P>
2078 If the vector is too small to hold all the captured substring offsets, it is
2079 used as far as possible (up to two-thirds of its length), and the function
2080 returns a value of zero. If neither the actual string matched nor any captured
2081 substrings are of interest, <b>pcre_exec()</b> may be called with <i>ovector</i>
2082 passed as NULL and <i>ovecsize</i> as zero. However, if the pattern contains
2083 back references and the <i>ovector</i> is not big enough to remember the related
2084 substrings, PCRE has to get additional memory for use during matching. Thus it
2085 is usually advisable to supply an <i>ovector</i> of reasonable size.
2086 </P>
2087 <P>
2088 There are some cases where zero is returned (indicating vector overflow) when
2089 in fact the vector is exactly the right size for the final match. For example,
2090 consider the pattern
2091 <pre>
2092 (a)(?:(b)c|bd)
2093 </pre>
2094 If a vector of 6 elements (allowing for only 1 captured substring) is given
2095 with subject string "abd", <b>pcre_exec()</b> will try to set the second
2096 captured string, thereby recording a vector overflow, before failing to match
2097 "c" and backing up to try the second alternative. The zero return, however,
2098 does correctly indicate that the maximum number of slots (namely 2) have been
2099 filled. In similar cases where there is temporary overflow, but the final
2100 number of used slots is actually less than the maximum, a non-zero value is
2101 returned.
2102 </P>
2103 <P>
2104 The <b>pcre_fullinfo()</b> function can be used to find out how many capturing
2105 subpatterns there are in a compiled pattern. The smallest size for
2106 <i>ovector</i> that will allow for <i>n</i> captured substrings, in addition to
2107 the offsets of the substring matched by the whole pattern, is (<i>n</i>+1)*3.
2108 </P>
2109 <P>
2110 It is possible for capturing subpattern number <i>n+1</i> to match some part of
2111 the subject when subpattern <i>n</i> has not been used at all. For example, if
2112 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2113 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2114 happens, both values in the offset pairs corresponding to unused subpatterns
2115 are set to -1.
2116 </P>
2117 <P>
2118 Offset values that correspond to unused subpatterns at the end of the
2119 expression are also set to -1. For example, if the string "abc" is matched
2120 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2121 return from the function is 2, because the highest used capturing subpattern
2122 number is 1, and the offsets for for the second and third capturing subpatterns
2123 (assuming the vector is large enough, of course) are set to -1.
2124 </P>
2125 <P>
2126 <b>Note</b>: Elements in the first two-thirds of <i>ovector</i> that do not
2127 correspond to capturing parentheses in the pattern are never changed. That is,
2128 if a pattern contains <i>n</i> capturing parentheses, no more than
2129 <i>ovector[0]</i> to <i>ovector[2n+1]</i> are set by <b>pcre_exec()</b>. The other
2130 elements (in the first two-thirds) retain whatever values they previously had.
2131 </P>
2132 <P>
2133 Some convenience functions are provided for extracting the captured substrings
2134 as separate strings. These are described below.
2135 <a name="errorlist"></a></P>
2136 <br><b>
2137 Error return values from <b>pcre_exec()</b>
2138 </b><br>
2139 <P>
2140 If <b>pcre_exec()</b> fails, it returns a negative number. The following are
2141 defined in the header file:
2142 <pre>
2144 </pre>
2145 The subject string did not match the pattern.
2146 <pre>
2148 </pre>
2149 Either <i>code</i> or <i>subject</i> was passed as NULL, or <i>ovector</i> was
2150 NULL and <i>ovecsize</i> was not zero.
2151 <pre>
2153 </pre>
2154 An unrecognized bit was set in the <i>options</i> argument.
2155 <pre>
2157 </pre>
2158 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2159 the case when it is passed a junk pointer and to detect when a pattern that was
2160 compiled in an environment of one endianness is run in an environment with the
2161 other endianness. This is the error that PCRE gives when the magic number is
2162 not present.
2163 <pre>
2165 </pre>
2166 While running the pattern match, an unknown item was encountered in the
2167 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2168 of the compiled pattern.
2169 <pre>
2171 </pre>
2172 If a pattern contains back references, but the <i>ovector</i> that is passed to
2173 <b>pcre_exec()</b> is not big enough to remember the referenced substrings, PCRE
2174 gets a block of memory at the start of matching to use for this purpose. If the
2175 call via <b>pcre_malloc()</b> fails, this error is given. The memory is
2176 automatically freed at the end of matching.
2177 </P>
2178 <P>
2179 This error is also given if <b>pcre_stack_malloc()</b> fails in
2180 <b>pcre_exec()</b>. This can happen only when PCRE has been compiled with
2181 <b>--disable-stack-for-recursion</b>.
2182 <pre>
2184 </pre>
2185 This error is used by the <b>pcre_copy_substring()</b>,
2186 <b>pcre_get_substring()</b>, and <b>pcre_get_substring_list()</b> functions (see
2187 below). It is never returned by <b>pcre_exec()</b>.
2188 <pre>
2190 </pre>
2191 The backtracking limit, as specified by the <i>match_limit</i> field in a
2192 <b>pcre_extra</b> structure (or defaulted) was reached. See the description
2193 above.
2194 <pre>
2196 </pre>
2197 This error is never generated by <b>pcre_exec()</b> itself. It is provided for
2198 use by callout functions that want to yield a distinctive error code. See the
2199 <a href="pcrecallout.html"><b>pcrecallout</b></a>
2200 documentation for details.
2201 <pre>
2203 </pre>
2204 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2205 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2206 (<i>ovecsize</i>) is at least 2, the byte offset to the start of the the invalid
2207 UTF-8 character is placed in the first element, and a reason code is placed in
2208 the second element. The reason codes are listed in the
2209 <a href="#badutf8reasons">following section.</a>
2210 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2211 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2212 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2213 <pre>
2215 </pre>
2216 The UTF-8 byte sequence that was passed as a subject was checked and found to
2217 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2218 <i>startoffset</i> did not point to the beginning of a UTF-8 character or the
2219 end of the subject.
2220 <pre>
2222 </pre>
2223 The subject string did not match, but it did match partially. See the
2224 <a href="pcrepartial.html"><b>pcrepartial</b></a>
2225 documentation for details of partial matching.
2226 <pre>
2228 </pre>
2229 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2230 option was used with a compiled pattern containing items that were not
2231 supported for partial matching. From release 8.00 onwards, there are no
2232 restrictions on partial matching.
2233 <pre>
2235 </pre>
2236 An unexpected internal error has occurred. This error could be caused by a bug
2237 in PCRE or by overwriting of the compiled pattern.
2238 <pre>
2240 </pre>
2241 This error is given if the value of the <i>ovecsize</i> argument is negative.
2242 <pre>
2244 </pre>
2245 The internal recursion limit, as specified by the <i>match_limit_recursion</i>
2246 field in a <b>pcre_extra</b> structure (or defaulted) was reached. See the
2247 description above.
2248 <pre>
2250 </pre>
2251 An invalid combination of PCRE_NEWLINE_<i>xxx</i> options was given.
2252 <pre>
2254 </pre>
2255 The value of <i>startoffset</i> was negative or greater than the length of the
2256 subject, that is, the value in <i>length</i>.
2257 <pre>
2259 </pre>
2260 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2261 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2262 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2263 fact sufficient to detect this case, but this special error code for
2264 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2265 retained for backwards compatibility.
2266 <pre>
2268 </pre>
2269 This error is returned when <b>pcre_exec()</b> detects a recursion loop within
2270 the pattern. Specifically, it means that either the whole pattern or a
2271 subpattern has been called recursively for the second time at the same position
2272 in the subject string. Some simple patterns that might do this are detected and
2273 faulted at compile time, but more complicated cases, in particular mutual
2274 recursions between two different subpatterns, cannot be detected until run
2275 time.
2276 <pre>
2278 </pre>
2279 This error is returned when a pattern that was successfully studied using a
2280 JIT compile option is being matched, but the memory available for the
2281 just-in-time processing stack is not large enough. See the
2282 <a href="pcrejit.html"><b>pcrejit</b></a>
2283 documentation for more details.
2284 <pre>
2286 </pre>
2287 This error is given if a pattern that was compiled by the 8-bit library is
2288 passed to a 16-bit or 32-bit library function, or vice versa.
2289 <pre>
2291 </pre>
2292 This error is given if a pattern that was compiled and saved is reloaded on a
2293 host with different endianness. The utility function
2294 <b>pcre_pattern_to_host_byte_order()</b> can be used to convert such a pattern
2295 so that it runs on the new host.
2296 <pre>
2298 </pre>
2299 This error is returned when a pattern that was successfully studied using a JIT
2300 compile option is being matched, but the matching mode (partial or complete
2301 match) does not correspond to any JIT compilation mode. When the JIT fast path
2302 function is used, this error may be also given for invalid options. See the
2303 <a href="pcrejit.html"><b>pcrejit</b></a>
2304 documentation for more details.
2305 <pre>
2307 </pre>
2308 This error is given if <b>pcre_exec()</b> is called with a negative value for
2309 the <i>length</i> argument.
2310 </P>
2311 <P>
2312 Error numbers -16 to -20, -22, and 30 are not used by <b>pcre_exec()</b>.
2313 <a name="badutf8reasons"></a></P>
2314 <br><b>
2315 Reason codes for invalid UTF-8 strings
2316 </b><br>
2317 <P>
2318 This section applies only to the 8-bit library. The corresponding information
2319 for the 16-bit and 32-bit libraries is given in the
2320 <a href="pcre16.html"><b>pcre16</b></a>
2321 and
2322 <a href="pcre32.html"><b>pcre32</b></a>
2323 pages.
2324 </P>
2325 <P>
2326 When <b>pcre_exec()</b> returns either PCRE_ERROR_BADUTF8 or
2327 PCRE_ERROR_SHORTUTF8, and the size of the output vector (<i>ovecsize</i>) is at
2328 least 2, the offset of the start of the invalid UTF-8 character is placed in
2329 the first output vector element (<i>ovector[0]</i>) and a reason code is placed
2330 in the second element (<i>ovector[1]</i>). The reason codes are given names in
2331 the <b>pcre.h</b> header file:
2332 <pre>
2338 </pre>
2339 The string ends with a truncated UTF-8 character; the code specifies how many
2340 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2341 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2342 allows for up to 6 bytes, and this is checked first; hence the possibility of
2343 4 or 5 missing bytes.
2344 <pre>
2349 PCRE_UTF8_ERR10
2350 </pre>
2351 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2352 character do not have the binary value 0b10 (that is, either the most
2353 significant bit is 0, or the next bit is 1).
2354 <pre>
2355 PCRE_UTF8_ERR11
2356 PCRE_UTF8_ERR12
2357 </pre>
2358 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2359 these code points are excluded by RFC 3629.
2360 <pre>
2361 PCRE_UTF8_ERR13
2362 </pre>
2363 A 4-byte character has a value greater than 0x10fff; these code points are
2364 excluded by RFC 3629.
2365 <pre>
2366 PCRE_UTF8_ERR14
2367 </pre>
2368 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2369 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2370 from UTF-8.
2371 <pre>
2372 PCRE_UTF8_ERR15
2373 PCRE_UTF8_ERR16
2374 PCRE_UTF8_ERR17
2375 PCRE_UTF8_ERR18
2376 PCRE_UTF8_ERR19
2377 </pre>
2378 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2379 value that can be represented by fewer bytes, which is invalid. For example,
2380 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2381 one byte.
2382 <pre>
2383 PCRE_UTF8_ERR20
2384 </pre>
2385 The two most significant bits of the first byte of a character have the binary
2386 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2387 byte can only validly occur as the second or subsequent byte of a multi-byte
2388 character.
2389 <pre>
2390 PCRE_UTF8_ERR21
2391 </pre>
2392 The first byte of a character has the value 0xfe or 0xff. These values can
2393 never occur in a valid UTF-8 string.
2394 <pre>
2395 PCRE_UTF8_ERR22
2396 </pre>
2397 This error code was formerly used when the presence of a so-called
2398 "non-character" caused an error. Unicode corrigendum #9 makes it clear that
2399 such characters should not cause a string to be rejected, and so this code is
2400 no longer in use and is never returned.
2401 </P>
2402 <br><a name="SEC18" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NUMBER</a><br>
2403 <P>
2404 <b>int pcre_copy_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
2405 <b>int <i>stringcount</i>, int <i>stringnumber</i>, char *<i>buffer</i>,</b>
2406 <b>int <i>buffersize</i>);</b>
2407 </P>
2408 <P>
2409 <b>int pcre_get_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
2410 <b>int <i>stringcount</i>, int <i>stringnumber</i>,</b>
2411 <b>const char **<i>stringptr</i>);</b>
2412 </P>
2413 <P>
2414 <b>int pcre_get_substring_list(const char *<i>subject</i>,</b>
2415 <b>int *<i>ovector</i>, int <i>stringcount</i>, const char ***<i>listptr</i>);</b>
2416 </P>
2417 <P>
2418 Captured substrings can be accessed directly by using the offsets returned by
2419 <b>pcre_exec()</b> in <i>ovector</i>. For convenience, the functions
2420 <b>pcre_copy_substring()</b>, <b>pcre_get_substring()</b>, and
2421 <b>pcre_get_substring_list()</b> are provided for extracting captured substrings
2422 as new, separate, zero-terminated strings. These functions identify substrings
2423 by number. The next section describes functions for extracting named
2424 substrings.
2425 </P>
2426 <P>
2427 A substring that contains a binary zero is correctly extracted and has a
2428 further zero added on the end, but the result is not, of course, a C string.
2429 However, you can process such a string by referring to the length that is
2430 returned by <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>.
2431 Unfortunately, the interface to <b>pcre_get_substring_list()</b> is not adequate
2432 for handling strings containing binary zeros, because the end of the final
2433 string is not independently indicated.
2434 </P>
2435 <P>
2436 The first three arguments are the same for all three of these functions:
2437 <i>subject</i> is the subject string that has just been successfully matched,
2438 <i>ovector</i> is a pointer to the vector of integer offsets that was passed to
2439 <b>pcre_exec()</b>, and <i>stringcount</i> is the number of substrings that were
2440 captured by the match, including the substring that matched the entire regular
2441 expression. This is the value returned by <b>pcre_exec()</b> if it is greater
2442 than zero. If <b>pcre_exec()</b> returned zero, indicating that it ran out of
2443 space in <i>ovector</i>, the value passed as <i>stringcount</i> should be the
2444 number of elements in the vector divided by three.
2445 </P>
2446 <P>
2447 The functions <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>
2448 extract a single substring, whose number is given as <i>stringnumber</i>. A
2449 value of zero extracts the substring that matched the entire pattern, whereas
2450 higher values extract the captured substrings. For <b>pcre_copy_substring()</b>,
2451 the string is placed in <i>buffer</i>, whose length is given by
2452 <i>buffersize</i>, while for <b>pcre_get_substring()</b> a new block of memory is
2453 obtained via <b>pcre_malloc</b>, and its address is returned via
2454 <i>stringptr</i>. The yield of the function is the length of the string, not
2455 including the terminating zero, or one of these error codes:
2456 <pre>
2458 </pre>
2459 The buffer was too small for <b>pcre_copy_substring()</b>, or the attempt to get
2460 memory failed for <b>pcre_get_substring()</b>.
2461 <pre>
2463 </pre>
2464 There is no substring whose number is <i>stringnumber</i>.
2465 </P>
2466 <P>
2467 The <b>pcre_get_substring_list()</b> function extracts all available substrings
2468 and builds a list of pointers to them. All this is done in a single block of
2469 memory that is obtained via <b>pcre_malloc</b>. The address of the memory block
2470 is returned via <i>listptr</i>, which is also the start of the list of string
2471 pointers. The end of the list is marked by a NULL pointer. The yield of the
2472 function is zero if all went well, or the error code
2473 <pre>
2475 </pre>
2476 if the attempt to get the memory block failed.
2477 </P>
2478 <P>
2479 When any of these functions encounter a substring that is unset, which can
2480 happen when capturing subpattern number <i>n+1</i> matches some part of the
2481 subject, but subpattern <i>n</i> has not been used at all, they return an empty
2482 string. This can be distinguished from a genuine zero-length substring by
2483 inspecting the appropriate offset in <i>ovector</i>, which is negative for unset
2484 substrings.
2485 </P>
2486 <P>
2487 The two convenience functions <b>pcre_free_substring()</b> and
2488 <b>pcre_free_substring_list()</b> can be used to free the memory returned by
2489 a previous call of <b>pcre_get_substring()</b> or
2490 <b>pcre_get_substring_list()</b>, respectively. They do nothing more than call
2491 the function pointed to by <b>pcre_free</b>, which of course could be called
2492 directly from a C program. However, PCRE is used in some situations where it is
2493 linked via a special interface to another programming language that cannot use
2494 <b>pcre_free</b> directly; it is for these cases that the functions are
2495 provided.
2496 </P>
2497 <br><a name="SEC19" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NAME</a><br>
2498 <P>
2499 <b>int pcre_get_stringnumber(const pcre *<i>code</i>,</b>
2500 <b>const char *<i>name</i>);</b>
2501 </P>
2502 <P>
2503 <b>int pcre_copy_named_substring(const pcre *<i>code</i>,</b>
2504 <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
2505 <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
2506 <b>char *<i>buffer</i>, int <i>buffersize</i>);</b>
2507 </P>
2508 <P>
2509 <b>int pcre_get_named_substring(const pcre *<i>code</i>,</b>
2510 <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
2511 <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
2512 <b>const char **<i>stringptr</i>);</b>
2513 </P>
2514 <P>
2515 To extract a substring by name, you first have to find associated number.
2516 For example, for this pattern
2517 <pre>
2518 (a+)b(?&#60;xxx&#62;\d+)...
2519 </pre>
2520 the number of the subpattern called "xxx" is 2. If the name is known to be
2521 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2522 calling <b>pcre_get_stringnumber()</b>. The first argument is the compiled
2523 pattern, and the second is the name. The yield of the function is the
2524 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2525 that name.
2526 </P>
2527 <P>
2528 Given the number, you can extract the substring directly, or use one of the
2529 functions described in the previous section. For convenience, there are also
2530 two functions that do the whole job.
2531 </P>
2532 <P>
2533 Most of the arguments of <b>pcre_copy_named_substring()</b> and
2534 <b>pcre_get_named_substring()</b> are the same as those for the similarly named
2535 functions that extract by number. As these are described in the previous
2536 section, they are not re-described here. There are just two differences:
2537 </P>
2538 <P>
2539 First, instead of a substring number, a substring name is given. Second, there
2540 is an extra argument, given at the start, which is a pointer to the compiled
2541 pattern. This is needed in order to gain access to the name-to-number
2542 translation table.
2543 </P>
2544 <P>
2545 These functions call <b>pcre_get_stringnumber()</b>, and if it succeeds, they
2546 then call <b>pcre_copy_substring()</b> or <b>pcre_get_substring()</b>, as
2547 appropriate. <b>NOTE:</b> If PCRE_DUPNAMES is set and there are duplicate names,
2548 the behaviour may not be what you want (see the next section).
2549 </P>
2550 <P>
2551 <b>Warning:</b> If the pattern uses the (?| feature to set up multiple
2552 subpatterns with the same number, as described in the
2553 <a href="pcrepattern.html#dupsubpatternnumber">section on duplicate subpattern numbers</a>
2554 in the
2555 <a href="pcrepattern.html"><b>pcrepattern</b></a>
2556 page, you cannot use names to distinguish the different subpatterns, because
2557 names are not included in the compiled code. The matching process uses only
2558 numbers. For this reason, the use of different names for subpatterns of the
2559 same number causes an error at compile time.
2560 </P>
2561 <br><a name="SEC20" href="#TOC1">DUPLICATE SUBPATTERN NAMES</a><br>
2562 <P>
2563 <b>int pcre_get_stringtable_entries(const pcre *<i>code</i>,</b>
2564 <b>const char *<i>name</i>, char **<i>first</i>, char **<i>last</i>);</b>
2565 </P>
2566 <P>
2567 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2568 are not required to be unique. (Duplicate names are always allowed for
2569 subpatterns with the same number, created by using the (?| feature. Indeed, if
2570 such subpatterns are named, they are required to use the same names.)
2571 </P>
2572 <P>
2573 Normally, patterns with duplicate names are such that in any one match, only
2574 one of the named subpatterns participates. An example is shown in the
2575 <a href="pcrepattern.html"><b>pcrepattern</b></a>
2576 documentation.
2577 </P>
2578 <P>
2579 When duplicates are present, <b>pcre_copy_named_substring()</b> and
2580 <b>pcre_get_named_substring()</b> return the first substring corresponding to
2581 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2582 returned; no data is returned. The <b>pcre_get_stringnumber()</b> function
2583 returns one of the numbers that are associated with the name, but it is not
2584 defined which it is.
2585 </P>
2586 <P>
2587 If you want to get full details of all captured substrings for a given name,
2588 you must use the <b>pcre_get_stringtable_entries()</b> function. The first
2589 argument is the compiled pattern, and the second is the name. The third and
2590 fourth are pointers to variables which are updated by the function. After it
2591 has run, they point to the first and last entries in the name-to-number table
2592 for the given name. The function itself returns the length of each entry, or
2593 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2594 described above in the section entitled <i>Information about a pattern</i>
2595 <a href="#infoaboutpattern">above.</a>
2596 Given all the relevant entries for the name, you can extract each of their
2597 numbers, and hence the captured data, if any.
2598 </P>
2599 <br><a name="SEC21" href="#TOC1">FINDING ALL POSSIBLE MATCHES</a><br>
2600 <P>
2601 The traditional matching function uses a similar algorithm to Perl, which stops
2602 when it finds the first match, starting at a given point in the subject. If you
2603 want to find all possible matches, or the longest possible match, consider
2604 using the alternative matching function (see below) instead. If you cannot use
2605 the alternative function, but still need to find all possible matches, you
2606 can kludge it up by making use of the callout facility, which is described in
2607 the
2608 <a href="pcrecallout.html"><b>pcrecallout</b></a>
2609 documentation.
2610 </P>
2611 <P>
2612 What you have to do is to insert a callout right at the end of the pattern.
2613 When your callout function is called, extract and save the current matched
2614 substring. Then return 1, which forces <b>pcre_exec()</b> to backtrack and try
2615 other alternatives. Ultimately, when it runs out of matches, <b>pcre_exec()</b>
2616 will yield PCRE_ERROR_NOMATCH.
2617 </P>
2618 <br><a name="SEC22" href="#TOC1">OBTAINING AN ESTIMATE OF STACK USAGE</a><br>
2619 <P>
2620 Matching certain patterns using <b>pcre_exec()</b> can use a lot of process
2621 stack, which in certain environments can be rather limited in size. Some users
2622 find it helpful to have an estimate of the amount of stack that is used by
2623 <b>pcre_exec()</b>, to help them set recursion limits, as described in the
2624 <a href="pcrestack.html"><b>pcrestack</b></a>
2625 documentation. The estimate that is output by <b>pcretest</b> when called with
2626 the <b>-m</b> and <b>-C</b> options is obtained by calling <b>pcre_exec</b> with
2627 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2628 </P>
2629 <P>
2630 Normally, if its first argument is NULL, <b>pcre_exec()</b> immediately returns
2631 the negative error code PCRE_ERROR_NULL, but with this special combination of
2632 arguments, it returns instead a negative number whose absolute value is the
2633 approximate stack frame size in bytes. (A negative number is used so that it is
2634 clear that no match has happened.) The value is approximate because in some
2635 cases, recursive calls to <b>pcre_exec()</b> occur when there are one or two
2636 additional variables on the stack.
2637 </P>
2638 <P>
2639 If PCRE has been compiled to use the heap instead of the stack for recursion,
2640 the value returned is the size of each block that is obtained from the heap.
2641 <a name="dfamatch"></a></P>
2642 <br><a name="SEC23" href="#TOC1">MATCHING A PATTERN: THE ALTERNATIVE FUNCTION</a><br>
2643 <P>
2644 <b>int pcre_dfa_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
2645 <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
2646 <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>,</b>
2647 <b>int *<i>workspace</i>, int <i>wscount</i>);</b>
2648 </P>
2649 <P>
2650 The function <b>pcre_dfa_exec()</b> is called to match a subject string against
2651 a compiled pattern, using a matching algorithm that scans the subject string
2652 just once, and does not backtrack. This has different characteristics to the
2653 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2654 patterns are not supported. Nevertheless, there are times when this kind of
2655 matching can be useful. For a discussion of the two matching algorithms, and a
2656 list of features that <b>pcre_dfa_exec()</b> does not support, see the
2657 <a href="pcrematching.html"><b>pcrematching</b></a>
2658 documentation.
2659 </P>
2660 <P>
2661 The arguments for the <b>pcre_dfa_exec()</b> function are the same as for
2662 <b>pcre_exec()</b>, plus two extras. The <i>ovector</i> argument is used in a
2663 different way, and this is described below. The other common arguments are used
2664 in the same way as for <b>pcre_exec()</b>, so their description is not repeated
2665 here.
2666 </P>
2667 <P>
2668 The two additional arguments provide workspace for the function. The workspace
2669 vector should contain at least 20 elements. It is used for keeping track of
2670 multiple paths through the pattern tree. More workspace will be needed for
2671 patterns and subjects where there are a lot of potential matches.
2672 </P>
2673 <P>
2674 Here is an example of a simple call to <b>pcre_dfa_exec()</b>:
2675 <pre>
2676 int rc;
2677 int ovector[10];
2678 int wspace[20];
2679 rc = pcre_dfa_exec(
2680 re, /* result of pcre_compile() */
2681 NULL, /* we didn't study the pattern */
2682 "some string", /* the subject string */
2683 11, /* the length of the subject string */
2684 0, /* start at offset 0 in the subject */
2685 0, /* default options */
2686 ovector, /* vector of integers for substring information */
2687 10, /* number of elements (NOT size in bytes) */
2688 wspace, /* working space vector */
2689 20); /* number of elements (NOT size in bytes) */
2690 </PRE>
2691 </P>
2692 <br><b>
2693 Option bits for <b>pcre_dfa_exec()</b>
2694 </b><br>
2695 <P>
2696 The unused bits of the <i>options</i> argument for <b>pcre_dfa_exec()</b> must be
2697 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
2701 All but the last four of these are exactly the same as for <b>pcre_exec()</b>,
2702 so their description is not repeated here.
2703 <pre>
2706 </pre>
2707 These have the same general effect as they do for <b>pcre_exec()</b>, but the
2708 details are slightly different. When PCRE_PARTIAL_HARD is set for
2709 <b>pcre_dfa_exec()</b>, it returns PCRE_ERROR_PARTIAL if the end of the subject
2710 is reached and there is still at least one matching possibility that requires
2711 additional characters. This happens even if some complete matches have also
2712 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2713 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2714 there have been no complete matches, but there is still at least one matching
2715 possibility. The portion of the string that was inspected when the longest
2716 partial match was found is set as the first matching string in both cases.
2717 There is a more detailed discussion of partial and multi-segment matching, with
2718 examples, in the
2719 <a href="pcrepartial.html"><b>pcrepartial</b></a>
2720 documentation.
2721 <pre>
2723 </pre>
2724 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2725 soon as it has found one match. Because of the way the alternative algorithm
2726 works, this is necessarily the shortest possible match at the first possible
2727 matching point in the subject string.
2728 <pre>
2730 </pre>
2731 When <b>pcre_dfa_exec()</b> returns a partial match, it is possible to call it
2732 again, with additional subject characters, and have it continue with the same
2733 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2734 <i>workspace</i> and <i>wscount</i> options must reference the same vector as
2735 before because data about the match so far is left in them after a partial
2736 match. There is more discussion of this facility in the
2737 <a href="pcrepartial.html"><b>pcrepartial</b></a>
2738 documentation.
2739 </P>
2740 <br><b>
2741 Successful returns from <b>pcre_dfa_exec()</b>
2742 </b><br>
2743 <P>
2744 When <b>pcre_dfa_exec()</b> succeeds, it may have matched more than one
2745 substring in the subject. Note, however, that all the matches from one run of
2746 the function start at the same point in the subject. The shorter matches are
2747 all initial substrings of the longer matches. For example, if the pattern
2748 <pre>
2749 &#60;.*&#62;
2750 </pre>
2751 is matched against the string
2752 <pre>
2753 This is &#60;something&#62; &#60;something else&#62; &#60;something further&#62; no more
2754 </pre>
2755 the three matched strings are
2756 <pre>
2757 &#60;something&#62;
2758 &#60;something&#62; &#60;something else&#62;
2759 &#60;something&#62; &#60;something else&#62; &#60;something further&#62;
2760 </pre>
2761 On success, the yield of the function is a number greater than zero, which is
2762 the number of matched substrings. The substrings themselves are returned in
2763 <i>ovector</i>. Each string uses two elements; the first is the offset to the
2764 start, and the second is the offset to the end. In fact, all the strings have
2765 the same start offset. (Space could have been saved by giving this only once,
2766 but it was decided to retain some compatibility with the way <b>pcre_exec()</b>
2767 returns data, even though the meaning of the strings is different.)
2768 </P>
2769 <P>
2770 The strings are returned in reverse order of length; that is, the longest
2771 matching string is given first. If there were too many matches to fit into
2772 <i>ovector</i>, the yield of the function is zero, and the vector is filled with
2773 the longest matches. Unlike <b>pcre_exec()</b>, <b>pcre_dfa_exec()</b> can use
2774 the entire <i>ovector</i> for returning matched strings.
2775 </P>
2776 <br><b>
2777 Error returns from <b>pcre_dfa_exec()</b>
2778 </b><br>
2779 <P>
2780 The <b>pcre_dfa_exec()</b> function returns a negative number when it fails.
2781 Many of the errors are the same as for <b>pcre_exec()</b>, and these are
2782 described
2783 <a href="#errorlist">above.</a>
2784 There are in addition the following errors that are specific to
2785 <b>pcre_dfa_exec()</b>:
2786 <pre>
2788 </pre>
2789 This return is given if <b>pcre_dfa_exec()</b> encounters an item in the pattern
2790 that it does not support, for instance, the use of \C or a back reference.
2791 <pre>
2793 </pre>
2794 This return is given if <b>pcre_dfa_exec()</b> encounters a condition item that
2795 uses a back reference for the condition, or a test for recursion in a specific
2796 group. These are not supported.
2797 <pre>
2799 </pre>
2800 This return is given if <b>pcre_dfa_exec()</b> is called with an <i>extra</i>
2801 block that contains a setting of the <i>match_limit</i> or
2802 <i>match_limit_recursion</i> fields. This is not supported (these fields are
2803 meaningless for DFA matching).
2804 <pre>
2806 </pre>
2807 This return is given if <b>pcre_dfa_exec()</b> runs out of space in the
2808 <i>workspace</i> vector.
2809 <pre>
2811 </pre>
2812 When a recursive subpattern is processed, the matching function calls itself
2813 recursively, using private vectors for <i>ovector</i> and <i>workspace</i>. This
2814 error is given if the output vector is not large enough. This should be
2815 extremely rare, as a vector of size 1000 is used.
2816 <pre>
2818 </pre>
2819 When <b>pcre_dfa_exec()</b> is called with the <b>PCRE_DFA_RESTART</b> option,
2820 some plausibility checks are made on the contents of the workspace, which
2821 should contain data about the previous partial match. If any of these checks
2822 fail, this error is given.
2823 </P>
2824 <br><a name="SEC24" href="#TOC1">SEE ALSO</a><br>
2825 <P>
2826 <b>pcre16</b>(3), <b>pcre32</b>(3), <b>pcrebuild</b>(3), <b>pcrecallout</b>(3),
2827 <b>pcrecpp(3)</b>(3), <b>pcrematching</b>(3), <b>pcrepartial</b>(3),
2828 <b>pcreposix</b>(3), <b>pcreprecompile</b>(3), <b>pcresample</b>(3),
2829 <b>pcrestack</b>(3).
2830 </P>
2831 <br><a name="SEC25" href="#TOC1">AUTHOR</a><br>
2832 <P>
2833 Philip Hazel
2834 <br>
2835 University Computing Service
2836 <br>
2837 Cambridge CB2 3QH, England.
2838 <br>
2839 </P>
2840 <br><a name="SEC26" href="#TOC1">REVISION</a><br>
2841 <P>
2842 Last updated: 26 April 2013
2843 <br>
2844 Copyright &copy; 1997-2013 University of Cambridge.
2845 <br>
2846 <p>
2847 Return to the <a href="index.html">PCRE index page</a>.
2848 </p>


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

  ViewVC Help
Powered by ViewVC 1.1.5