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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1298 - (show annotations)
Fri Mar 22 16:13:13 2013 UTC (6 years, 5 months ago) by ph10
File MIME type: text/html
File size: 128722 byte(s)
Fix COMMIT in recursion; document backtracking verbs in assertions and 
subroutines.
1 <html>
2 <head>
3 <title>pcreapi specification</title>
4 </head>
5 <body bgcolor="#FFFFFF" text="#00005A" link="#0066FF" alink="#3399FF" vlink="#2222BB">
6 <h1>pcreapi man page</h1>
7 <p>
8 Return to the <a href="index.html">PCRE index page</a>.
9 </p>
10 <p>
11 This page is part of the PCRE HTML documentation. It was generated automatically
12 from the original man page. If there is any nonsense in it, please consult the
13 man page, in case the conversion went wrong.
14 <br>
15 <ul>
16 <li><a name="TOC1" href="#SEC1">PCRE NATIVE API BASIC FUNCTIONS</a>
17 <li><a name="TOC2" href="#SEC2">PCRE NATIVE API STRING EXTRACTION FUNCTIONS</a>
18 <li><a name="TOC3" href="#SEC3">PCRE NATIVE API AUXILIARY FUNCTIONS</a>
19 <li><a name="TOC4" href="#SEC4">PCRE NATIVE API INDIRECTED 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>
24 <li><a name="TOC9" href="#SEC9">SAVING PRECOMPILED PATTERNS FOR LATER USE</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>
32 <li><a name="TOC17" href="#SEC17">MATCHING A PATTERN: THE TRADITIONAL FUNCTION</a>
33 <li><a name="TOC18" href="#SEC18">EXTRACTING CAPTURED SUBSTRINGS BY NUMBER</a>
34 <li><a name="TOC19" href="#SEC19">EXTRACTING CAPTURED SUBSTRINGS BY NAME</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>
38 <li><a name="TOC23" href="#SEC23">MATCHING A PATTERN: THE ALTERNATIVE FUNCTION</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>
76 <br><a name="SEC2" href="#TOC1">PCRE NATIVE API STRING EXTRACTION FUNCTIONS</a><br>
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>
412 PCRE_CONFIG_UTF8
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>
419 PCRE_CONFIG_UTF16
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>
426 PCRE_CONFIG_UTF32
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>
433 PCRE_CONFIG_UNICODE_PROPERTIES
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>
438 PCRE_CONFIG_JIT
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>
443 PCRE_CONFIG_JITTARGET
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>
450 PCRE_CONFIG_NEWLINE
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>
461 PCRE_CONFIG_BSR
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>
468 PCRE_CONFIG_LINK_SIZE
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>
479 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
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>
487 PCRE_CONFIG_MATCH_LIMIT
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>
493 PCRE_CONFIG_MATCH_LIMIT_RECURSION
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>
499 PCRE_CONFIG_STACKRECURSE
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>
602 PCRE_ANCHORED
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>
610 PCRE_AUTO_CALLOUT
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>
618 PCRE_BSR_ANYCRLF
619 PCRE_BSR_UNICODE
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>
627 PCRE_CASELESS
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>
639 PCRE_DOLLAR_ENDONLY
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>
648 PCRE_DOTALL
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>
658 PCRE_DUPNAMES
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>
667 PCRE_EXTENDED
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>
691 PCRE_EXTRA
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>
703 PCRE_FIRSTLINE
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>
709 PCRE_JAVASCRIPT_COMPAT
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>
742 PCRE_MULTILINE
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>
759 PCRE_NEWLINE_CR
760 PCRE_NEWLINE_LF
761 PCRE_NEWLINE_CRLF
762 PCRE_NEWLINE_ANYCRLF
763 PCRE_NEWLINE_ANY
764 </pre>
765 These options override the default newline definition that was chosen when PCRE
766 was built. Setting the first or the second specifies that a newline is
767 indicated by a single character (CR or LF, respectively). Setting
768 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
769 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
770 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
771 that any Unicode newline sequence should be recognized.
772 </P>
773 <P>
774 In an ASCII/Unicode environment, the Unicode newline sequences are the three
775 just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
776 feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
777 (paragraph separator, U+2029). For the 8-bit library, the last two are
778 recognized only in UTF-8 mode.
779 </P>
780 <P>
781 When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
782 CR is 0x0d, the same as ASCII. However, the character code for LF is normally
783 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
784 not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
785 less than 256. For more details, see the
786 <a href="pcrebuild.html"><b>pcrebuild</b></a>
787 documentation.
788 </P>
789 <P>
790 The newline setting in the options word uses three bits that are treated
791 as a number, giving eight possibilities. Currently only six are used (default
792 plus the five values above). This means that if you set more than one newline
793 option, the combination may or may not be sensible. For example,
794 PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
795 other combinations may yield unused numbers and cause an error.
796 </P>
797 <P>
798 The only time that a line break in a pattern is specially recognized when
799 compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
800 and so are ignored in this mode. Also, an unescaped # outside a character class
801 indicates a comment that lasts until after the next line break sequence. In
802 other circumstances, line break sequences in patterns are treated as literal
803 data.
804 </P>
805 <P>
806 The newline option that is set at compile time becomes the default that is used
807 for <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, but it can be overridden.
808 <pre>
809 PCRE_NO_AUTO_CAPTURE
810 </pre>
811 If this option is set, it disables the use of numbered capturing parentheses in
812 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
813 were followed by ?: but named parentheses can still be used for capturing (and
814 they acquire numbers in the usual way). There is no equivalent of this option
815 in Perl.
816 <pre>
817 NO_START_OPTIMIZE
818 </pre>
819 This is an option that acts at matching time; that is, it is really an option
820 for <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b>. If it is set at compile time,
821 it is remembered with the compiled pattern and assumed at matching time. For
822 details see the discussion of PCRE_NO_START_OPTIMIZE
823 <a href="#execoptions">below.</a>
824 <pre>
825 PCRE_UCP
826 </pre>
827 This option changes the way PCRE processes \B, \b, \D, \d, \S, \s, \W,
828 \w, and some of the POSIX character classes. By default, only ASCII characters
829 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
830 classify characters. More details are given in the section on
831 <a href="pcre.html#genericchartypes">generic character types</a>
832 in the
833 <a href="pcrepattern.html"><b>pcrepattern</b></a>
834 page. If you set PCRE_UCP, matching one of the items it affects takes much
835 longer. The option is available only if PCRE has been compiled with Unicode
836 property support.
837 <pre>
838 PCRE_UNGREEDY
839 </pre>
840 This option inverts the "greediness" of the quantifiers so that they are not
841 greedy by default, but become greedy if followed by "?". It is not compatible
842 with Perl. It can also be set by a (?U) option setting within the pattern.
843 <pre>
844 PCRE_UTF8
845 </pre>
846 This option causes PCRE to regard both the pattern and the subject as strings
847 of UTF-8 characters instead of single-byte strings. However, it is available
848 only when PCRE is built to include UTF support. If not, the use of this option
849 provokes an error. Details of how this option changes the behaviour of PCRE are
850 given in the
851 <a href="pcreunicode.html"><b>pcreunicode</b></a>
852 page.
853 <pre>
854 PCRE_NO_UTF8_CHECK
855 </pre>
856 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
857 automatically checked. There is a discussion about the
858 <a href="pcreunicode.html#utf8strings">validity of UTF-8 strings</a>
859 in the
860 <a href="pcreunicode.html"><b>pcreunicode</b></a>
861 page. If an invalid UTF-8 sequence is found, <b>pcre_compile()</b> returns an
862 error. If you already know that your pattern is valid, and you want to skip
863 this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
864 When it is set, the effect of passing an invalid UTF-8 string as a pattern is
865 undefined. It may cause your program to crash. Note that this option can also
866 be passed to <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, to suppress the
867 validity checking of subject strings only. If the same string is being matched
868 many times, the option can be safely set for the second and subsequent
869 matchings to improve performance.
870 </P>
871 <br><a name="SEC12" href="#TOC1">COMPILATION ERROR CODES</a><br>
872 <P>
873 The following table lists the error codes than may be returned by
874 <b>pcre_compile2()</b>, along with the error messages that may be returned by
875 both compiling functions. Note that error messages are always 8-bit ASCII
876 strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
877 have fallen out of use. To avoid confusion, they have not been re-used.
878 <pre>
879 0 no error
880 1 \ at end of pattern
881 2 \c at end of pattern
882 3 unrecognized character follows \
883 4 numbers out of order in {} quantifier
884 5 number too big in {} quantifier
885 6 missing terminating ] for character class
886 7 invalid escape sequence in character class
887 8 range out of order in character class
888 9 nothing to repeat
889 10 [this code is not in use]
890 11 internal error: unexpected repeat
891 12 unrecognized character after (? or (?-
892 13 POSIX named classes are supported only within a class
893 14 missing )
894 15 reference to non-existent subpattern
895 16 erroffset passed as NULL
896 17 unknown option bit(s) set
897 18 missing ) after comment
898 19 [this code is not in use]
899 20 regular expression is too large
900 21 failed to get memory
901 22 unmatched parentheses
902 23 internal error: code overflow
903 24 unrecognized character after (?&#60;
904 25 lookbehind assertion is not fixed length
905 26 malformed number or name after (?(
906 27 conditional group contains more than two branches
907 28 assertion expected after (?(
908 29 (?R or (?[+-]digits must be followed by )
909 30 unknown POSIX class name
910 31 POSIX collating elements are not supported
911 32 this version of PCRE is compiled without UTF support
912 33 [this code is not in use]
913 34 character value in \x{...} sequence is too large
914 35 invalid condition (?(0)
915 36 \C not allowed in lookbehind assertion
916 37 PCRE does not support \L, \l, \N{name}, \U, or \u
917 38 number after (?C is &#62; 255
918 39 closing ) for (?C expected
919 40 recursive call could loop indefinitely
920 41 unrecognized character after (?P
921 42 syntax error in subpattern name (missing terminator)
922 43 two named subpatterns have the same name
923 44 invalid UTF-8 string (specifically UTF-8)
924 45 support for \P, \p, and \X has not been compiled
925 46 malformed \P or \p sequence
926 47 unknown property name after \P or \p
927 48 subpattern name is too long (maximum 32 characters)
928 49 too many named subpatterns (maximum 10000)
929 50 [this code is not in use]
930 51 octal value is greater than \377 in 8-bit non-UTF-8 mode
931 52 internal error: overran compiling workspace
932 53 internal error: previously-checked referenced subpattern
933 not found
934 54 DEFINE group contains more than one branch
935 55 repeating a DEFINE group is not allowed
936 56 inconsistent NEWLINE options
937 57 \g is not followed by a braced, angle-bracketed, or quoted
938 name/number or by a plain number
939 58 a numbered reference must not be zero
940 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
941 60 (*VERB) not recognized
942 61 number is too big
943 62 subpattern name expected
944 63 digit expected after (?+
945 64 ] is an invalid data character in JavaScript compatibility mode
946 65 different names for subpatterns of the same number are
947 not allowed
948 66 (*MARK) must have an argument
949 67 this version of PCRE is not compiled with Unicode property
950 support
951 68 \c must be followed by an ASCII character
952 69 \k is not followed by a braced, angle-bracketed, or quoted name
953 70 internal error: unknown opcode in find_fixedlength()
954 71 \N is not supported in a class
955 72 too many forward references
956 73 disallowed Unicode code point (&#62;= 0xd800 && &#60;= 0xdfff)
957 74 invalid UTF-16 string (specifically UTF-16)
958 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
959 76 character value in \u.... sequence is too large
960 77 invalid UTF-32 string (specifically UTF-32)
961 </pre>
962 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
963 be used if the limits were changed when PCRE was built.
964 <a name="studyingapattern"></a></P>
965 <br><a name="SEC13" href="#TOC1">STUDYING A PATTERN</a><br>
966 <P>
967 <b>pcre_extra *pcre_study(const pcre *<i>code</i>, int <i>options</i></b>
968 <b>const char **<i>errptr</i>);</b>
969 </P>
970 <P>
971 If a compiled pattern is going to be used several times, it is worth spending
972 more time analyzing it in order to speed up the time taken for matching. The
973 function <b>pcre_study()</b> takes a pointer to a compiled pattern as its first
974 argument. If studying the pattern produces additional information that will
975 help speed up matching, <b>pcre_study()</b> returns a pointer to a
976 <b>pcre_extra</b> block, in which the <i>study_data</i> field points to the
977 results of the study.
978 </P>
979 <P>
980 The returned value from <b>pcre_study()</b> can be passed directly to
981 <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b>. However, a <b>pcre_extra</b> block
982 also contains other fields that can be set by the caller before the block is
983 passed; these are described
984 <a href="#extradata">below</a>
985 in the section on matching a pattern.
986 </P>
987 <P>
988 If studying the pattern does not produce any useful information,
989 <b>pcre_study()</b> returns NULL by default. In that circumstance, if the
990 calling program wants to pass any of the other fields to <b>pcre_exec()</b> or
991 <b>pcre_dfa_exec()</b>, it must set up its own <b>pcre_extra</b> block. However,
992 if <b>pcre_study()</b> is called with the PCRE_STUDY_EXTRA_NEEDED option, it
993 returns a <b>pcre_extra</b> block even if studying did not find any additional
994 information. It may still return NULL, however, if an error occurs in
995 <b>pcre_study()</b>.
996 </P>
997 <P>
998 The second argument of <b>pcre_study()</b> contains option bits. There are three
999 further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1000 <pre>
1001 PCRE_STUDY_JIT_COMPILE
1002 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1003 PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1004 </pre>
1005 If any of these are set, and the just-in-time compiler is available, the
1006 pattern is further compiled into machine code that executes much faster than
1007 the <b>pcre_exec()</b> interpretive matching function. If the just-in-time
1008 compiler is not available, these options are ignored. All undefined bits in the
1009 <i>options</i> argument must be zero.
1010 </P>
1011 <P>
1012 JIT compilation is a heavyweight optimization. It can take some time for
1013 patterns to be analyzed, and for one-off matches and simple patterns the
1014 benefit of faster execution might be offset by a much slower study time.
1015 Not all patterns can be optimized by the JIT compiler. For those that cannot be
1016 handled, matching automatically falls back to the <b>pcre_exec()</b>
1017 interpreter. For more details, see the
1018 <a href="pcrejit.html"><b>pcrejit</b></a>
1019 documentation.
1020 </P>
1021 <P>
1022 The third argument for <b>pcre_study()</b> is a pointer for an error message. If
1023 studying succeeds (even if no data is returned), the variable it points to is
1024 set to NULL. Otherwise it is set to point to a textual error message. This is a
1025 static string that is part of the library. You must not try to free it. You
1026 should test the error pointer for NULL after calling <b>pcre_study()</b>, to be
1027 sure that it has run successfully.
1028 </P>
1029 <P>
1030 When you are finished with a pattern, you can free the memory used for the
1031 study data by calling <b>pcre_free_study()</b>. This function was added to the
1032 API for release 8.20. For earlier versions, the memory could be freed with
1033 <b>pcre_free()</b>, just like the pattern itself. This will still work in cases
1034 where JIT optimization is not used, but it is advisable to change to the new
1035 function when convenient.
1036 </P>
1037 <P>
1038 This is a typical way in which <b>pcre_study</b>() is used (except that in a
1039 real application there should be tests for errors):
1040 <pre>
1041 int rc;
1042 pcre *re;
1043 pcre_extra *sd;
1044 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1045 sd = pcre_study(
1046 re, /* result of pcre_compile() */
1047 0, /* no options */
1048 &error); /* set to NULL or points to a message */
1049 rc = pcre_exec( /* see below for details of pcre_exec() options */
1050 re, sd, "subject", 7, 0, 0, ovector, 30);
1051 ...
1052 pcre_free_study(sd);
1053 pcre_free(re);
1054 </pre>
1055 Studying a pattern does two things: first, a lower bound for the length of
1056 subject string that is needed to match the pattern is computed. This does not
1057 mean that there are any strings of that length that match, but it does
1058 guarantee that no shorter strings match. The value is used to avoid wasting
1059 time by trying to match strings that are shorter than the lower bound. You can
1060 find out the value in a calling program via the <b>pcre_fullinfo()</b> function.
1061 </P>
1062 <P>
1063 Studying a pattern is also useful for non-anchored patterns that do not have a
1064 single fixed starting character. A bitmap of possible starting bytes is
1065 created. This speeds up finding a position in the subject at which to start
1066 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1067 In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1068 </P>
1069 <P>
1070 These two optimizations apply to both <b>pcre_exec()</b> and
1071 <b>pcre_dfa_exec()</b>, and the information is also used by the JIT compiler.
1072 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option
1073 when calling <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b>, but if this is done,
1074 JIT execution is also disabled. You might want to do this if your pattern
1075 contains callouts or (*MARK) and you want to make use of these facilities in
1076 cases where matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
1077 <a href="#execoptions">below.</a>
1078 <a name="localesupport"></a></P>
1079 <br><a name="SEC14" href="#TOC1">LOCALE SUPPORT</a><br>
1080 <P>
1081 PCRE handles caseless matching, and determines whether characters are letters,
1082 digits, or whatever, by reference to a set of tables, indexed by character
1083 value. When running in UTF-8 mode, this applies only to characters
1084 with codes less than 128. By default, higher-valued codes never match escapes
1085 such as \w or \d, but they can be tested with \p if PCRE is built with
1086 Unicode character property support. Alternatively, the PCRE_UCP option can be
1087 set at compile time; this causes \w and friends to use Unicode property
1088 support instead of built-in tables. The use of locales with Unicode is
1089 discouraged. If you are handling characters with codes greater than 128, you
1090 should either use UTF-8 and Unicode, or use locales, but not try to mix the
1091 two.
1092 </P>
1093 <P>
1094 PCRE contains an internal set of tables that are used when the final argument
1095 of <b>pcre_compile()</b> is NULL. These are sufficient for many applications.
1096 Normally, the internal tables recognize only ASCII characters. However, when
1097 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1098 default "C" locale of the local system, which may cause them to be different.
1099 </P>
1100 <P>
1101 The internal tables can always be overridden by tables supplied by the
1102 application that calls PCRE. These may be created in a different locale from
1103 the default. As more and more applications change to using Unicode, the need
1104 for this locale support is expected to die away.
1105 </P>
1106 <P>
1107 External tables are built by calling the <b>pcre_maketables()</b> function,
1108 which has no arguments, in the relevant locale. The result can then be passed
1109 to <b>pcre_compile()</b> or <b>pcre_exec()</b> as often as necessary. For
1110 example, to build and use tables that are appropriate for the French locale
1111 (where accented characters with values greater than 128 are treated as letters),
1112 the following code could be used:
1113 <pre>
1114 setlocale(LC_CTYPE, "fr_FR");
1115 tables = pcre_maketables();
1116 re = pcre_compile(..., tables);
1117 </pre>
1118 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1119 are using Windows, the name for the French locale is "french".
1120 </P>
1121 <P>
1122 When <b>pcre_maketables()</b> runs, the tables are built in memory that is
1123 obtained via <b>pcre_malloc</b>. It is the caller's responsibility to ensure
1124 that the memory containing the tables remains available for as long as it is
1125 needed.
1126 </P>
1127 <P>
1128 The pointer that is passed to <b>pcre_compile()</b> is saved with the compiled
1129 pattern, and the same tables are used via this pointer by <b>pcre_study()</b>
1130 and normally also by <b>pcre_exec()</b>. Thus, by default, for any single
1131 pattern, compilation, studying and matching all happen in the same locale, but
1132 different patterns can be compiled in different locales.
1133 </P>
1134 <P>
1135 It is possible to pass a table pointer or NULL (indicating the use of the
1136 internal tables) to <b>pcre_exec()</b>. Although not intended for this purpose,
1137 this facility could be used to match a pattern in a different locale from the
1138 one in which it was compiled. Passing table pointers at run time is discussed
1139 below in the section on matching a pattern.
1140 <a name="infoaboutpattern"></a></P>
1141 <br><a name="SEC15" href="#TOC1">INFORMATION ABOUT A PATTERN</a><br>
1142 <P>
1143 <b>int pcre_fullinfo(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
1144 <b>int <i>what</i>, void *<i>where</i>);</b>
1145 </P>
1146 <P>
1147 The <b>pcre_fullinfo()</b> function returns information about a compiled
1148 pattern. It replaces the <b>pcre_info()</b> function, which was removed from the
1149 library at version 8.30, after more than 10 years of obsolescence.
1150 </P>
1151 <P>
1152 The first argument for <b>pcre_fullinfo()</b> is a pointer to the compiled
1153 pattern. The second argument is the result of <b>pcre_study()</b>, or NULL if
1154 the pattern was not studied. The third argument specifies which piece of
1155 information is required, and the fourth argument is a pointer to a variable
1156 to receive the data. The yield of the function is zero for success, or one of
1157 the following negative numbers:
1158 <pre>
1159 PCRE_ERROR_NULL the argument <i>code</i> was NULL
1160 the argument <i>where</i> was NULL
1161 PCRE_ERROR_BADMAGIC the "magic number" was not found
1162 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1163 endianness
1164 PCRE_ERROR_BADOPTION the value of <i>what</i> was invalid
1165 </pre>
1166 The "magic number" is placed at the start of each compiled pattern as an simple
1167 check against passing an arbitrary memory pointer. The endianness error can
1168 occur if a compiled pattern is saved and reloaded on a different host. Here is
1169 a typical call of <b>pcre_fullinfo()</b>, to obtain the length of the compiled
1170 pattern:
1171 <pre>
1172 int rc;
1173 size_t length;
1174 rc = pcre_fullinfo(
1175 re, /* result of pcre_compile() */
1176 sd, /* result of pcre_study(), or NULL */
1177 PCRE_INFO_SIZE, /* what is required */
1178 &length); /* where to put the data */
1179 </pre>
1180 The possible values for the third argument are defined in <b>pcre.h</b>, and are
1181 as follows:
1182 <pre>
1183 PCRE_INFO_BACKREFMAX
1184 </pre>
1185 Return the number of the highest back reference in the pattern. The fourth
1186 argument should point to an <b>int</b> variable. Zero is returned if there are
1187 no back references.
1188 <pre>
1189 PCRE_INFO_CAPTURECOUNT
1190 </pre>
1191 Return the number of capturing subpatterns in the pattern. The fourth argument
1192 should point to an <b>int</b> variable.
1193 <pre>
1194 PCRE_INFO_DEFAULT_TABLES
1195 </pre>
1196 Return a pointer to the internal default character tables within PCRE. The
1197 fourth argument should point to an <b>unsigned char *</b> variable. This
1198 information call is provided for internal use by the <b>pcre_study()</b>
1199 function. External callers can cause PCRE to use its internal tables by passing
1200 a NULL table pointer.
1201 <pre>
1202 PCRE_INFO_FIRSTBYTE
1203 </pre>
1204 Return information about the first data unit of any matched string, for a
1205 non-anchored pattern. (The name of this option refers to the 8-bit library,
1206 where data units are bytes.) The fourth argument should point to an <b>int</b>
1207 variable.
1208 </P>
1209 <P>
1210 If there is a fixed first value, for example, the letter "c" from a pattern
1211 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1212 value is always less than 256. In the 16-bit library the value can be up to
1213 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1214 </P>
1215 <P>
1216 If there is no fixed first value, and if either
1217 <br>
1218 <br>
1219 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1220 starts with "^", or
1221 <br>
1222 <br>
1223 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1224 (if it were set, the pattern would be anchored),
1225 <br>
1226 <br>
1227 -1 is returned, indicating that the pattern matches only at the start of a
1228 subject string or after any newline within the string. Otherwise -2 is
1229 returned. For anchored patterns, -2 is returned.
1230 </P>
1231 <P>
1232 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1233 to return the full 32-bit range of the character, this value is deprecated;
1234 instead the PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER values
1235 should be used.
1236 <pre>
1237 PCRE_INFO_FIRSTTABLE
1238 </pre>
1239 If the pattern was studied, and this resulted in the construction of a 256-bit
1240 table indicating a fixed set of values for the first data unit in any matching
1241 string, a pointer to the table is returned. Otherwise NULL is returned. The
1242 fourth argument should point to an <b>unsigned char *</b> variable.
1243 <pre>
1244 PCRE_INFO_HASCRORLF
1245 </pre>
1246 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1247 otherwise 0. The fourth argument should point to an <b>int</b> variable. An
1248 explicit match is either a literal CR or LF character, or \r or \n.
1249 <pre>
1250 PCRE_INFO_JCHANGED
1251 </pre>
1252 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1253 0. The fourth argument should point to an <b>int</b> variable. (?J) and
1254 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1255 <pre>
1256 PCRE_INFO_JIT
1257 </pre>
1258 Return 1 if the pattern was studied with one of the JIT options, and
1259 just-in-time compiling was successful. The fourth argument should point to an
1260 <b>int</b> variable. A return value of 0 means that JIT support is not available
1261 in this version of PCRE, or that the pattern was not studied with a JIT option,
1262 or that the JIT compiler could not handle this particular pattern. See the
1263 <a href="pcrejit.html"><b>pcrejit</b></a>
1264 documentation for details of what can and cannot be handled.
1265 <pre>
1266 PCRE_INFO_JITSIZE
1267 </pre>
1268 If the pattern was successfully studied with a JIT option, return the size of
1269 the JIT compiled code, otherwise return zero. The fourth argument should point
1270 to a <b>size_t</b> variable.
1271 <pre>
1272 PCRE_INFO_LASTLITERAL
1273 </pre>
1274 Return the value of the rightmost literal data unit that must exist in any
1275 matched string, other than at its start, if such a value has been recorded. The
1276 fourth argument should point to an <b>int</b> variable. If there is no such
1277 value, -1 is returned. For anchored patterns, a last literal value is recorded
1278 only if it follows something of variable length. For example, for the pattern
1279 /^a\d+z\d+/ the returned value is "z", but for /^a\dz\d/ the returned value
1280 is -1.
1281 </P>
1282 <P>
1283 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1284 to return the full 32-bit range of the character, this value is deprecated;
1285 instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1286 be used.
1287 <pre>
1288 PCRE_INFO_MAXLOOKBEHIND
1289 </pre>
1290 Return the number of characters (NB not bytes) in the longest lookbehind
1291 assertion in the pattern. This information is useful when doing multi-segment
1292 matching using the partial matching facilities. Note that the simple assertions
1293 \b and \B require a one-character lookbehind. \A also registers a
1294 one-character lookbehind, though it does not actually inspect the previous
1295 character. This is to ensure that at least one character from the old segment
1296 is retained when a new segment is processed. Otherwise, if there are no
1297 lookbehinds in the pattern, \A might match incorrectly at the start of a new
1298 segment.
1299 <pre>
1300 PCRE_INFO_MINLENGTH
1301 </pre>
1302 If the pattern was studied and a minimum length for matching subject strings
1303 was computed, its value is returned. Otherwise the returned value is -1. The
1304 value is a number of characters, which in UTF-8 mode may be different from the
1305 number of bytes. The fourth argument should point to an <b>int</b> variable. A
1306 non-negative value is a lower bound to the length of any matching string. There
1307 may not be any strings of that length that do actually match, but every string
1308 that does match is at least that long.
1309 <pre>
1310 PCRE_INFO_NAMECOUNT
1311 PCRE_INFO_NAMEENTRYSIZE
1312 PCRE_INFO_NAMETABLE
1313 </pre>
1314 PCRE supports the use of named as well as numbered capturing parentheses. The
1315 names are just an additional way of identifying the parentheses, which still
1316 acquire numbers. Several convenience functions such as
1317 <b>pcre_get_named_substring()</b> are provided for extracting captured
1318 substrings by name. It is also possible to extract the data directly, by first
1319 converting the name to a number in order to access the correct pointers in the
1320 output vector (described with <b>pcre_exec()</b> below). To do the conversion,
1321 you need to use the name-to-number map, which is described by these three
1322 values.
1323 </P>
1324 <P>
1325 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1326 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1327 entry; both of these return an <b>int</b> value. The entry size depends on the
1328 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1329 entry of the table. This is a pointer to <b>char</b> in the 8-bit library, where
1330 the first two bytes of each entry are the number of the capturing parenthesis,
1331 most significant byte first. In the 16-bit library, the pointer points to
1332 16-bit data units, the first of which contains the parenthesis number.
1333 In the 32-bit library, the pointer points to 32-bit data units, the first of
1334 which contains the parenthesis number. The rest
1335 of the entry is the corresponding name, zero terminated.
1336 </P>
1337 <P>
1338 The names are in alphabetical order. Duplicate names may appear if (?| is used
1339 to create multiple groups with the same number, as described in the
1340 <a href="pcrepattern.html#dupsubpatternnumber">section on duplicate subpattern numbers</a>
1341 in the
1342 <a href="pcrepattern.html"><b>pcrepattern</b></a>
1343 page. Duplicate names for subpatterns with different numbers are permitted only
1344 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1345 table in the order in which they were found in the pattern. In the absence of
1346 (?| this is the order of increasing number; when (?| is used this is not
1347 necessarily the case because later subpatterns may have lower numbers.
1348 </P>
1349 <P>
1350 As a simple example of the name/number table, consider the following pattern
1351 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1352 space - including newlines - is ignored):
1353 <pre>
1354 (?&#60;date&#62; (?&#60;year&#62;(\d\d)?\d\d) - (?&#60;month&#62;\d\d) - (?&#60;day&#62;\d\d) )
1355 </pre>
1356 There are four named subpatterns, so the table has four entries, and each entry
1357 in the table is eight bytes long. The table is as follows, with non-printing
1358 bytes shows in hexadecimal, and undefined bytes shown as ??:
1359 <pre>
1360 00 01 d a t e 00 ??
1361 00 05 d a y 00 ?? ??
1362 00 04 m o n t h 00
1363 00 02 y e a r 00 ??
1364 </pre>
1365 When writing code to extract data from named subpatterns using the
1366 name-to-number map, remember that the length of the entries is likely to be
1367 different for each compiled pattern.
1368 <pre>
1369 PCRE_INFO_OKPARTIAL
1370 </pre>
1371 Return 1 if the pattern can be used for partial matching with
1372 <b>pcre_exec()</b>, otherwise 0. The fourth argument should point to an
1373 <b>int</b> variable. From release 8.00, this always returns 1, because the
1374 restrictions that previously applied to partial matching have been lifted. The
1375 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1376 documentation gives details of partial matching.
1377 <pre>
1378 PCRE_INFO_OPTIONS
1379 </pre>
1380 Return a copy of the options with which the pattern was compiled. The fourth
1381 argument should point to an <b>unsigned long int</b> variable. These option bits
1382 are those specified in the call to <b>pcre_compile()</b>, modified by any
1383 top-level option settings at the start of the pattern itself. In other words,
1384 they are the options that will be in force when matching starts. For example,
1385 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1386 result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1387 </P>
1388 <P>
1389 A pattern is automatically anchored by PCRE if all of its top-level
1390 alternatives begin with one of the following:
1391 <pre>
1392 ^ unless PCRE_MULTILINE is set
1393 \A always
1394 \G always
1395 .* if PCRE_DOTALL is set and there are no back references to the subpattern in which .* appears
1396 </pre>
1397 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1398 <b>pcre_fullinfo()</b>.
1399 <pre>
1400 PCRE_INFO_SIZE
1401 </pre>
1402 Return the size of the compiled pattern in bytes (for both libraries). The
1403 fourth argument should point to a <b>size_t</b> variable. This value does not
1404 include the size of the <b>pcre</b> structure that is returned by
1405 <b>pcre_compile()</b>. The value that is passed as the argument to
1406 <b>pcre_malloc()</b> when <b>pcre_compile()</b> is getting memory in which to
1407 place the compiled data is the value returned by this option plus the size of
1408 the <b>pcre</b> structure. Studying a compiled pattern, with or without JIT,
1409 does not alter the value returned by this option.
1410 <pre>
1411 PCRE_INFO_STUDYSIZE
1412 </pre>
1413 Return the size in bytes of the data block pointed to by the <i>study_data</i>
1414 field in a <b>pcre_extra</b> block. If <b>pcre_extra</b> is NULL, or there is no
1415 study data, zero is returned. The fourth argument should point to a
1416 <b>size_t</b> variable. The <i>study_data</i> field is set by <b>pcre_study()</b>
1417 to record information that will speed up matching (see the section entitled
1418 <a href="#studyingapattern">"Studying a pattern"</a>
1419 above). The format of the <i>study_data</i> block is private, but its length
1420 is made available via this option so that it can be saved and restored (see the
1421 <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1422 documentation for details).
1423 <pre>
1424 PCRE_INFO_FIRSTCHARACTERFLAGS
1425 </pre>
1426 Return information about the first data unit of any matched string, for a
1427 non-anchored pattern. The fourth argument should point to an <b>int</b>
1428 variable.
1429 </P>
1430 <P>
1431 If there is a fixed first value, for example, the letter "c" from a pattern
1432 such as (cat|cow|coyote), 1 is returned, and the character value can be
1433 retrieved using PCRE_INFO_FIRSTCHARACTER.
1434 </P>
1435 <P>
1436 If there is no fixed first value, and if either
1437 <br>
1438 <br>
1439 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1440 starts with "^", or
1441 <br>
1442 <br>
1443 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1444 (if it were set, the pattern would be anchored),
1445 <br>
1446 <br>
1447 2 is returned, indicating that the pattern matches only at the start of a
1448 subject string or after any newline within the string. Otherwise 0 is
1449 returned. For anchored patterns, 0 is returned.
1450 <pre>
1451 PCRE_INFO_FIRSTCHARACTER
1452 </pre>
1453 Return the fixed first character value, if PCRE_INFO_FIRSTCHARACTERFLAGS
1454 returned 1; otherwise returns 0. The fourth argument should point to an
1455 <b>uint_t</b> variable.
1456 </P>
1457 <P>
1458 In the 8-bit library, the value is always less than 256. In the 16-bit library
1459 the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1460 can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1461 </P>
1462 <P>
1463 If there is no fixed first value, and if either
1464 <br>
1465 <br>
1466 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1467 starts with "^", or
1468 <br>
1469 <br>
1470 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1471 (if it were set, the pattern would be anchored),
1472 <br>
1473 <br>
1474 -1 is returned, indicating that the pattern matches only at the start of a
1475 subject string or after any newline within the string. Otherwise -2 is
1476 returned. For anchored patterns, -2 is returned.
1477 <pre>
1478 PCRE_INFO_REQUIREDCHARFLAGS
1479 </pre>
1480 Returns 1 if there is a rightmost literal data unit that must exist in any
1481 matched string, other than at its start. The fourth argument should point to
1482 an <b>int</b> variable. If there is no such value, 0 is returned. If returning
1483 1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1484 </P>
1485 <P>
1486 For anchored patterns, a last literal value is recorded only if it follows
1487 something of variable length. For example, for the pattern /^a\d+z\d+/ the
1488 returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1489 /^a\dz\d/ the returned value is 0.
1490 <pre>
1491 PCRE_INFO_REQUIREDCHAR
1492 </pre>
1493 Return the value of the rightmost literal data unit that must exist in any
1494 matched string, other than at its start, if such a value has been recorded. The
1495 fourth argument should point to an <b>uint32_t</b> variable. If there is no such
1496 value, 0 is returned.
1497 </P>
1498 <br><a name="SEC16" href="#TOC1">REFERENCE COUNTS</a><br>
1499 <P>
1500 <b>int pcre_refcount(pcre *<i>code</i>, int <i>adjust</i>);</b>
1501 </P>
1502 <P>
1503 The <b>pcre_refcount()</b> function is used to maintain a reference count in the
1504 data block that contains a compiled pattern. It is provided for the benefit of
1505 applications that operate in an object-oriented manner, where different parts
1506 of the application may be using the same compiled pattern, but you want to free
1507 the block when they are all done.
1508 </P>
1509 <P>
1510 When a pattern is compiled, the reference count field is initialized to zero.
1511 It is changed only by calling this function, whose action is to add the
1512 <i>adjust</i> value (which may be positive or negative) to it. The yield of the
1513 function is the new value. However, the value of the count is constrained to
1514 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1515 it is forced to the appropriate limit value.
1516 </P>
1517 <P>
1518 Except when it is zero, the reference count is not correctly preserved if a
1519 pattern is compiled on one host and then transferred to a host whose byte-order
1520 is different. (This seems a highly unlikely scenario.)
1521 </P>
1522 <br><a name="SEC17" href="#TOC1">MATCHING A PATTERN: THE TRADITIONAL FUNCTION</a><br>
1523 <P>
1524 <b>int pcre_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
1525 <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
1526 <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>);</b>
1527 </P>
1528 <P>
1529 The function <b>pcre_exec()</b> is called to match a subject string against a
1530 compiled pattern, which is passed in the <i>code</i> argument. If the
1531 pattern was studied, the result of the study should be passed in the
1532 <i>extra</i> argument. You can call <b>pcre_exec()</b> with the same <i>code</i>
1533 and <i>extra</i> arguments as many times as you like, in order to match
1534 different subject strings with the same pattern.
1535 </P>
1536 <P>
1537 This function is the main matching facility of the library, and it operates in
1538 a Perl-like manner. For specialist use there is also an alternative matching
1539 function, which is described
1540 <a href="#dfamatch">below</a>
1541 in the section about the <b>pcre_dfa_exec()</b> function.
1542 </P>
1543 <P>
1544 In most applications, the pattern will have been compiled (and optionally
1545 studied) in the same process that calls <b>pcre_exec()</b>. However, it is
1546 possible to save compiled patterns and study data, and then use them later
1547 in different processes, possibly even on different hosts. For a discussion
1548 about this, see the
1549 <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1550 documentation.
1551 </P>
1552 <P>
1553 Here is an example of a simple call to <b>pcre_exec()</b>:
1554 <pre>
1555 int rc;
1556 int ovector[30];
1557 rc = pcre_exec(
1558 re, /* result of pcre_compile() */
1559 NULL, /* we didn't study the pattern */
1560 "some string", /* the subject string */
1561 11, /* the length of the subject string */
1562 0, /* start at offset 0 in the subject */
1563 0, /* default options */
1564 ovector, /* vector of integers for substring information */
1565 30); /* number of elements (NOT size in bytes) */
1566 <a name="extradata"></a></PRE>
1567 </P>
1568 <br><b>
1569 Extra data for <b>pcre_exec()</b>
1570 </b><br>
1571 <P>
1572 If the <i>extra</i> argument is not NULL, it must point to a <b>pcre_extra</b>
1573 data block. The <b>pcre_study()</b> function returns such a block (when it
1574 doesn't return NULL), but you can also create one for yourself, and pass
1575 additional information in it. The <b>pcre_extra</b> block contains the following
1576 fields (not necessarily in this order):
1577 <pre>
1578 unsigned long int <i>flags</i>;
1579 void *<i>study_data</i>;
1580 void *<i>executable_jit</i>;
1581 unsigned long int <i>match_limit</i>;
1582 unsigned long int <i>match_limit_recursion</i>;
1583 void *<i>callout_data</i>;
1584 const unsigned char *<i>tables</i>;
1585 unsigned char **<i>mark</i>;
1586 </pre>
1587 In the 16-bit version of this structure, the <i>mark</i> field has type
1588 "PCRE_UCHAR16 **".
1589 <br>
1590 <br>
1591 In the 32-bit version of this structure, the <i>mark</i> field has type
1592 "PCRE_UCHAR32 **".
1593 </P>
1594 <P>
1595 The <i>flags</i> field is used to specify which of the other fields are set. The
1596 flag bits are:
1597 <pre>
1598 PCRE_EXTRA_CALLOUT_DATA
1599 PCRE_EXTRA_EXECUTABLE_JIT
1600 PCRE_EXTRA_MARK
1601 PCRE_EXTRA_MATCH_LIMIT
1602 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1603 PCRE_EXTRA_STUDY_DATA
1604 PCRE_EXTRA_TABLES
1605 </pre>
1606 Other flag bits should be set to zero. The <i>study_data</i> field and sometimes
1607 the <i>executable_jit</i> field are set in the <b>pcre_extra</b> block that is
1608 returned by <b>pcre_study()</b>, together with the appropriate flag bits. You
1609 should not set these yourself, but you may add to the block by setting other
1610 fields and their corresponding flag bits.
1611 </P>
1612 <P>
1613 The <i>match_limit</i> field provides a means of preventing PCRE from using up a
1614 vast amount of resources when running patterns that are not going to match,
1615 but which have a very large number of possibilities in their search trees. The
1616 classic example is a pattern that uses nested unlimited repeats.
1617 </P>
1618 <P>
1619 Internally, <b>pcre_exec()</b> uses a function called <b>match()</b>, which it
1620 calls repeatedly (sometimes recursively). The limit set by <i>match_limit</i> is
1621 imposed on the number of times this function is called during a match, which
1622 has the effect of limiting the amount of backtracking that can take place. For
1623 patterns that are not anchored, the count restarts from zero for each position
1624 in the subject string.
1625 </P>
1626 <P>
1627 When <b>pcre_exec()</b> is called with a pattern that was successfully studied
1628 with a JIT option, the way that the matching is executed is entirely different.
1629 However, there is still the possibility of runaway matching that goes on for a
1630 very long time, and so the <i>match_limit</i> value is also used in this case
1631 (but in a different way) to limit how long the matching can continue.
1632 </P>
1633 <P>
1634 The default value for the limit can be set when PCRE is built; the default
1635 default is 10 million, which handles all but the most extreme cases. You can
1636 override the default by suppling <b>pcre_exec()</b> with a <b>pcre_extra</b>
1637 block in which <i>match_limit</i> is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1638 the <i>flags</i> field. If the limit is exceeded, <b>pcre_exec()</b> returns
1639 PCRE_ERROR_MATCHLIMIT.
1640 </P>
1641 <P>
1642 The <i>match_limit_recursion</i> field is similar to <i>match_limit</i>, but
1643 instead of limiting the total number of times that <b>match()</b> is called, it
1644 limits the depth of recursion. The recursion depth is a smaller number than the
1645 total number of calls, because not all calls to <b>match()</b> are recursive.
1646 This limit is of use only if it is set smaller than <i>match_limit</i>.
1647 </P>
1648 <P>
1649 Limiting the recursion depth limits the amount of machine stack that can be
1650 used, or, when PCRE has been compiled to use memory on the heap instead of the
1651 stack, the amount of heap memory that can be used. This limit is not relevant,
1652 and is ignored, when matching is done using JIT compiled code.
1653 </P>
1654 <P>
1655 The default value for <i>match_limit_recursion</i> can be set when PCRE is
1656 built; the default default is the same value as the default for
1657 <i>match_limit</i>. You can override the default by suppling <b>pcre_exec()</b>
1658 with a <b>pcre_extra</b> block in which <i>match_limit_recursion</i> is set, and
1659 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the <i>flags</i> field. If the limit
1660 is exceeded, <b>pcre_exec()</b> returns PCRE_ERROR_RECURSIONLIMIT.
1661 </P>
1662 <P>
1663 The <i>callout_data</i> field is used in conjunction with the "callout" feature,
1664 and is described in the
1665 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1666 documentation.
1667 </P>
1668 <P>
1669 The <i>tables</i> field is used to pass a character tables pointer to
1670 <b>pcre_exec()</b>; this overrides the value that is stored with the compiled
1671 pattern. A non-NULL value is stored with the compiled pattern only if custom
1672 tables were supplied to <b>pcre_compile()</b> via its <i>tableptr</i> argument.
1673 If NULL is passed to <b>pcre_exec()</b> using this mechanism, it forces PCRE's
1674 internal tables to be used. This facility is helpful when re-using patterns
1675 that have been saved after compiling with an external set of tables, because
1676 the external tables might be at a different address when <b>pcre_exec()</b> is
1677 called. See the
1678 <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1679 documentation for a discussion of saving compiled patterns for later use.
1680 </P>
1681 <P>
1682 If PCRE_EXTRA_MARK is set in the <i>flags</i> field, the <i>mark</i> field must
1683 be set to point to a suitable variable. If the pattern contains any
1684 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1685 a name to pass back, a pointer to the name string (zero terminated) is placed
1686 in the variable pointed to by the <i>mark</i> field. The names are within the
1687 compiled pattern; if you wish to retain such a name you must copy it before
1688 freeing the memory of a compiled pattern. If there is no name to pass back, the
1689 variable pointed to by the <i>mark</i> field is set to NULL. For details of the
1690 backtracking control verbs, see the section entitled
1691 <a href="pcrepattern#backtrackcontrol">"Backtracking control"</a>
1692 in the
1693 <a href="pcrepattern.html"><b>pcrepattern</b></a>
1694 documentation.
1695 <a name="execoptions"></a></P>
1696 <br><b>
1697 Option bits for <b>pcre_exec()</b>
1698 </b><br>
1699 <P>
1700 The unused bits of the <i>options</i> argument for <b>pcre_exec()</b> must be
1701 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
1702 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1703 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1704 PCRE_PARTIAL_SOFT.
1705 </P>
1706 <P>
1707 If the pattern was successfully studied with one of the just-in-time (JIT)
1708 compile options, the only supported options for JIT execution are
1709 PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1710 PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1711 unsupported option is used, JIT execution is disabled and the normal
1712 interpretive code in <b>pcre_exec()</b> is run.
1713 <pre>
1714 PCRE_ANCHORED
1715 </pre>
1716 The PCRE_ANCHORED option limits <b>pcre_exec()</b> to matching at the first
1717 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1718 to be anchored by virtue of its contents, it cannot be made unachored at
1719 matching time.
1720 <pre>
1721 PCRE_BSR_ANYCRLF
1722 PCRE_BSR_UNICODE
1723 </pre>
1724 These options (which are mutually exclusive) control what the \R escape
1725 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1726 match any Unicode newline sequence. These options override the choice that was
1727 made or defaulted when the pattern was compiled.
1728 <pre>
1729 PCRE_NEWLINE_CR
1730 PCRE_NEWLINE_LF
1731 PCRE_NEWLINE_CRLF
1732 PCRE_NEWLINE_ANYCRLF
1733 PCRE_NEWLINE_ANY
1734 </pre>
1735 These options override the newline definition that was chosen or defaulted when
1736 the pattern was compiled. For details, see the description of
1737 <b>pcre_compile()</b> above. During matching, the newline choice affects the
1738 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1739 the way the match position is advanced after a match failure for an unanchored
1740 pattern.
1741 </P>
1742 <P>
1743 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1744 match attempt for an unanchored pattern fails when the current position is at a
1745 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1746 characters, the match position is advanced by two characters instead of one, in
1747 other words, to after the CRLF.
1748 </P>
1749 <P>
1750 The above rule is a compromise that makes the most common cases work as
1751 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1752 set), it does not match the string "\r\nA" because, after failing at the
1753 start, it skips both the CR and the LF before retrying. However, the pattern
1754 [\r\n]A does match that string, because it contains an explicit CR or LF
1755 reference, and so advances only by one character after the first failure.
1756 </P>
1757 <P>
1758 An explicit match for CR of LF is either a literal appearance of one of those
1759 characters, or one of the \r or \n escape sequences. Implicit matches such as
1760 [^X] do not count, nor does \s (which includes CR and LF in the characters
1761 that it matches).
1762 </P>
1763 <P>
1764 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1765 valid newline sequence and explicit \r or \n escapes appear in the pattern.
1766 <pre>
1767 PCRE_NOTBOL
1768 </pre>
1769 This option specifies that first character of the subject string is not the
1770 beginning of a line, so the circumflex metacharacter should not match before
1771 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1772 never to match. This option affects only the behaviour of the circumflex
1773 metacharacter. It does not affect \A.
1774 <pre>
1775 PCRE_NOTEOL
1776 </pre>
1777 This option specifies that the end of the subject string is not the end of a
1778 line, so the dollar metacharacter should not match it nor (except in multiline
1779 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1780 compile time) causes dollar never to match. This option affects only the
1781 behaviour of the dollar metacharacter. It does not affect \Z or \z.
1782 <pre>
1783 PCRE_NOTEMPTY
1784 </pre>
1785 An empty string is not considered to be a valid match if this option is set. If
1786 there are alternatives in the pattern, they are tried. If all the alternatives
1787 match the empty string, the entire match fails. For example, if the pattern
1788 <pre>
1789 a?b?
1790 </pre>
1791 is applied to a string not beginning with "a" or "b", it matches an empty
1792 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1793 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1794 <pre>
1795 PCRE_NOTEMPTY_ATSTART
1796 </pre>
1797 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1798 the start of the subject is permitted. If the pattern is anchored, such a match
1799 can occur only if the pattern contains \K.
1800 </P>
1801 <P>
1802 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1803 does make a special case of a pattern match of the empty string within its
1804 <b>split()</b> function, and when using the /g modifier. It is possible to
1805 emulate Perl's behaviour after matching a null string by first trying the match
1806 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1807 if that fails, by advancing the starting offset (see below) and trying an
1808 ordinary match again. There is some code that demonstrates how to do this in
1809 the
1810 <a href="pcredemo.html"><b>pcredemo</b></a>
1811 sample program. In the most general case, you have to check to see if the
1812 newline convention recognizes CRLF as a newline, and if so, and the current
1813 character is CR followed by LF, advance the starting offset by two characters
1814 instead of one.
1815 <pre>
1816 PCRE_NO_START_OPTIMIZE
1817 </pre>
1818 There are a number of optimizations that <b>pcre_exec()</b> uses at the start of
1819 a match, in order to speed up the process. For example, if it is known that an
1820 unanchored match must start with a specific character, it searches the subject
1821 for that character, and fails immediately if it cannot find it, without
1822 actually running the main matching function. This means that a special item
1823 such as (*COMMIT) at the start of a pattern is not considered until after a
1824 suitable starting point for the match has been found. When callouts or (*MARK)
1825 items are in use, these "start-up" optimizations can cause them to be skipped
1826 if the pattern is never actually used. The start-up optimizations are in effect
1827 a pre-scan of the subject that takes place before the pattern is run.
1828 </P>
1829 <P>
1830 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1831 causing performance to suffer, but ensuring that in cases where the result is
1832 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1833 are considered at every possible starting position in the subject string. If
1834 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1835 time. The use of PCRE_NO_START_OPTIMIZE disables JIT execution; when it is set,
1836 matching is always done using interpretively.
1837 </P>
1838 <P>
1839 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1840 Consider the pattern
1841 <pre>
1842 (*COMMIT)ABC
1843 </pre>
1844 When this is compiled, PCRE records the fact that a match must start with the
1845 character "A". Suppose the subject string is "DEFABC". The start-up
1846 optimization scans along the subject, finds "A" and runs the first match
1847 attempt from there. The (*COMMIT) item means that the pattern must match the
1848 current starting position, which in this case, it does. However, if the same
1849 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1850 subject string does not happen. The first match attempt is run starting from
1851 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1852 the overall result is "no match". If the pattern is studied, more start-up
1853 optimizations may be used. For example, a minimum length for the subject may be
1854 recorded. Consider the pattern
1855 <pre>
1856 (*MARK:A)(X|Y)
1857 </pre>
1858 The minimum length for a match is one character. If the subject is "ABC", there
1859 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1860 If the pattern is studied, the final attempt does not take place, because PCRE
1861 knows that the subject is too short, and so the (*MARK) is never encountered.
1862 In this case, studying the pattern does not affect the overall match result,
1863 which is still "no match", but it does affect the auxiliary information that is
1864 returned.
1865 <pre>
1866 PCRE_NO_UTF8_CHECK
1867 </pre>
1868 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1869 string is automatically checked when <b>pcre_exec()</b> is subsequently called.
1870 The entire string is checked before any other processing takes place. The value
1871 of <i>startoffset</i> is also checked to ensure that it points to the start of a
1872 UTF-8 character. There is a discussion about the
1873 <a href="pcreunicode.html#utf8strings">validity of UTF-8 strings</a>
1874 in the
1875 <a href="pcreunicode.html"><b>pcreunicode</b></a>
1876 page. If an invalid sequence of bytes is found, <b>pcre_exec()</b> returns the
1877 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1878 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1879 cases, information about the precise nature of the error may also be returned
1880 (see the descriptions of these errors in the section entitled \fIError return
1881 values from\fP <b>pcre_exec()</b>
1882 <a href="#errorlist">below).</a>
1883 If <i>startoffset</i> contains a value that does not point to the start of a
1884 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1885 returned.
1886 </P>
1887 <P>
1888 If you already know that your subject is valid, and you want to skip these
1889 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1890 calling <b>pcre_exec()</b>. You might want to do this for the second and
1891 subsequent calls to <b>pcre_exec()</b> if you are making repeated calls to find
1892 all the matches in a single subject string. However, you should be sure that
1893 the value of <i>startoffset</i> points to the start of a character (or the end
1894 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1895 invalid string as a subject or an invalid value of <i>startoffset</i> is
1896 undefined. Your program may crash.
1897 <pre>
1898 PCRE_PARTIAL_HARD
1899 PCRE_PARTIAL_SOFT
1900 </pre>
1901 These options turn on the partial matching feature. For backwards
1902 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1903 occurs if the end of the subject string is reached successfully, but there are
1904 not enough subject characters to complete the match. If this happens when
1905 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1906 testing any remaining alternatives. Only if no complete match can be found is
1907 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1908 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1909 but only if no complete match can be found.
1910 </P>
1911 <P>
1912 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1913 partial match is found, <b>pcre_exec()</b> immediately returns
1914 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1915 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1916 important that an alternative complete match.
1917 </P>
1918 <P>
1919 In both cases, the portion of the string that was inspected when the partial
1920 match was found is set as the first matching string. There is a more detailed
1921 discussion of partial and multi-segment matching, with examples, in the
1922 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1923 documentation.
1924 </P>
1925 <br><b>
1926 The string to be matched by <b>pcre_exec()</b>
1927 </b><br>
1928 <P>
1929 The subject string is passed to <b>pcre_exec()</b> as a pointer in
1930 <i>subject</i>, a length in bytes in <i>length</i>, and a starting byte offset
1931 in <i>startoffset</i>. If this is negative or greater than the length of the
1932 subject, <b>pcre_exec()</b> returns PCRE_ERROR_BADOFFSET. When the starting
1933 offset is zero, the search for a match starts at the beginning of the subject,
1934 and this is by far the most common case. In UTF-8 mode, the byte offset must
1935 point to the start of a UTF-8 character (or the end of the subject). Unlike the
1936 pattern string, the subject may contain binary zero bytes.
1937 </P>
1938 <P>
1939 A non-zero starting offset is useful when searching for another match in the
1940 same subject by calling <b>pcre_exec()</b> again after a previous success.
1941 Setting <i>startoffset</i> differs from just passing over a shortened string and
1942 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1943 lookbehind. For example, consider the pattern
1944 <pre>
1945 \Biss\B
1946 </pre>
1947 which finds occurrences of "iss" in the middle of words. (\B matches only if
1948 the current position in the subject is not a word boundary.) When applied to
1949 the string "Mississipi" the first call to <b>pcre_exec()</b> finds the first
1950 occurrence. If <b>pcre_exec()</b> is called again with just the remainder of the
1951 subject, namely "issipi", it does not match, because \B is always false at the
1952 start of the subject, which is deemed to be a word boundary. However, if
1953 <b>pcre_exec()</b> is passed the entire string again, but with <i>startoffset</i>
1954 set to 4, it finds the second occurrence of "iss" because it is able to look
1955 behind the starting point to discover that it is preceded by a letter.
1956 </P>
1957 <P>
1958 Finding all the matches in a subject is tricky when the pattern can match an
1959 empty string. It is possible to emulate Perl's /g behaviour by first trying the
1960 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1961 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1962 and trying an ordinary match again. There is some code that demonstrates how to
1963 do this in the
1964 <a href="pcredemo.html"><b>pcredemo</b></a>
1965 sample program. In the most general case, you have to check to see if the
1966 newline convention recognizes CRLF as a newline, and if so, and the current
1967 character is CR followed by LF, advance the starting offset by two characters
1968 instead of one.
1969 </P>
1970 <P>
1971 If a non-zero starting offset is passed when the pattern is anchored, one
1972 attempt to match at the given offset is made. This can only succeed if the
1973 pattern does not require the match to be at the start of the subject.
1974 </P>
1975 <br><b>
1976 How <b>pcre_exec()</b> returns captured substrings
1977 </b><br>
1978 <P>
1979 In general, a pattern matches a certain portion of the subject, and in
1980 addition, further substrings from the subject may be picked out by parts of the
1981 pattern. Following the usage in Jeffrey Friedl's book, this is called
1982 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1983 a fragment of a pattern that picks out a substring. PCRE supports several other
1984 kinds of parenthesized subpattern that do not cause substrings to be captured.
1985 </P>
1986 <P>
1987 Captured substrings are returned to the caller via a vector of integers whose
1988 address is passed in <i>ovector</i>. The number of elements in the vector is
1989 passed in <i>ovecsize</i>, which must be a non-negative number. <b>Note</b>: this
1990 argument is NOT the size of <i>ovector</i> in bytes.
1991 </P>
1992 <P>
1993 The first two-thirds of the vector is used to pass back captured substrings,
1994 each substring using a pair of integers. The remaining third of the vector is
1995 used as workspace by <b>pcre_exec()</b> while matching capturing subpatterns,
1996 and is not available for passing back information. The number passed in
1997 <i>ovecsize</i> should always be a multiple of three. If it is not, it is
1998 rounded down.
1999 </P>
2000 <P>
2001 When a match is successful, information about captured substrings is returned
2002 in pairs of integers, starting at the beginning of <i>ovector</i>, and
2003 continuing up to two-thirds of its length at the most. The first element of
2004 each pair is set to the byte offset of the first character in a substring, and
2005 the second is set to the byte offset of the first character after the end of a
2006 substring. <b>Note</b>: these values are always byte offsets, even in UTF-8
2007 mode. They are not character counts.
2008 </P>
2009 <P>
2010 The first pair of integers, <i>ovector[0]</i> and <i>ovector[1]</i>, identify the
2011 portion of the subject string matched by the entire pattern. The next pair is
2012 used for the first capturing subpattern, and so on. The value returned by
2013 <b>pcre_exec()</b> is one more than the highest numbered pair that has been set.
2014 For example, if two substrings have been captured, the returned value is 3. If
2015 there are no capturing subpatterns, the return value from a successful match is
2016 1, indicating that just the first pair of offsets has been set.
2017 </P>
2018 <P>
2019 If a capturing subpattern is matched repeatedly, it is the last portion of the
2020 string that it matched that is returned.
2021 </P>
2022 <P>
2023 If the vector is too small to hold all the captured substring offsets, it is
2024 used as far as possible (up to two-thirds of its length), and the function
2025 returns a value of zero. If neither the actual string matched nor any captured
2026 substrings are of interest, <b>pcre_exec()</b> may be called with <i>ovector</i>
2027 passed as NULL and <i>ovecsize</i> as zero. However, if the pattern contains
2028 back references and the <i>ovector</i> is not big enough to remember the related
2029 substrings, PCRE has to get additional memory for use during matching. Thus it
2030 is usually advisable to supply an <i>ovector</i> of reasonable size.
2031 </P>
2032 <P>
2033 There are some cases where zero is returned (indicating vector overflow) when
2034 in fact the vector is exactly the right size for the final match. For example,
2035 consider the pattern
2036 <pre>
2037 (a)(?:(b)c|bd)
2038 </pre>
2039 If a vector of 6 elements (allowing for only 1 captured substring) is given
2040 with subject string "abd", <b>pcre_exec()</b> will try to set the second
2041 captured string, thereby recording a vector overflow, before failing to match
2042 "c" and backing up to try the second alternative. The zero return, however,
2043 does correctly indicate that the maximum number of slots (namely 2) have been
2044 filled. In similar cases where there is temporary overflow, but the final
2045 number of used slots is actually less than the maximum, a non-zero value is
2046 returned.
2047 </P>
2048 <P>
2049 The <b>pcre_fullinfo()</b> function can be used to find out how many capturing
2050 subpatterns there are in a compiled pattern. The smallest size for
2051 <i>ovector</i> that will allow for <i>n</i> captured substrings, in addition to
2052 the offsets of the substring matched by the whole pattern, is (<i>n</i>+1)*3.
2053 </P>
2054 <P>
2055 It is possible for capturing subpattern number <i>n+1</i> to match some part of
2056 the subject when subpattern <i>n</i> has not been used at all. For example, if
2057 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2058 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2059 happens, both values in the offset pairs corresponding to unused subpatterns
2060 are set to -1.
2061 </P>
2062 <P>
2063 Offset values that correspond to unused subpatterns at the end of the
2064 expression are also set to -1. For example, if the string "abc" is matched
2065 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2066 return from the function is 2, because the highest used capturing subpattern
2067 number is 1, and the offsets for for the second and third capturing subpatterns
2068 (assuming the vector is large enough, of course) are set to -1.
2069 </P>
2070 <P>
2071 <b>Note</b>: Elements in the first two-thirds of <i>ovector</i> that do not
2072 correspond to capturing parentheses in the pattern are never changed. That is,
2073 if a pattern contains <i>n</i> capturing parentheses, no more than
2074 <i>ovector[0]</i> to <i>ovector[2n+1]</i> are set by <b>pcre_exec()</b>. The other
2075 elements (in the first two-thirds) retain whatever values they previously had.
2076 </P>
2077 <P>
2078 Some convenience functions are provided for extracting the captured substrings
2079 as separate strings. These are described below.
2080 <a name="errorlist"></a></P>
2081 <br><b>
2082 Error return values from <b>pcre_exec()</b>
2083 </b><br>
2084 <P>
2085 If <b>pcre_exec()</b> fails, it returns a negative number. The following are
2086 defined in the header file:
2087 <pre>
2088 PCRE_ERROR_NOMATCH (-1)
2089 </pre>
2090 The subject string did not match the pattern.
2091 <pre>
2092 PCRE_ERROR_NULL (-2)
2093 </pre>
2094 Either <i>code</i> or <i>subject</i> was passed as NULL, or <i>ovector</i> was
2095 NULL and <i>ovecsize</i> was not zero.
2096 <pre>
2097 PCRE_ERROR_BADOPTION (-3)
2098 </pre>
2099 An unrecognized bit was set in the <i>options</i> argument.
2100 <pre>
2101 PCRE_ERROR_BADMAGIC (-4)
2102 </pre>
2103 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2104 the case when it is passed a junk pointer and to detect when a pattern that was
2105 compiled in an environment of one endianness is run in an environment with the
2106 other endianness. This is the error that PCRE gives when the magic number is
2107 not present.
2108 <pre>
2109 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2110 </pre>
2111 While running the pattern match, an unknown item was encountered in the
2112 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2113 of the compiled pattern.
2114 <pre>
2115 PCRE_ERROR_NOMEMORY (-6)
2116 </pre>
2117 If a pattern contains back references, but the <i>ovector</i> that is passed to
2118 <b>pcre_exec()</b> is not big enough to remember the referenced substrings, PCRE
2119 gets a block of memory at the start of matching to use for this purpose. If the
2120 call via <b>pcre_malloc()</b> fails, this error is given. The memory is
2121 automatically freed at the end of matching.
2122 </P>
2123 <P>
2124 This error is also given if <b>pcre_stack_malloc()</b> fails in
2125 <b>pcre_exec()</b>. This can happen only when PCRE has been compiled with
2126 <b>--disable-stack-for-recursion</b>.
2127 <pre>
2128 PCRE_ERROR_NOSUBSTRING (-7)
2129 </pre>
2130 This error is used by the <b>pcre_copy_substring()</b>,
2131 <b>pcre_get_substring()</b>, and <b>pcre_get_substring_list()</b> functions (see
2132 below). It is never returned by <b>pcre_exec()</b>.
2133 <pre>
2134 PCRE_ERROR_MATCHLIMIT (-8)
2135 </pre>
2136 The backtracking limit, as specified by the <i>match_limit</i> field in a
2137 <b>pcre_extra</b> structure (or defaulted) was reached. See the description
2138 above.
2139 <pre>
2140 PCRE_ERROR_CALLOUT (-9)
2141 </pre>
2142 This error is never generated by <b>pcre_exec()</b> itself. It is provided for
2143 use by callout functions that want to yield a distinctive error code. See the
2144 <a href="pcrecallout.html"><b>pcrecallout</b></a>
2145 documentation for details.
2146 <pre>
2147 PCRE_ERROR_BADUTF8 (-10)
2148 </pre>
2149 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2150 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2151 (<i>ovecsize</i>) is at least 2, the byte offset to the start of the the invalid
2152 UTF-8 character is placed in the first element, and a reason code is placed in
2153 the second element. The reason codes are listed in the
2154 <a href="#badutf8reasons">following section.</a>
2155 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2156 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2157 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2158 <pre>
2159 PCRE_ERROR_BADUTF8_OFFSET (-11)
2160 </pre>
2161 The UTF-8 byte sequence that was passed as a subject was checked and found to
2162 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2163 <i>startoffset</i> did not point to the beginning of a UTF-8 character or the
2164 end of the subject.
2165 <pre>
2166 PCRE_ERROR_PARTIAL (-12)
2167 </pre>
2168 The subject string did not match, but it did match partially. See the
2169 <a href="pcrepartial.html"><b>pcrepartial</b></a>
2170 documentation for details of partial matching.
2171 <pre>
2172 PCRE_ERROR_BADPARTIAL (-13)
2173 </pre>
2174 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2175 option was used with a compiled pattern containing items that were not
2176 supported for partial matching. From release 8.00 onwards, there are no
2177 restrictions on partial matching.
2178 <pre>
2179 PCRE_ERROR_INTERNAL (-14)
2180 </pre>
2181 An unexpected internal error has occurred. This error could be caused by a bug
2182 in PCRE or by overwriting of the compiled pattern.
2183 <pre>
2184 PCRE_ERROR_BADCOUNT (-15)
2185 </pre>
2186 This error is given if the value of the <i>ovecsize</i> argument is negative.
2187 <pre>
2188 PCRE_ERROR_RECURSIONLIMIT (-21)
2189 </pre>
2190 The internal recursion limit, as specified by the <i>match_limit_recursion</i>
2191 field in a <b>pcre_extra</b> structure (or defaulted) was reached. See the
2192 description above.
2193 <pre>
2194 PCRE_ERROR_BADNEWLINE (-23)
2195 </pre>
2196 An invalid combination of PCRE_NEWLINE_<i>xxx</i> options was given.
2197 <pre>
2198 PCRE_ERROR_BADOFFSET (-24)
2199 </pre>
2200 The value of <i>startoffset</i> was negative or greater than the length of the
2201 subject, that is, the value in <i>length</i>.
2202 <pre>
2203 PCRE_ERROR_SHORTUTF8 (-25)
2204 </pre>
2205 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2206 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2207 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2208 fact sufficient to detect this case, but this special error code for
2209 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2210 retained for backwards compatibility.
2211 <pre>
2212 PCRE_ERROR_RECURSELOOP (-26)
2213 </pre>
2214 This error is returned when <b>pcre_exec()</b> detects a recursion loop within
2215 the pattern. Specifically, it means that either the whole pattern or a
2216 subpattern has been called recursively for the second time at the same position
2217 in the subject string. Some simple patterns that might do this are detected and
2218 faulted at compile time, but more complicated cases, in particular mutual
2219 recursions between two different subpatterns, cannot be detected until run
2220 time.
2221 <pre>
2222 PCRE_ERROR_JIT_STACKLIMIT (-27)
2223 </pre>
2224 This error is returned when a pattern that was successfully studied using a
2225 JIT compile option is being matched, but the memory available for the
2226 just-in-time processing stack is not large enough. See the
2227 <a href="pcrejit.html"><b>pcrejit</b></a>
2228 documentation for more details.
2229 <pre>
2230 PCRE_ERROR_BADMODE (-28)
2231 </pre>
2232 This error is given if a pattern that was compiled by the 8-bit library is
2233 passed to a 16-bit or 32-bit library function, or vice versa.
2234 <pre>
2235 PCRE_ERROR_BADENDIANNESS (-29)
2236 </pre>
2237 This error is given if a pattern that was compiled and saved is reloaded on a
2238 host with different endianness. The utility function
2239 <b>pcre_pattern_to_host_byte_order()</b> can be used to convert such a pattern
2240 so that it runs on the new host.
2241 <pre>
2242 PCRE_ERROR_JIT_BADOPTION
2243 </pre>
2244 This error is returned when a pattern that was successfully studied using a JIT
2245 compile option is being matched, but the matching mode (partial or complete
2246 match) does not correspond to any JIT compilation mode. When the JIT fast path
2247 function is used, this error may be also given for invalid options. See the
2248 <a href="pcrejit.html"><b>pcrejit</b></a>
2249 documentation for more details.
2250 <pre>
2251 PCRE_ERROR_BADLENGTH (-32)
2252 </pre>
2253 This error is given if <b>pcre_exec()</b> is called with a negative value for
2254 the <i>length</i> argument.
2255 </P>
2256 <P>
2257 Error numbers -16 to -20, -22, and 30 are not used by <b>pcre_exec()</b>.
2258 <a name="badutf8reasons"></a></P>
2259 <br><b>
2260 Reason codes for invalid UTF-8 strings
2261 </b><br>
2262 <P>
2263 This section applies only to the 8-bit library. The corresponding information
2264 for the 16-bit and 32-bit libraries is given in the
2265 <a href="pcre16.html"><b>pcre16</b></a>
2266 and
2267 <a href="pcre32.html"><b>pcre32</b></a>
2268 pages.
2269 </P>
2270 <P>
2271 When <b>pcre_exec()</b> returns either PCRE_ERROR_BADUTF8 or
2272 PCRE_ERROR_SHORTUTF8, and the size of the output vector (<i>ovecsize</i>) is at
2273 least 2, the offset of the start of the invalid UTF-8 character is placed in
2274 the first output vector element (<i>ovector[0]</i>) and a reason code is placed
2275 in the second element (<i>ovector[1]</i>). The reason codes are given names in
2276 the <b>pcre.h</b> header file:
2277 <pre>
2278 PCRE_UTF8_ERR1
2279 PCRE_UTF8_ERR2
2280 PCRE_UTF8_ERR3
2281 PCRE_UTF8_ERR4
2282 PCRE_UTF8_ERR5
2283 </pre>
2284 The string ends with a truncated UTF-8 character; the code specifies how many
2285 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2286 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2287 allows for up to 6 bytes, and this is checked first; hence the possibility of
2288 4 or 5 missing bytes.
2289 <pre>
2290 PCRE_UTF8_ERR6
2291 PCRE_UTF8_ERR7
2292 PCRE_UTF8_ERR8
2293 PCRE_UTF8_ERR9
2294 PCRE_UTF8_ERR10
2295 </pre>
2296 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2297 character do not have the binary value 0b10 (that is, either the most
2298 significant bit is 0, or the next bit is 1).
2299 <pre>
2300 PCRE_UTF8_ERR11
2301 PCRE_UTF8_ERR12
2302 </pre>
2303 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2304 these code points are excluded by RFC 3629.
2305 <pre>
2306 PCRE_UTF8_ERR13
2307 </pre>
2308 A 4-byte character has a value greater than 0x10fff; these code points are
2309 excluded by RFC 3629.
2310 <pre>
2311 PCRE_UTF8_ERR14
2312 </pre>
2313 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2314 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2315 from UTF-8.
2316 <pre>
2317 PCRE_UTF8_ERR15
2318 PCRE_UTF8_ERR16
2319 PCRE_UTF8_ERR17
2320 PCRE_UTF8_ERR18
2321 PCRE_UTF8_ERR19
2322 </pre>
2323 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2324 value that can be represented by fewer bytes, which is invalid. For example,
2325 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2326 one byte.
2327 <pre>
2328 PCRE_UTF8_ERR20
2329 </pre>
2330 The two most significant bits of the first byte of a character have the binary
2331 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2332 byte can only validly occur as the second or subsequent byte of a multi-byte
2333 character.
2334 <pre>
2335 PCRE_UTF8_ERR21
2336 </pre>
2337 The first byte of a character has the value 0xfe or 0xff. These values can
2338 never occur in a valid UTF-8 string.
2339 <pre>
2340 PCRE_UTF8_ERR22
2341 </pre>
2342 This error code was formerly used when the presence of a so-called
2343 "non-character" caused an error. Unicode corrigendum #9 makes it clear that
2344 such characters should not cause a string to be rejected, and so this code is
2345 no longer in use and is never returned.
2346 </P>
2347 <br><a name="SEC18" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NUMBER</a><br>
2348 <P>
2349 <b>int pcre_copy_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
2350 <b>int <i>stringcount</i>, int <i>stringnumber</i>, char *<i>buffer</i>,</b>
2351 <b>int <i>buffersize</i>);</b>
2352 </P>
2353 <P>
2354 <b>int pcre_get_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
2355 <b>int <i>stringcount</i>, int <i>stringnumber</i>,</b>
2356 <b>const char **<i>stringptr</i>);</b>
2357 </P>
2358 <P>
2359 <b>int pcre_get_substring_list(const char *<i>subject</i>,</b>
2360 <b>int *<i>ovector</i>, int <i>stringcount</i>, const char ***<i>listptr</i>);</b>
2361 </P>
2362 <P>
2363 Captured substrings can be accessed directly by using the offsets returned by
2364 <b>pcre_exec()</b> in <i>ovector</i>. For convenience, the functions
2365 <b>pcre_copy_substring()</b>, <b>pcre_get_substring()</b>, and
2366 <b>pcre_get_substring_list()</b> are provided for extracting captured substrings
2367 as new, separate, zero-terminated strings. These functions identify substrings
2368 by number. The next section describes functions for extracting named
2369 substrings.
2370 </P>
2371 <P>
2372 A substring that contains a binary zero is correctly extracted and has a
2373 further zero added on the end, but the result is not, of course, a C string.
2374 However, you can process such a string by referring to the length that is
2375 returned by <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>.
2376 Unfortunately, the interface to <b>pcre_get_substring_list()</b> is not adequate
2377 for handling strings containing binary zeros, because the end of the final
2378 string is not independently indicated.
2379 </P>
2380 <P>
2381 The first three arguments are the same for all three of these functions:
2382 <i>subject</i> is the subject string that has just been successfully matched,
2383 <i>ovector</i> is a pointer to the vector of integer offsets that was passed to
2384 <b>pcre_exec()</b>, and <i>stringcount</i> is the number of substrings that were
2385 captured by the match, including the substring that matched the entire regular
2386 expression. This is the value returned by <b>pcre_exec()</b> if it is greater
2387 than zero. If <b>pcre_exec()</b> returned zero, indicating that it ran out of
2388 space in <i>ovector</i>, the value passed as <i>stringcount</i> should be the
2389 number of elements in the vector divided by three.
2390 </P>
2391 <P>
2392 The functions <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>
2393 extract a single substring, whose number is given as <i>stringnumber</i>. A
2394 value of zero extracts the substring that matched the entire pattern, whereas
2395 higher values extract the captured substrings. For <b>pcre_copy_substring()</b>,
2396 the string is placed in <i>buffer</i>, whose length is given by
2397 <i>buffersize</i>, while for <b>pcre_get_substring()</b> a new block of memory is
2398 obtained via <b>pcre_malloc</b>, and its address is returned via
2399 <i>stringptr</i>. The yield of the function is the length of the string, not
2400 including the terminating zero, or one of these error codes:
2401 <pre>
2402 PCRE_ERROR_NOMEMORY (-6)
2403 </pre>
2404 The buffer was too small for <b>pcre_copy_substring()</b>, or the attempt to get
2405 memory failed for <b>pcre_get_substring()</b>.
2406 <pre>
2407 PCRE_ERROR_NOSUBSTRING (-7)
2408 </pre>
2409 There is no substring whose number is <i>stringnumber</i>.
2410 </P>
2411 <P>
2412 The <b>pcre_get_substring_list()</b> function extracts all available substrings
2413 and builds a list of pointers to them. All this is done in a single block of
2414 memory that is obtained via <b>pcre_malloc</b>. The address of the memory block
2415 is returned via <i>listptr</i>, which is also the start of the list of string
2416 pointers. The end of the list is marked by a NULL pointer. The yield of the
2417 function is zero if all went well, or the error code
2418 <pre>
2419 PCRE_ERROR_NOMEMORY (-6)
2420 </pre>
2421 if the attempt to get the memory block failed.
2422 </P>
2423 <P>
2424 When any of these functions encounter a substring that is unset, which can
2425 happen when capturing subpattern number <i>n+1</i> matches some part of the
2426 subject, but subpattern <i>n</i> has not been used at all, they return an empty
2427 string. This can be distinguished from a genuine zero-length substring by
2428 inspecting the appropriate offset in <i>ovector</i>, which is negative for unset
2429 substrings.
2430 </P>
2431 <P>
2432 The two convenience functions <b>pcre_free_substring()</b> and
2433 <b>pcre_free_substring_list()</b> can be used to free the memory returned by
2434 a previous call of <b>pcre_get_substring()</b> or
2435 <b>pcre_get_substring_list()</b>, respectively. They do nothing more than call
2436 the function pointed to by <b>pcre_free</b>, which of course could be called
2437 directly from a C program. However, PCRE is used in some situations where it is
2438 linked via a special interface to another programming language that cannot use
2439 <b>pcre_free</b> directly; it is for these cases that the functions are
2440 provided.
2441 </P>
2442 <br><a name="SEC19" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NAME</a><br>
2443 <P>
2444 <b>int pcre_get_stringnumber(const pcre *<i>code</i>,</b>
2445 <b>const char *<i>name</i>);</b>
2446 </P>
2447 <P>
2448 <b>int pcre_copy_named_substring(const pcre *<i>code</i>,</b>
2449 <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
2450 <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
2451 <b>char *<i>buffer</i>, int <i>buffersize</i>);</b>
2452 </P>
2453 <P>
2454 <b>int pcre_get_named_substring(const pcre *<i>code</i>,</b>
2455 <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
2456 <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
2457 <b>const char **<i>stringptr</i>);</b>
2458 </P>
2459 <P>
2460 To extract a substring by name, you first have to find associated number.
2461 For example, for this pattern
2462 <pre>
2463 (a+)b(?&#60;xxx&#62;\d+)...
2464 </pre>
2465 the number of the subpattern called "xxx" is 2. If the name is known to be
2466 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2467 calling <b>pcre_get_stringnumber()</b>. The first argument is the compiled
2468 pattern, and the second is the name. The yield of the function is the
2469 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2470 that name.
2471 </P>
2472 <P>
2473 Given the number, you can extract the substring directly, or use one of the
2474 functions described in the previous section. For convenience, there are also
2475 two functions that do the whole job.
2476 </P>
2477 <P>
2478 Most of the arguments of <b>pcre_copy_named_substring()</b> and
2479 <b>pcre_get_named_substring()</b> are the same as those for the similarly named
2480 functions that extract by number. As these are described in the previous
2481 section, they are not re-described here. There are just two differences:
2482 </P>
2483 <P>
2484 First, instead of a substring number, a substring name is given. Second, there
2485 is an extra argument, given at the start, which is a pointer to the compiled
2486 pattern. This is needed in order to gain access to the name-to-number
2487 translation table.
2488 </P>
2489 <P>
2490 These functions call <b>pcre_get_stringnumber()</b>, and if it succeeds, they
2491 then call <b>pcre_copy_substring()</b> or <b>pcre_get_substring()</b>, as
2492 appropriate. <b>NOTE:</b> If PCRE_DUPNAMES is set and there are duplicate names,
2493 the behaviour may not be what you want (see the next section).
2494 </P>
2495 <P>
2496 <b>Warning:</b> If the pattern uses the (?| feature to set up multiple
2497 subpatterns with the same number, as described in the
2498 <a href="pcrepattern.html#dupsubpatternnumber">section on duplicate subpattern numbers</a>
2499 in the
2500 <a href="pcrepattern.html"><b>pcrepattern</b></a>
2501 page, you cannot use names to distinguish the different subpatterns, because
2502 names are not included in the compiled code. The matching process uses only
2503 numbers. For this reason, the use of different names for subpatterns of the
2504 same number causes an error at compile time.
2505 </P>
2506 <br><a name="SEC20" href="#TOC1">DUPLICATE SUBPATTERN NAMES</a><br>
2507 <P>
2508 <b>int pcre_get_stringtable_entries(const pcre *<i>code</i>,</b>
2509 <b>const char *<i>name</i>, char **<i>first</i>, char **<i>last</i>);</b>
2510 </P>
2511 <P>
2512 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2513 are not required to be unique. (Duplicate names are always allowed for
2514 subpatterns with the same number, created by using the (?| feature. Indeed, if
2515 such subpatterns are named, they are required to use the same names.)
2516 </P>
2517 <P>
2518 Normally, patterns with duplicate names are such that in any one match, only
2519 one of the named subpatterns participates. An example is shown in the
2520 <a href="pcrepattern.html"><b>pcrepattern</b></a>
2521 documentation.
2522 </P>
2523 <P>
2524 When duplicates are present, <b>pcre_copy_named_substring()</b> and
2525 <b>pcre_get_named_substring()</b> return the first substring corresponding to
2526 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2527 returned; no data is returned. The <b>pcre_get_stringnumber()</b> function
2528 returns one of the numbers that are associated with the name, but it is not
2529 defined which it is.
2530 </P>
2531 <P>
2532 If you want to get full details of all captured substrings for a given name,
2533 you must use the <b>pcre_get_stringtable_entries()</b> function. The first
2534 argument is the compiled pattern, and the second is the name. The third and
2535 fourth are pointers to variables which are updated by the function. After it
2536 has run, they point to the first and last entries in the name-to-number table
2537 for the given name. The function itself returns the length of each entry, or
2538 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2539 described above in the section entitled <i>Information about a pattern</i>
2540 <a href="#infoaboutpattern">above.</a>
2541 Given all the relevant entries for the name, you can extract each of their
2542 numbers, and hence the captured data, if any.
2543 </P>
2544 <br><a name="SEC21" href="#TOC1">FINDING ALL POSSIBLE MATCHES</a><br>
2545 <P>
2546 The traditional matching function uses a similar algorithm to Perl, which stops
2547 when it finds the first match, starting at a given point in the subject. If you
2548 want to find all possible matches, or the longest possible match, consider
2549 using the alternative matching function (see below) instead. If you cannot use
2550 the alternative function, but still need to find all possible matches, you
2551 can kludge it up by making use of the callout facility, which is described in
2552 the
2553 <a href="pcrecallout.html"><b>pcrecallout</b></a>
2554 documentation.
2555 </P>
2556 <P>
2557 What you have to do is to insert a callout right at the end of the pattern.
2558 When your callout function is called, extract and save the current matched
2559 substring. Then return 1, which forces <b>pcre_exec()</b> to backtrack and try
2560 other alternatives. Ultimately, when it runs out of matches, <b>pcre_exec()</b>
2561 will yield PCRE_ERROR_NOMATCH.
2562 </P>
2563 <br><a name="SEC22" href="#TOC1">OBTAINING AN ESTIMATE OF STACK USAGE</a><br>
2564 <P>
2565 Matching certain patterns using <b>pcre_exec()</b> can use a lot of process
2566 stack, which in certain environments can be rather limited in size. Some users
2567 find it helpful to have an estimate of the amount of stack that is used by
2568 <b>pcre_exec()</b>, to help them set recursion limits, as described in the
2569 <a href="pcrestack.html"><b>pcrestack</b></a>
2570 documentation. The estimate that is output by <b>pcretest</b> when called with
2571 the <b>-m</b> and <b>-C</b> options is obtained by calling <b>pcre_exec</b> with
2572 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2573 </P>
2574 <P>
2575 Normally, if its first argument is NULL, <b>pcre_exec()</b> immediately returns
2576 the negative error code PCRE_ERROR_NULL, but with this special combination of
2577 arguments, it returns instead a negative number whose absolute value is the
2578 approximate stack frame size in bytes. (A negative number is used so that it is
2579 clear that no match has happened.) The value is approximate because in some
2580 cases, recursive calls to <b>pcre_exec()</b> occur when there are one or two
2581 additional variables on the stack.
2582 </P>
2583 <P>
2584 If PCRE has been compiled to use the heap instead of the stack for recursion,
2585 the value returned is the size of each block that is obtained from the heap.
2586 <a name="dfamatch"></a></P>
2587 <br><a name="SEC23" href="#TOC1">MATCHING A PATTERN: THE ALTERNATIVE FUNCTION</a><br>
2588 <P>
2589 <b>int pcre_dfa_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
2590 <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
2591 <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>,</b>
2592 <b>int *<i>workspace</i>, int <i>wscount</i>);</b>
2593 </P>
2594 <P>
2595 The function <b>pcre_dfa_exec()</b> is called to match a subject string against
2596 a compiled pattern, using a matching algorithm that scans the subject string
2597 just once, and does not backtrack. This has different characteristics to the
2598 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2599 patterns are not supported. Nevertheless, there are times when this kind of
2600 matching can be useful. For a discussion of the two matching algorithms, and a
2601 list of features that <b>pcre_dfa_exec()</b> does not support, see the
2602 <a href="pcrematching.html"><b>pcrematching</b></a>
2603 documentation.
2604 </P>
2605 <P>
2606 The arguments for the <b>pcre_dfa_exec()</b> function are the same as for
2607 <b>pcre_exec()</b>, plus two extras. The <i>ovector</i> argument is used in a
2608 different way, and this is described below. The other common arguments are used
2609 in the same way as for <b>pcre_exec()</b>, so their description is not repeated
2610 here.
2611 </P>
2612 <P>
2613 The two additional arguments provide workspace for the function. The workspace
2614 vector should contain at least 20 elements. It is used for keeping track of
2615 multiple paths through the pattern tree. More workspace will be needed for
2616 patterns and subjects where there are a lot of potential matches.
2617 </P>
2618 <P>
2619 Here is an example of a simple call to <b>pcre_dfa_exec()</b>:
2620 <pre>
2621 int rc;
2622 int ovector[10];
2623 int wspace[20];
2624 rc = pcre_dfa_exec(
2625 re, /* result of pcre_compile() */
2626 NULL, /* we didn't study the pattern */
2627 "some string", /* the subject string */
2628 11, /* the length of the subject string */
2629 0, /* start at offset 0 in the subject */
2630 0, /* default options */
2631 ovector, /* vector of integers for substring information */
2632 10, /* number of elements (NOT size in bytes) */
2633 wspace, /* working space vector */
2634 20); /* number of elements (NOT size in bytes) */
2635 </PRE>
2636 </P>
2637 <br><b>
2638 Option bits for <b>pcre_dfa_exec()</b>
2639 </b><br>
2640 <P>
2641 The unused bits of the <i>options</i> argument for <b>pcre_dfa_exec()</b> must be
2642 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
2643 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2644 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2645 PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2646 All but the last four of these are exactly the same as for <b>pcre_exec()</b>,
2647 so their description is not repeated here.
2648 <pre>
2649 PCRE_PARTIAL_HARD
2650 PCRE_PARTIAL_SOFT
2651 </pre>
2652 These have the same general effect as they do for <b>pcre_exec()</b>, but the
2653 details are slightly different. When PCRE_PARTIAL_HARD is set for
2654 <b>pcre_dfa_exec()</b>, it returns PCRE_ERROR_PARTIAL if the end of the subject
2655 is reached and there is still at least one matching possibility that requires
2656 additional characters. This happens even if some complete matches have also
2657 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2658 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2659 there have been no complete matches, but there is still at least one matching
2660 possibility. The portion of the string that was inspected when the longest
2661 partial match was found is set as the first matching string in both cases.
2662 There is a more detailed discussion of partial and multi-segment matching, with
2663 examples, in the
2664 <a href="pcrepartial.html"><b>pcrepartial</b></a>
2665 documentation.
2666 <pre>
2667 PCRE_DFA_SHORTEST
2668 </pre>
2669 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2670 soon as it has found one match. Because of the way the alternative algorithm
2671 works, this is necessarily the shortest possible match at the first possible
2672 matching point in the subject string.
2673 <pre>
2674 PCRE_DFA_RESTART
2675 </pre>
2676 When <b>pcre_dfa_exec()</b> returns a partial match, it is possible to call it
2677 again, with additional subject characters, and have it continue with the same
2678 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2679 <i>workspace</i> and <i>wscount</i> options must reference the same vector as
2680 before because data about the match so far is left in them after a partial
2681 match. There is more discussion of this facility in the
2682 <a href="pcrepartial.html"><b>pcrepartial</b></a>
2683 documentation.
2684 </P>
2685 <br><b>
2686 Successful returns from <b>pcre_dfa_exec()</b>
2687 </b><br>
2688 <P>
2689 When <b>pcre_dfa_exec()</b> succeeds, it may have matched more than one
2690 substring in the subject. Note, however, that all the matches from one run of
2691 the function start at the same point in the subject. The shorter matches are
2692 all initial substrings of the longer matches. For example, if the pattern
2693 <pre>
2694 &#60;.*&#62;
2695 </pre>
2696 is matched against the string
2697 <pre>
2698 This is &#60;something&#62; &#60;something else&#62; &#60;something further&#62; no more
2699 </pre>
2700 the three matched strings are
2701 <pre>
2702 &#60;something&#62;
2703 &#60;something&#62; &#60;something else&#62;
2704 &#60;something&#62; &#60;something else&#62; &#60;something further&#62;
2705 </pre>
2706 On success, the yield of the function is a number greater than zero, which is
2707 the number of matched substrings. The substrings themselves are returned in
2708 <i>ovector</i>. Each string uses two elements; the first is the offset to the
2709 start, and the second is the offset to the end. In fact, all the strings have
2710 the same start offset. (Space could have been saved by giving this only once,
2711 but it was decided to retain some compatibility with the way <b>pcre_exec()</b>
2712 returns data, even though the meaning of the strings is different.)
2713 </P>
2714 <P>
2715 The strings are returned in reverse order of length; that is, the longest
2716 matching string is given first. If there were too many matches to fit into
2717 <i>ovector</i>, the yield of the function is zero, and the vector is filled with
2718 the longest matches. Unlike <b>pcre_exec()</b>, <b>pcre_dfa_exec()</b> can use
2719 the entire <i>ovector</i> for returning matched strings.
2720 </P>
2721 <br><b>
2722 Error returns from <b>pcre_dfa_exec()</b>
2723 </b><br>
2724 <P>
2725 The <b>pcre_dfa_exec()</b> function returns a negative number when it fails.
2726 Many of the errors are the same as for <b>pcre_exec()</b>, and these are
2727 described
2728 <a href="#errorlist">above.</a>
2729 There are in addition the following errors that are specific to
2730 <b>pcre_dfa_exec()</b>:
2731 <pre>
2732 PCRE_ERROR_DFA_UITEM (-16)
2733 </pre>
2734 This return is given if <b>pcre_dfa_exec()</b> encounters an item in the pattern
2735 that it does not support, for instance, the use of \C or a back reference.
2736 <pre>
2737 PCRE_ERROR_DFA_UCOND (-17)
2738 </pre>
2739 This return is given if <b>pcre_dfa_exec()</b> encounters a condition item that
2740 uses a back reference for the condition, or a test for recursion in a specific
2741 group. These are not supported.
2742 <pre>
2743 PCRE_ERROR_DFA_UMLIMIT (-18)
2744 </pre>
2745 This return is given if <b>pcre_dfa_exec()</b> is called with an <i>extra</i>
2746 block that contains a setting of the <i>match_limit</i> or
2747 <i>match_limit_recursion</i> fields. This is not supported (these fields are
2748 meaningless for DFA matching).
2749 <pre>
2750 PCRE_ERROR_DFA_WSSIZE (-19)
2751 </pre>
2752 This return is given if <b>pcre_dfa_exec()</b> runs out of space in the
2753 <i>workspace</i> vector.
2754 <pre>
2755 PCRE_ERROR_DFA_RECURSE (-20)
2756 </pre>
2757 When a recursive subpattern is processed, the matching function calls itself
2758 recursively, using private vectors for <i>ovector</i> and <i>workspace</i>. This
2759 error is given if the output vector is not large enough. This should be
2760 extremely rare, as a vector of size 1000 is used.
2761 <pre>
2762 PCRE_ERROR_DFA_BADRESTART (-30)
2763 </pre>
2764 When <b>pcre_dfa_exec()</b> is called with the <b>PCRE_DFA_RESTART</b> option,
2765 some plausibility checks are made on the contents of the workspace, which
2766 should contain data about the previous partial match. If any of these checks
2767 fail, this error is given.
2768 </P>
2769 <br><a name="SEC24" href="#TOC1">SEE ALSO</a><br>
2770 <P>
2771 <b>pcre16</b>(3), <b>pcre32</b>(3), <b>pcrebuild</b>(3), <b>pcrecallout</b>(3),
2772 <b>pcrecpp(3)</b>(3), <b>pcrematching</b>(3), <b>pcrepartial</b>(3),
2773 <b>pcreposix</b>(3), <b>pcreprecompile</b>(3), <b>pcresample</b>(3),
2774 <b>pcrestack</b>(3).
2775 </P>
2776 <br><a name="SEC25" href="#TOC1">AUTHOR</a><br>
2777 <P>
2778 Philip Hazel
2779 <br>
2780 University Computing Service
2781 <br>
2782 Cambridge CB2 3QH, England.
2783 <br>
2784 </P>
2785 <br><a name="SEC26" href="#TOC1">REVISION</a><br>
2786 <P>
2787 Last updated: 27 February 2013
2788 <br>
2789 Copyright &copy; 1997-2013 University of Cambridge.
2790 <br>
2791 <p>
2792 Return to the <a href="index.html">PCRE index page</a>.
2793 </p>

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5