/[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 835 - (hide annotations) (download) (as text)
Wed Dec 28 16:10:09 2011 UTC (2 years, 9 months ago) by ph10
File MIME type: text/html
File size: 114819 byte(s)
Rolled back trunk to r755 to prepare for merging the 16-bit branch.

1 nigel 63 <html>
2     <head>
3     <title>pcreapi specification</title>
4     </head>
5     <body bgcolor="#FFFFFF" text="#00005A" link="#0066FF" alink="#3399FF" vlink="#2222BB">
6 nigel 75 <h1>pcreapi man page</h1>
7     <p>
8     Return to the <a href="index.html">PCRE index page</a>.
9     </p>
10 ph10 111 <p>
11 nigel 75 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 ph10 111 <br>
15 nigel 63 <ul>
16 ph10 691 <li><a name="TOC1" href="#SEC1">PCRE NATIVE API BASIC FUNCTIONS</a>
17     <li><a name="TOC2" href="#SEC2">PCRE NATIVE API AUXILIARY FUNCTIONS</a>
18     <li><a name="TOC3" href="#SEC3">PCRE NATIVE API INDIRECTED FUNCTIONS</a>
19     <li><a name="TOC4" href="#SEC4">PCRE API OVERVIEW</a>
20     <li><a name="TOC5" href="#SEC5">NEWLINES</a>
21     <li><a name="TOC6" href="#SEC6">MULTITHREADING</a>
22     <li><a name="TOC7" href="#SEC7">SAVING PRECOMPILED PATTERNS FOR LATER USE</a>
23     <li><a name="TOC8" href="#SEC8">CHECKING BUILD-TIME OPTIONS</a>
24     <li><a name="TOC9" href="#SEC9">COMPILING A PATTERN</a>
25     <li><a name="TOC10" href="#SEC10">COMPILATION ERROR CODES</a>
26     <li><a name="TOC11" href="#SEC11">STUDYING A PATTERN</a>
27     <li><a name="TOC12" href="#SEC12">LOCALE SUPPORT</a>
28     <li><a name="TOC13" href="#SEC13">INFORMATION ABOUT A PATTERN</a>
29     <li><a name="TOC14" href="#SEC14">OBSOLETE INFO FUNCTION</a>
30     <li><a name="TOC15" href="#SEC15">REFERENCE COUNTS</a>
31     <li><a name="TOC16" href="#SEC16">MATCHING A PATTERN: THE TRADITIONAL FUNCTION</a>
32     <li><a name="TOC17" href="#SEC17">EXTRACTING CAPTURED SUBSTRINGS BY NUMBER</a>
33     <li><a name="TOC18" href="#SEC18">EXTRACTING CAPTURED SUBSTRINGS BY NAME</a>
34     <li><a name="TOC19" href="#SEC19">DUPLICATE SUBPATTERN NAMES</a>
35     <li><a name="TOC20" href="#SEC20">FINDING ALL POSSIBLE MATCHES</a>
36     <li><a name="TOC21" href="#SEC21">MATCHING A PATTERN: THE ALTERNATIVE FUNCTION</a>
37     <li><a name="TOC22" href="#SEC22">SEE ALSO</a>
38     <li><a name="TOC23" href="#SEC23">AUTHOR</a>
39     <li><a name="TOC24" href="#SEC24">REVISION</a>
40 nigel 63 </ul>
41 ph10 691 <br><a name="SEC1" href="#TOC1">PCRE NATIVE API BASIC FUNCTIONS</a><br>
42 nigel 63 <P>
43     <b>#include &#60;pcre.h&#62;</b>
44     </P>
45     <P>
46     <b>pcre *pcre_compile(const char *<i>pattern</i>, int <i>options</i>,</b>
47     <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
48     <b>const unsigned char *<i>tableptr</i>);</b>
49     </P>
50     <P>
51 nigel 77 <b>pcre *pcre_compile2(const char *<i>pattern</i>, int <i>options</i>,</b>
52     <b>int *<i>errorcodeptr</i>,</b>
53     <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
54     <b>const unsigned char *<i>tableptr</i>);</b>
55     </P>
56     <P>
57 nigel 63 <b>pcre_extra *pcre_study(const pcre *<i>code</i>, int <i>options</i>,</b>
58     <b>const char **<i>errptr</i>);</b>
59     </P>
60     <P>
61 ph10 691 <b>void pcre_free_study(pcre_extra *<i>extra</i>);</b>
62     </P>
63     <P>
64 nigel 63 <b>int pcre_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
65     <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
66     <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>);</b>
67     </P>
68 ph10 691 <br><a name="SEC2" href="#TOC1">PCRE NATIVE API AUXILIARY FUNCTIONS</a><br>
69 nigel 63 <P>
70 ph10 691 <b>pcre_jit_stack *pcre_jit_stack_alloc(int <i>startsize</i>, int <i>maxsize</i>);</b>
71     </P>
72     <P>
73     <b>void pcre_jit_stack_free(pcre_jit_stack *<i>stack</i>);</b>
74     </P>
75     <P>
76 ph10 692 <b>void pcre_assign_jit_stack(pcre_extra *<i>extra</i>,</b>
77 ph10 691 <b>pcre_jit_callback <i>callback</i>, void *<i>data</i>);</b>
78     </P>
79     <P>
80 nigel 77 <b>int pcre_dfa_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
81     <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
82     <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>,</b>
83     <b>int *<i>workspace</i>, int <i>wscount</i>);</b>
84     </P>
85     <P>
86 nigel 63 <b>int pcre_copy_named_substring(const pcre *<i>code</i>,</b>
87     <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
88     <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
89     <b>char *<i>buffer</i>, int <i>buffersize</i>);</b>
90     </P>
91     <P>
92     <b>int pcre_copy_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
93     <b>int <i>stringcount</i>, int <i>stringnumber</i>, char *<i>buffer</i>,</b>
94     <b>int <i>buffersize</i>);</b>
95     </P>
96     <P>
97     <b>int pcre_get_named_substring(const pcre *<i>code</i>,</b>
98     <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
99     <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
100     <b>const char **<i>stringptr</i>);</b>
101     </P>
102     <P>
103     <b>int pcre_get_stringnumber(const pcre *<i>code</i>,</b>
104     <b>const char *<i>name</i>);</b>
105     </P>
106     <P>
107 nigel 91 <b>int pcre_get_stringtable_entries(const pcre *<i>code</i>,</b>
108     <b>const char *<i>name</i>, char **<i>first</i>, char **<i>last</i>);</b>
109     </P>
110     <P>
111 nigel 63 <b>int pcre_get_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
112     <b>int <i>stringcount</i>, int <i>stringnumber</i>,</b>
113     <b>const char **<i>stringptr</i>);</b>
114     </P>
115     <P>
116     <b>int pcre_get_substring_list(const char *<i>subject</i>,</b>
117     <b>int *<i>ovector</i>, int <i>stringcount</i>, const char ***<i>listptr</i>);</b>
118     </P>
119     <P>
120     <b>void pcre_free_substring(const char *<i>stringptr</i>);</b>
121     </P>
122     <P>
123     <b>void pcre_free_substring_list(const char **<i>stringptr</i>);</b>
124     </P>
125     <P>
126     <b>const unsigned char *pcre_maketables(void);</b>
127     </P>
128     <P>
129     <b>int pcre_fullinfo(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
130     <b>int <i>what</i>, void *<i>where</i>);</b>
131     </P>
132     <P>
133     <b>int pcre_info(const pcre *<i>code</i>, int *<i>optptr</i>, int</b>
134     <b>*<i>firstcharptr</i>);</b>
135     </P>
136     <P>
137 nigel 77 <b>int pcre_refcount(pcre *<i>code</i>, int <i>adjust</i>);</b>
138     </P>
139     <P>
140 nigel 63 <b>int pcre_config(int <i>what</i>, void *<i>where</i>);</b>
141     </P>
142     <P>
143     <b>char *pcre_version(void);</b>
144     </P>
145 ph10 691 <br><a name="SEC3" href="#TOC1">PCRE NATIVE API INDIRECTED FUNCTIONS</a><br>
146 nigel 63 <P>
147     <b>void *(*pcre_malloc)(size_t);</b>
148     </P>
149     <P>
150     <b>void (*pcre_free)(void *);</b>
151     </P>
152     <P>
153 nigel 73 <b>void *(*pcre_stack_malloc)(size_t);</b>
154     </P>
155     <P>
156     <b>void (*pcre_stack_free)(void *);</b>
157     </P>
158     <P>
159 nigel 63 <b>int (*pcre_callout)(pcre_callout_block *);</b>
160     </P>
161 ph10 691 <br><a name="SEC4" href="#TOC1">PCRE API OVERVIEW</a><br>
162 nigel 63 <P>
163 nigel 93 PCRE has its own native API, which is described in this document. There are
164     also some wrapper functions that correspond to the POSIX regular expression
165 ph10 691 API, but they do not give access to all the functionality. They are described
166     in the
167 nigel 75 <a href="pcreposix.html"><b>pcreposix</b></a>
168 nigel 77 documentation. Both of these APIs define a set of C function calls. A C++
169 ph10 691 wrapper is also distributed with PCRE. It is documented in the
170 nigel 77 <a href="pcrecpp.html"><b>pcrecpp</b></a>
171     page.
172 nigel 63 </P>
173     <P>
174 nigel 77 The native API C function prototypes are defined in the header file
175     <b>pcre.h</b>, and on Unix systems the library itself is called <b>libpcre</b>.
176     It can normally be accessed by adding <b>-lpcre</b> to the command for linking
177     an application that uses PCRE. The header file defines the macros PCRE_MAJOR
178     and PCRE_MINOR to contain the major and minor release numbers for the library.
179 nigel 75 Applications can use these to include support for different releases of PCRE.
180 nigel 63 </P>
181     <P>
182 ph10 535 In a Windows environment, if you want to statically link an application program
183     against a non-dll <b>pcre.a</b> file, you must define PCRE_STATIC before
184     including <b>pcre.h</b> or <b>pcrecpp.h</b>, because otherwise the
185     <b>pcre_malloc()</b> and <b>pcre_free()</b> exported functions will be declared
186     <b>__declspec(dllimport)</b>, with unwanted results.
187     </P>
188     <P>
189 nigel 77 The functions <b>pcre_compile()</b>, <b>pcre_compile2()</b>, <b>pcre_study()</b>,
190     and <b>pcre_exec()</b> are used for compiling and matching regular expressions
191     in a Perl-compatible manner. A sample program that demonstrates the simplest
192 ph10 429 way of using them is provided in the file called <i>pcredemo.c</i> in the PCRE
193     source distribution. A listing of this program is given in the
194     <a href="pcredemo.html"><b>pcredemo</b></a>
195     documentation, and the
196 nigel 75 <a href="pcresample.html"><b>pcresample</b></a>
197 ph10 313 documentation describes how to compile and run it.
198 nigel 63 </P>
199     <P>
200 ph10 691 Just-in-time compiler support is an optional feature of PCRE that can be built
201     in appropriate hardware environments. It greatly speeds up the matching
202     performance of many patterns. Simple programs can easily request that it be
203     used if available, by setting an option that is ignored when it is not
204     relevant. More complicated programs might need to make use of the functions
205     <b>pcre_jit_stack_alloc()</b>, <b>pcre_jit_stack_free()</b>, and
206     <b>pcre_assign_jit_stack()</b> in order to control the JIT code's memory usage.
207     These functions are discussed in the
208     <a href="pcrejit.html"><b>pcrejit</b></a>
209     documentation.
210     </P>
211     <P>
212 nigel 77 A second matching function, <b>pcre_dfa_exec()</b>, which is not
213     Perl-compatible, is also provided. This uses a different algorithm for the
214 nigel 91 matching. The alternative algorithm finds all possible matches (at a given
215 ph10 453 point in the subject), and scans the subject just once (unless there are
216     lookbehind assertions). However, this algorithm does not return captured
217     substrings. A description of the two matching algorithms and their advantages
218     and disadvantages is given in the
219 nigel 77 <a href="pcrematching.html"><b>pcrematching</b></a>
220     documentation.
221     </P>
222     <P>
223 nigel 75 In addition to the main compiling and matching functions, there are convenience
224 nigel 77 functions for extracting captured substrings from a subject string that is
225     matched by <b>pcre_exec()</b>. They are:
226 nigel 63 <pre>
227     <b>pcre_copy_substring()</b>
228     <b>pcre_copy_named_substring()</b>
229     <b>pcre_get_substring()</b>
230     <b>pcre_get_named_substring()</b>
231     <b>pcre_get_substring_list()</b>
232 nigel 75 <b>pcre_get_stringnumber()</b>
233 nigel 91 <b>pcre_get_stringtable_entries()</b>
234 nigel 75 </pre>
235 nigel 63 <b>pcre_free_substring()</b> and <b>pcre_free_substring_list()</b> are also
236     provided, to free the memory used for extracted strings.
237     </P>
238     <P>
239 nigel 75 The function <b>pcre_maketables()</b> is used to build a set of character tables
240 nigel 77 in the current locale for passing to <b>pcre_compile()</b>, <b>pcre_exec()</b>,
241     or <b>pcre_dfa_exec()</b>. This is an optional facility that is provided for
242     specialist use. Most commonly, no special tables are passed, in which case
243     internal tables that are generated when PCRE is built are used.
244 nigel 63 </P>
245     <P>
246     The function <b>pcre_fullinfo()</b> is used to find out information about a
247 nigel 75 compiled pattern; <b>pcre_info()</b> is an obsolete version that returns only
248 nigel 63 some of the available information, but is retained for backwards compatibility.
249     The function <b>pcre_version()</b> returns a pointer to a string containing the
250     version of PCRE and its date of release.
251     </P>
252     <P>
253 nigel 77 The function <b>pcre_refcount()</b> maintains a reference count in a data block
254     containing a compiled pattern. This is provided for the benefit of
255     object-oriented applications.
256     </P>
257     <P>
258 nigel 63 The global variables <b>pcre_malloc</b> and <b>pcre_free</b> initially contain
259 nigel 75 the entry points of the standard <b>malloc()</b> and <b>free()</b> functions,
260 nigel 63 respectively. PCRE calls the memory management functions via these variables,
261     so a calling program can replace them if it wishes to intercept the calls. This
262     should be done before calling any PCRE functions.
263     </P>
264     <P>
265 nigel 73 The global variables <b>pcre_stack_malloc</b> and <b>pcre_stack_free</b> are also
266     indirections to memory management functions. These special functions are used
267     only when PCRE is compiled to use the heap for remembering data, instead of
268 nigel 91 recursive function calls, when running the <b>pcre_exec()</b> function. See the
269     <a href="pcrebuild.html"><b>pcrebuild</b></a>
270     documentation for details of how to do this. It is a non-standard way of
271     building PCRE, for use in environments that have limited stacks. Because of the
272     greater use of memory management, it runs more slowly. Separate functions are
273     provided so that special-purpose external code can be used for this case. When
274     used, these functions are always called in a stack-like manner (last obtained,
275     first freed), and always for memory blocks of the same size. There is a
276     discussion about PCRE's stack usage in the
277     <a href="pcrestack.html"><b>pcrestack</b></a>
278     documentation.
279 nigel 73 </P>
280     <P>
281 nigel 63 The global variable <b>pcre_callout</b> initially contains NULL. It can be set
282     by the caller to a "callout" function, which PCRE will then call at specified
283 nigel 75 points during a matching operation. Details are given in the
284     <a href="pcrecallout.html"><b>pcrecallout</b></a>
285 nigel 63 documentation.
286 ph10 227 <a name="newlines"></a></P>
287 ph10 691 <br><a name="SEC5" href="#TOC1">NEWLINES</a><br>
288 nigel 63 <P>
289 ph10 150 PCRE supports five different conventions for indicating line breaks in
290 nigel 93 strings: a single CR (carriage return) character, a single LF (linefeed)
291 ph10 150 character, the two-character sequence CRLF, any of the three preceding, or any
292     Unicode newline sequence. The Unicode newline sequences are the three just
293     mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
294     U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
295     (paragraph separator, U+2029).
296 nigel 93 </P>
297     <P>
298     Each of the first three conventions is used by at least one operating system as
299     its standard newline sequence. When PCRE is built, a default can be specified.
300     The default default is LF, which is the Unix standard. When PCRE is run, the
301     default can be overridden, either when a pattern is compiled, or when it is
302     matched.
303     </P>
304     <P>
305 ph10 227 At compile time, the newline convention can be specified by the <i>options</i>
306     argument of <b>pcre_compile()</b>, or it can be specified by special text at the
307     start of the pattern itself; this overrides any other settings. See the
308     <a href="pcrepattern.html"><b>pcrepattern</b></a>
309     page for details of the special character sequences.
310     </P>
311     <P>
312 nigel 91 In the PCRE documentation the word "newline" is used to mean "the character or
313 nigel 93 pair of characters that indicate a line break". The choice of newline
314     convention affects the handling of the dot, circumflex, and dollar
315     metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
316     recognized line ending sequence, the match position advancement for a
317 ph10 227 non-anchored pattern. There is more detail about this in the
318     <a href="#execoptions">section on <b>pcre_exec()</b> options</a>
319 ph10 231 below.
320 nigel 91 </P>
321 ph10 231 <P>
322     The choice of newline convention does not affect the interpretation of
323     the \n or \r escape sequences, nor does it affect what \R matches, which is
324     controlled in a similar way, but by separate options.
325     </P>
326 ph10 691 <br><a name="SEC6" href="#TOC1">MULTITHREADING</a><br>
327 nigel 91 <P>
328 nigel 63 The PCRE functions can be used in multi-threading applications, with the
329 nigel 73 proviso that the memory management functions pointed to by <b>pcre_malloc</b>,
330     <b>pcre_free</b>, <b>pcre_stack_malloc</b>, and <b>pcre_stack_free</b>, and the
331     callout function pointed to by <b>pcre_callout</b>, are shared by all threads.
332 nigel 63 </P>
333     <P>
334     The compiled form of a regular expression is not altered during matching, so
335     the same compiled pattern can safely be used by several threads at once.
336     </P>
337     <P>
338 ph10 691 If the just-in-time optimization feature is being used, it needs separate
339     memory stack areas for each thread. See the
340     <a href="pcrejit.html"><b>pcrejit</b></a>
341     documentation for more details.
342     </P>
343     <br><a name="SEC7" href="#TOC1">SAVING PRECOMPILED PATTERNS FOR LATER USE</a><br>
344     <P>
345 nigel 75 The compiled form of a regular expression can be saved and re-used at a later
346     time, possibly by a different program, and even on a host other than the one on
347     which it was compiled. Details are given in the
348     <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
349 ph10 155 documentation. However, compiling a regular expression with one version of PCRE
350     for use with a different version is not guaranteed to work and may cause
351     crashes.
352 nigel 75 </P>
353 ph10 691 <br><a name="SEC8" href="#TOC1">CHECKING BUILD-TIME OPTIONS</a><br>
354 nigel 75 <P>
355 nigel 63 <b>int pcre_config(int <i>what</i>, void *<i>where</i>);</b>
356     </P>
357     <P>
358     The function <b>pcre_config()</b> makes it possible for a PCRE client to
359     discover which optional features have been compiled into the PCRE library. The
360     <a href="pcrebuild.html"><b>pcrebuild</b></a>
361     documentation has more details about these optional features.
362     </P>
363     <P>
364     The first argument for <b>pcre_config()</b> is an integer, specifying which
365     information is required; the second argument is a pointer to a variable into
366     which the information is placed. The following information is available:
367     <pre>
368     PCRE_CONFIG_UTF8
369 nigel 75 </pre>
370 nigel 63 The output is an integer that is set to one if UTF-8 support is available;
371     otherwise it is set to zero.
372     <pre>
373 nigel 75 PCRE_CONFIG_UNICODE_PROPERTIES
374     </pre>
375     The output is an integer that is set to one if support for Unicode character
376     properties is available; otherwise it is set to zero.
377     <pre>
378 ph10 691 PCRE_CONFIG_JIT
379     </pre>
380     The output is an integer that is set to one if support for just-in-time
381     compiling is available; otherwise it is set to zero.
382     <pre>
383 nigel 63 PCRE_CONFIG_NEWLINE
384 nigel 75 </pre>
385 nigel 91 The output is an integer whose value specifies the default character sequence
386 nigel 93 that is recognized as meaning "newline". The four values that are supported
387 ph10 392 are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF, and -1 for ANY.
388     Though they are derived from ASCII, the same values are returned in EBCDIC
389     environments. The default should normally correspond to the standard sequence
390     for your operating system.
391 nigel 63 <pre>
392 ph10 231 PCRE_CONFIG_BSR
393     </pre>
394     The output is an integer whose value indicates what character sequences the \R
395     escape sequence matches by default. A value of 0 means that \R matches any
396     Unicode line ending sequence; a value of 1 means that \R matches only CR, LF,
397     or CRLF. The default can be overridden when a pattern is compiled or matched.
398     <pre>
399 nigel 63 PCRE_CONFIG_LINK_SIZE
400 nigel 75 </pre>
401 nigel 63 The output is an integer that contains the number of bytes used for internal
402     linkage in compiled regular expressions. The value is 2, 3, or 4. Larger values
403     allow larger regular expressions to be compiled, at the expense of slower
404     matching. The default value of 2 is sufficient for all but the most massive
405     patterns, since it allows the compiled pattern to be up to 64K in size.
406     <pre>
407     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
408 nigel 75 </pre>
409 nigel 63 The output is an integer that contains the threshold above which the POSIX
410     interface uses <b>malloc()</b> for output vectors. Further details are given in
411 nigel 75 the
412     <a href="pcreposix.html"><b>pcreposix</b></a>
413     documentation.
414 nigel 63 <pre>
415     PCRE_CONFIG_MATCH_LIMIT
416 nigel 75 </pre>
417 ph10 392 The output is a long integer that gives the default limit for the number of
418 nigel 63 internal matching function calls in a <b>pcre_exec()</b> execution. Further
419     details are given with <b>pcre_exec()</b> below.
420 nigel 73 <pre>
421 nigel 87 PCRE_CONFIG_MATCH_LIMIT_RECURSION
422     </pre>
423 ph10 392 The output is a long integer that gives the default limit for the depth of
424 nigel 87 recursion when calling the internal matching function in a <b>pcre_exec()</b>
425     execution. Further details are given with <b>pcre_exec()</b> below.
426     <pre>
427 nigel 73 PCRE_CONFIG_STACKRECURSE
428 nigel 75 </pre>
429 nigel 77 The output is an integer that is set to one if internal recursion when running
430     <b>pcre_exec()</b> is implemented by recursive function calls that use the stack
431     to remember their state. This is the usual way that PCRE is compiled. The
432     output is zero if PCRE was compiled to use blocks of data on the heap instead
433     of recursive function calls. In this case, <b>pcre_stack_malloc</b> and
434     <b>pcre_stack_free</b> are called to manage memory blocks on the heap, thus
435     avoiding the use of the stack.
436 nigel 73 </P>
437 ph10 691 <br><a name="SEC9" href="#TOC1">COMPILING A PATTERN</a><br>
438 nigel 63 <P>
439     <b>pcre *pcre_compile(const char *<i>pattern</i>, int <i>options</i>,</b>
440     <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
441     <b>const unsigned char *<i>tableptr</i>);</b>
442 nigel 77 <b>pcre *pcre_compile2(const char *<i>pattern</i>, int <i>options</i>,</b>
443     <b>int *<i>errorcodeptr</i>,</b>
444     <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
445     <b>const unsigned char *<i>tableptr</i>);</b>
446 nigel 63 </P>
447     <P>
448 nigel 77 Either of the functions <b>pcre_compile()</b> or <b>pcre_compile2()</b> can be
449     called to compile a pattern into an internal form. The only difference between
450     the two interfaces is that <b>pcre_compile2()</b> has an additional argument,
451 ph10 461 <i>errorcodeptr</i>, via which a numerical error code can be returned. To avoid
452     too much repetition, we refer just to <b>pcre_compile()</b> below, but the
453     information applies equally to <b>pcre_compile2()</b>.
454 nigel 63 </P>
455     <P>
456 nigel 77 The pattern is a C string terminated by a binary zero, and is passed in the
457     <i>pattern</i> argument. A pointer to a single block of memory that is obtained
458     via <b>pcre_malloc</b> is returned. This contains the compiled code and related
459     data. The <b>pcre</b> type is defined for the returned block; this is a typedef
460     for a structure whose contents are not externally defined. It is up to the
461 nigel 91 caller to free the memory (via <b>pcre_free</b>) when it is no longer required.
462 nigel 77 </P>
463     <P>
464 nigel 63 Although the compiled code of a PCRE regex is relocatable, that is, it does not
465     depend on memory location, the complete <b>pcre</b> data block is not
466 nigel 75 fully relocatable, because it may contain a copy of the <i>tableptr</i>
467     argument, which is an address (see below).
468 nigel 63 </P>
469     <P>
470 nigel 93 The <i>options</i> argument contains various bit settings that affect the
471 nigel 75 compilation. It should be zero if no options are required. The available
472 ph10 416 options are described below. Some of them (in particular, those that are
473 ph10 461 compatible with Perl, but some others as well) can also be set and unset from
474 ph10 416 within the pattern (see the detailed description in the
475 nigel 75 <a href="pcrepattern.html"><b>pcrepattern</b></a>
476 ph10 416 documentation). For those options that can be different in different parts of
477 ph10 461 the pattern, the contents of the <i>options</i> argument specifies their
478     settings at the start of compilation and execution. The PCRE_ANCHORED,
479 ph10 579 PCRE_BSR_<i>xxx</i>, PCRE_NEWLINE_<i>xxx</i>, PCRE_NO_UTF8_CHECK, and
480     PCRE_NO_START_OPT options can be set at the time of matching as well as at
481     compile time.
482 nigel 63 </P>
483     <P>
484     If <i>errptr</i> is NULL, <b>pcre_compile()</b> returns NULL immediately.
485     Otherwise, if compilation of a pattern fails, <b>pcre_compile()</b> returns
486     NULL, and sets the variable pointed to by <i>errptr</i> to point to a textual
487 nigel 87 error message. This is a static string that is part of the library. You must
488 ph10 654 not try to free it. Normally, the offset from the start of the pattern to the
489     byte that was being processed when the error was discovered is placed in the
490     variable pointed to by <i>erroffset</i>, which must not be NULL (if it is, an
491     immediate error is given). However, for an invalid UTF-8 string, the offset is
492     that of the first byte of the failing character. Also, some errors are not
493     detected until checks are carried out when the whole pattern has been scanned;
494     in these cases the offset passed back is the length of the pattern.
495 nigel 63 </P>
496     <P>
497 ph10 572 Note that the offset is in bytes, not characters, even in UTF-8 mode. It may
498 ph10 654 sometimes point into the middle of a UTF-8 character.
499 ph10 572 </P>
500     <P>
501 nigel 77 If <b>pcre_compile2()</b> is used instead of <b>pcre_compile()</b>, and the
502     <i>errorcodeptr</i> argument is not NULL, a non-zero error code number is
503     returned via this argument in the event of an error. This is in addition to the
504     textual error message. Error codes and messages are listed below.
505     </P>
506     <P>
507 nigel 63 If the final argument, <i>tableptr</i>, is NULL, PCRE uses a default set of
508 nigel 75 character tables that are built when PCRE is compiled, using the default C
509     locale. Otherwise, <i>tableptr</i> must be an address that is the result of a
510     call to <b>pcre_maketables()</b>. This value is stored with the compiled
511     pattern, and used again by <b>pcre_exec()</b>, unless another table pointer is
512     passed to it. For more discussion, see the section on locale support below.
513 nigel 63 </P>
514     <P>
515     This code fragment shows a typical straightforward call to <b>pcre_compile()</b>:
516     <pre>
517     pcre *re;
518     const char *error;
519     int erroffset;
520     re = pcre_compile(
521     "^A.*Z", /* the pattern */
522     0, /* default options */
523     &error, /* for error message */
524     &erroffset, /* for error offset */
525     NULL); /* use default character tables */
526 nigel 75 </pre>
527     The following names for option bits are defined in the <b>pcre.h</b> header
528     file:
529 nigel 63 <pre>
530     PCRE_ANCHORED
531 nigel 75 </pre>
532 nigel 63 If this bit is set, the pattern is forced to be "anchored", that is, it is
533 nigel 75 constrained to match only at the first matching point in the string that is
534 nigel 63 being searched (the "subject string"). This effect can also be achieved by
535     appropriate constructs in the pattern itself, which is the only way to do it in
536     Perl.
537     <pre>
538 nigel 75 PCRE_AUTO_CALLOUT
539     </pre>
540     If this bit is set, <b>pcre_compile()</b> automatically inserts callout items,
541     all with number 255, before each pattern item. For discussion of the callout
542     facility, see the
543     <a href="pcrecallout.html"><b>pcrecallout</b></a>
544     documentation.
545     <pre>
546 ph10 231 PCRE_BSR_ANYCRLF
547     PCRE_BSR_UNICODE
548     </pre>
549     These options (which are mutually exclusive) control what the \R escape
550     sequence matches. The choice is either to match only CR, LF, or CRLF, or to
551     match any Unicode newline sequence. The default is specified when PCRE is
552     built. It can be overridden from within the pattern, or by setting an option
553     when a compiled pattern is matched.
554     <pre>
555 nigel 63 PCRE_CASELESS
556 nigel 75 </pre>
557 nigel 63 If this bit is set, letters in the pattern match both upper and lower case
558     letters. It is equivalent to Perl's /i option, and it can be changed within a
559 nigel 77 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
560     concept of case for characters whose values are less than 128, so caseless
561     matching is always possible. For characters with higher values, the concept of
562     case is supported if PCRE is compiled with Unicode property support, but not
563     otherwise. If you want to use caseless matching for characters 128 and above,
564     you must ensure that PCRE is compiled with Unicode property support as well as
565     with UTF-8 support.
566 nigel 63 <pre>
567     PCRE_DOLLAR_ENDONLY
568 nigel 75 </pre>
569 nigel 63 If this bit is set, a dollar metacharacter in the pattern matches only at the
570     end of the subject string. Without this option, a dollar also matches
571 nigel 91 immediately before a newline at the end of the string (but not before any other
572     newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
573     There is no equivalent to this option in Perl, and no way to set it within a
574     pattern.
575 nigel 63 <pre>
576     PCRE_DOTALL
577 nigel 75 </pre>
578 ph10 572 If this bit is set, a dot metacharacter in the pattern matches a character of
579     any value, including one that indicates a newline. However, it only ever
580     matches one character, even if newlines are coded as CRLF. Without this option,
581     a dot does not match when the current position is at a newline. This option is
582     equivalent to Perl's /s option, and it can be changed within a pattern by a
583     (?s) option setting. A negative class such as [^a] always matches newline
584     characters, independent of the setting of this option.
585 nigel 63 <pre>
586 nigel 91 PCRE_DUPNAMES
587     </pre>
588     If this bit is set, names used to identify capturing subpatterns need not be
589     unique. This can be helpful for certain types of pattern when it is known that
590     only one instance of the named subpattern can ever be matched. There are more
591     details of named subpatterns below; see also the
592     <a href="pcrepattern.html"><b>pcrepattern</b></a>
593     documentation.
594     <pre>
595 nigel 63 PCRE_EXTENDED
596 nigel 75 </pre>
597 nigel 63 If this bit is set, whitespace data characters in the pattern are totally
598     ignored except when escaped or inside a character class. Whitespace does not
599     include the VT character (code 11). In addition, characters between an
600 nigel 91 unescaped # outside a character class and the next newline, inclusive, are also
601     ignored. This is equivalent to Perl's /x option, and it can be changed within a
602     pattern by a (?x) option setting.
603 nigel 63 </P>
604     <P>
605 ph10 654 Which characters are interpreted as newlines is controlled by the options
606     passed to <b>pcre_compile()</b> or by a special sequence at the start of the
607     pattern, as described in the section entitled
608 ph10 572 <a href="pcrepattern.html#newlines">"Newline conventions"</a>
609     in the <b>pcrepattern</b> documentation. Note that the end of this type of
610     comment is a literal newline sequence in the pattern; escape sequences that
611     happen to represent a newline do not count.
612     </P>
613     <P>
614 nigel 63 This option makes it possible to include comments inside complicated patterns.
615     Note, however, that this applies only to data characters. Whitespace characters
616     may never appear within special character sequences in a pattern, for example
617 ph10 572 within the sequence (?( that introduces a conditional subpattern.
618 nigel 63 <pre>
619     PCRE_EXTRA
620 nigel 75 </pre>
621 nigel 63 This option was invented in order to turn on additional functionality of PCRE
622     that is incompatible with Perl, but it is currently of very little use. When
623     set, any backslash in a pattern that is followed by a letter that has no
624     special meaning causes an error, thus reserving these combinations for future
625     expansion. By default, as in Perl, a backslash followed by a letter with no
626 nigel 91 special meaning is treated as a literal. (Perl can, however, be persuaded to
627 ph10 518 give an error for this, by running it with the -w option.) There are at present
628     no other features controlled by this option. It can also be set by a (?X)
629     option setting within a pattern.
630 nigel 63 <pre>
631 nigel 77 PCRE_FIRSTLINE
632     </pre>
633     If this option is set, an unanchored pattern is required to match before or at
634 nigel 91 the first newline in the subject string, though the matched text may continue
635     over the newline.
636 nigel 77 <pre>
637 ph10 345 PCRE_JAVASCRIPT_COMPAT
638     </pre>
639     If this option is set, PCRE's behaviour is changed in some ways so that it is
640     compatible with JavaScript rather than Perl. The changes are as follows:
641     </P>
642     <P>
643     (1) A lone closing square bracket in a pattern causes a compile-time error,
644     because this is illegal in JavaScript (by default it is treated as a data
645     character). Thus, the pattern AB]CD becomes illegal when this option is set.
646     </P>
647     <P>
648     (2) At run time, a back reference to an unset subpattern group matches an empty
649     string (by default this causes the current matching alternative to fail). A
650     pattern such as (\1)(a) succeeds when this option is set (assuming it can find
651     an "a" in the subject), whereas it fails by default, for Perl compatibility.
652     <pre>
653 nigel 63 PCRE_MULTILINE
654 nigel 75 </pre>
655     By default, PCRE treats the subject string as consisting of a single line of
656     characters (even if it actually contains newlines). The "start of line"
657 nigel 63 metacharacter (^) matches only at the start of the string, while the "end of
658     line" metacharacter ($) matches only at the end of the string, or before a
659     terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
660     Perl.
661     </P>
662     <P>
663     When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
664 nigel 91 match immediately following or immediately before internal newlines in the
665     subject string, respectively, as well as at the very start and end. This is
666     equivalent to Perl's /m option, and it can be changed within a pattern by a
667     (?m) option setting. If there are no newlines in a subject string, or no
668 nigel 63 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
669     <pre>
670 nigel 91 PCRE_NEWLINE_CR
671     PCRE_NEWLINE_LF
672     PCRE_NEWLINE_CRLF
673 ph10 150 PCRE_NEWLINE_ANYCRLF
674 nigel 93 PCRE_NEWLINE_ANY
675 nigel 91 </pre>
676     These options override the default newline definition that was chosen when PCRE
677     was built. Setting the first or the second specifies that a newline is
678 nigel 93 indicated by a single character (CR or LF, respectively). Setting
679     PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
680 ph10 150 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
681     preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
682     that any Unicode newline sequence should be recognized. The Unicode newline
683     sequences are the three just mentioned, plus the single characters VT (vertical
684     tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
685     separator, U+2028), and PS (paragraph separator, U+2029). The last two are
686     recognized only in UTF-8 mode.
687 nigel 91 </P>
688     <P>
689 nigel 93 The newline setting in the options word uses three bits that are treated
690 ph10 150 as a number, giving eight possibilities. Currently only six are used (default
691     plus the five values above). This means that if you set more than one newline
692 nigel 93 option, the combination may or may not be sensible. For example,
693     PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
694 ph10 150 other combinations may yield unused numbers and cause an error.
695 nigel 93 </P>
696     <P>
697 ph10 572 The only time that a line break in a pattern is specially recognized when
698     compiling is when PCRE_EXTENDED is set. CR and LF are whitespace characters,
699     and so are ignored in this mode. Also, an unescaped # outside a character class
700     indicates a comment that lasts until after the next line break sequence. In
701     other circumstances, line break sequences in patterns are treated as literal
702     data.
703 nigel 93 </P>
704     <P>
705     The newline option that is set at compile time becomes the default that is used
706     for <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, but it can be overridden.
707 nigel 91 <pre>
708 nigel 63 PCRE_NO_AUTO_CAPTURE
709 nigel 75 </pre>
710 nigel 63 If this option is set, it disables the use of numbered capturing parentheses in
711     the pattern. Any opening parenthesis that is not followed by ? behaves as if it
712     were followed by ?: but named parentheses can still be used for capturing (and
713     they acquire numbers in the usual way). There is no equivalent of this option
714     in Perl.
715     <pre>
716 ph10 579 NO_START_OPTIMIZE
717     </pre>
718     This is an option that acts at matching time; that is, it is really an option
719     for <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b>. If it is set at compile time,
720     it is remembered with the compiled pattern and assumed at matching time. For
721     details see the discussion of PCRE_NO_START_OPTIMIZE
722     <a href="#execoptions">below.</a>
723     <pre>
724 ph10 535 PCRE_UCP
725     </pre>
726 ph10 572 This option changes the way PCRE processes \B, \b, \D, \d, \S, \s, \W,
727     \w, and some of the POSIX character classes. By default, only ASCII characters
728     are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
729     classify characters. More details are given in the section on
730 ph10 535 <a href="pcre.html#genericchartypes">generic character types</a>
731     in the
732     <a href="pcrepattern.html"><b>pcrepattern</b></a>
733     page. If you set PCRE_UCP, matching one of the items it affects takes much
734     longer. The option is available only if PCRE has been compiled with Unicode
735     property support.
736     <pre>
737 nigel 63 PCRE_UNGREEDY
738 nigel 75 </pre>
739 nigel 63 This option inverts the "greediness" of the quantifiers so that they are not
740     greedy by default, but become greedy if followed by "?". It is not compatible
741     with Perl. It can also be set by a (?U) option setting within the pattern.
742     <pre>
743     PCRE_UTF8
744 nigel 75 </pre>
745 nigel 63 This option causes PCRE to regard both the pattern and the subject as strings
746     of UTF-8 characters instead of single-byte character strings. However, it is
747 nigel 75 available only when PCRE is built to include UTF-8 support. If not, the use
748 nigel 63 of this option provokes an error. Details of how this option changes the
749     behaviour of PCRE are given in the
750 ph10 678 <a href="pcreunicode.html"><b>pcreunicode</b></a>
751 nigel 63 page.
752 nigel 71 <pre>
753     PCRE_NO_UTF8_CHECK
754 nigel 75 </pre>
755 nigel 71 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
756 ph10 211 automatically checked. There is a discussion about the
757     <a href="pcre.html#utf8strings">validity of UTF-8 strings</a>
758     in the main
759     <a href="pcre.html"><b>pcre</b></a>
760     page. If an invalid UTF-8 sequence of bytes is found, <b>pcre_compile()</b>
761     returns an error. If you already know that your pattern is valid, and you want
762     to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
763     option. When it is set, the effect of passing an invalid UTF-8 string as a
764     pattern is undefined. It may cause your program to crash. Note that this option
765     can also be passed to <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, to suppress
766     the UTF-8 validity checking of subject strings.
767 nigel 71 </P>
768 ph10 691 <br><a name="SEC10" href="#TOC1">COMPILATION ERROR CODES</a><br>
769 nigel 63 <P>
770 nigel 77 The following table lists the error codes than may be returned by
771     <b>pcre_compile2()</b>, along with the error messages that may be returned by
772 nigel 93 both compiling functions. As PCRE has developed, some error codes have fallen
773     out of use. To avoid confusion, they have not been re-used.
774 nigel 77 <pre>
775     0 no error
776     1 \ at end of pattern
777     2 \c at end of pattern
778     3 unrecognized character follows \
779     4 numbers out of order in {} quantifier
780     5 number too big in {} quantifier
781     6 missing terminating ] for character class
782     7 invalid escape sequence in character class
783     8 range out of order in character class
784     9 nothing to repeat
785 nigel 93 10 [this code is not in use]
786 nigel 77 11 internal error: unexpected repeat
787 ph10 292 12 unrecognized character after (? or (?-
788 nigel 77 13 POSIX named classes are supported only within a class
789     14 missing )
790     15 reference to non-existent subpattern
791     16 erroffset passed as NULL
792     17 unknown option bit(s) set
793     18 missing ) after comment
794 nigel 93 19 [this code is not in use]
795 ph10 292 20 regular expression is too large
796 nigel 77 21 failed to get memory
797     22 unmatched parentheses
798     23 internal error: code overflow
799     24 unrecognized character after (?&#60;
800     25 lookbehind assertion is not fixed length
801 nigel 91 26 malformed number or name after (?(
802 nigel 77 27 conditional group contains more than two branches
803     28 assertion expected after (?(
804 ph10 182 29 (?R or (?[+-]digits must be followed by )
805 nigel 77 30 unknown POSIX class name
806     31 POSIX collating elements are not supported
807     32 this version of PCRE is not compiled with PCRE_UTF8 support
808 nigel 93 33 [this code is not in use]
809 nigel 77 34 character value in \x{...} sequence is too large
810     35 invalid condition (?(0)
811     36 \C not allowed in lookbehind assertion
812 ph10 656 37 PCRE does not support \L, \l, \N{name}, \U, or \u
813 nigel 77 38 number after (?C is &#62; 255
814     39 closing ) for (?C expected
815     40 recursive call could loop indefinitely
816     41 unrecognized character after (?P
817 nigel 93 42 syntax error in subpattern name (missing terminator)
818 nigel 91 43 two named subpatterns have the same name
819 nigel 77 44 invalid UTF-8 string
820     45 support for \P, \p, and \X has not been compiled
821     46 malformed \P or \p sequence
822     47 unknown property name after \P or \p
823 nigel 91 48 subpattern name is too long (maximum 32 characters)
824 ph10 292 49 too many named subpatterns (maximum 10000)
825 ph10 202 50 [this code is not in use]
826 nigel 91 51 octal value is greater than \377 (not in UTF-8 mode)
827 nigel 93 52 internal error: overran compiling workspace
828 ph10 548 53 internal error: previously-checked referenced subpattern
829     not found
830 nigel 93 54 DEFINE group contains more than one branch
831     55 repeating a DEFINE group is not allowed
832 ph10 231 56 inconsistent NEWLINE options
833 ph10 345 57 \g is not followed by a braced, angle-bracketed, or quoted
834     name/number or by a plain number
835     58 a numbered reference must not be zero
836 ph10 512 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
837 ph10 292 60 (*VERB) not recognized
838     61 number is too big
839     62 subpattern name expected
840     63 digit expected after (?+
841 ph10 345 64 ] is an invalid data character in JavaScript compatibility mode
842 ph10 548 65 different names for subpatterns of the same number are
843     not allowed
844 ph10 512 66 (*MARK) must have an argument
845 ph10 535 67 this version of PCRE is not compiled with PCRE_UCP support
846 ph10 656 68 \c must be followed by an ASCII character
847 ph10 659 69 \k is not followed by a braced, angle-bracketed, or quoted name
848 ph10 292 </pre>
849     The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
850     be used if the limits were changed when PCRE was built.
851 ph10 656 <a name="studyingapattern"></a></P>
852 ph10 691 <br><a name="SEC11" href="#TOC1">STUDYING A PATTERN</a><br>
853 nigel 77 <P>
854     <b>pcre_extra *pcre_study(const pcre *<i>code</i>, int <i>options</i></b>
855 nigel 63 <b>const char **<i>errptr</i>);</b>
856     </P>
857     <P>
858 nigel 75 If a compiled pattern is going to be used several times, it is worth spending
859     more time analyzing it in order to speed up the time taken for matching. The
860 nigel 63 function <b>pcre_study()</b> takes a pointer to a compiled pattern as its first
861 nigel 75 argument. If studying the pattern produces additional information that will
862     help speed up matching, <b>pcre_study()</b> returns a pointer to a
863     <b>pcre_extra</b> block, in which the <i>study_data</i> field points to the
864     results of the study.
865 nigel 63 </P>
866     <P>
867 nigel 75 The returned value from <b>pcre_study()</b> can be passed directly to
868 ph10 461 <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b>. However, a <b>pcre_extra</b> block
869     also contains other fields that can be set by the caller before the block is
870     passed; these are described
871 nigel 75 <a href="#extradata">below</a>
872     in the section on matching a pattern.
873 nigel 63 </P>
874     <P>
875 ph10 461 If studying the pattern does not produce any useful information,
876 nigel 75 <b>pcre_study()</b> returns NULL. In that circumstance, if the calling program
877 ph10 461 wants to pass any of the other fields to <b>pcre_exec()</b> or
878     <b>pcre_dfa_exec()</b>, it must set up its own <b>pcre_extra</b> block.
879 nigel 63 </P>
880     <P>
881 ph10 691 The second argument of <b>pcre_study()</b> contains option bits. There is only
882     one option: PCRE_STUDY_JIT_COMPILE. If this is set, and the just-in-time
883     compiler is available, the pattern is further compiled into machine code that
884     executes much faster than the <b>pcre_exec()</b> matching function. If
885     the just-in-time compiler is not available, this option is ignored. All other
886     bits in the <i>options</i> argument must be zero.
887 nigel 75 </P>
888     <P>
889 ph10 691 JIT compilation is a heavyweight optimization. It can take some time for
890     patterns to be analyzed, and for one-off matches and simple patterns the
891     benefit of faster execution might be offset by a much slower study time.
892     Not all patterns can be optimized by the JIT compiler. For those that cannot be
893     handled, matching automatically falls back to the <b>pcre_exec()</b>
894     interpreter. For more details, see the
895     <a href="pcrejit.html"><b>pcrejit</b></a>
896     documentation.
897     </P>
898     <P>
899 nigel 63 The third argument for <b>pcre_study()</b> is a pointer for an error message. If
900     studying succeeds (even if no data is returned), the variable it points to is
901 nigel 87 set to NULL. Otherwise it is set to point to a textual error message. This is a
902     static string that is part of the library. You must not try to free it. You
903     should test the error pointer for NULL after calling <b>pcre_study()</b>, to be
904     sure that it has run successfully.
905 nigel 63 </P>
906     <P>
907 ph10 691 When you are finished with a pattern, you can free the memory used for the
908     study data by calling <b>pcre_free_study()</b>. This function was added to the
909     API for release 8.20. For earlier versions, the memory could be freed with
910     <b>pcre_free()</b>, just like the pattern itself. This will still work in cases
911     where PCRE_STUDY_JIT_COMPILE is not used, but it is advisable to change to the
912     new function when convenient.
913     </P>
914     <P>
915     This is a typical way in which <b>pcre_study</b>() is used (except that in a
916     real application there should be tests for errors):
917 nigel 63 <pre>
918 ph10 691 int rc;
919     pcre *re;
920     pcre_extra *sd;
921     re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
922     sd = pcre_study(
923 nigel 63 re, /* result of pcre_compile() */
924 ph10 691 0, /* no options */
925 nigel 63 &error); /* set to NULL or points to a message */
926 ph10 691 rc = pcre_exec( /* see below for details of pcre_exec() options */
927     re, sd, "subject", 7, 0, 0, ovector, 30);
928     ...
929     pcre_free_study(sd);
930     pcre_free(re);
931 nigel 75 </pre>
932 ph10 461 Studying a pattern does two things: first, a lower bound for the length of
933     subject string that is needed to match the pattern is computed. This does not
934     mean that there are any strings of that length that match, but it does
935     guarantee that no shorter strings match. The value is used by
936     <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b> to avoid wasting time by trying to
937     match strings that are shorter than the lower bound. You can find out the value
938     in a calling program via the <b>pcre_fullinfo()</b> function.
939     </P>
940     <P>
941     Studying a pattern is also useful for non-anchored patterns that do not have a
942     single fixed starting character. A bitmap of possible starting bytes is
943     created. This speeds up finding a position in the subject at which to start
944     matching.
945 ph10 548 </P>
946     <P>
947 ph10 691 These two optimizations apply to both <b>pcre_exec()</b> and
948     <b>pcre_dfa_exec()</b>. However, they are not used by <b>pcre_exec()</b> if
949     <b>pcre_study()</b> is called with the PCRE_STUDY_JIT_COMPILE option, and
950     just-in-time compiling is successful. The optimizations can be disabled by
951     setting the PCRE_NO_START_OPTIMIZE option when calling <b>pcre_exec()</b> or
952 ph10 548 <b>pcre_dfa_exec()</b>. You might want to do this if your pattern contains
953 ph10 691 callouts or (*MARK) (which cannot be handled by the JIT compiler), and you want
954     to make use of these facilities in cases where matching fails. See the
955     discussion of PCRE_NO_START_OPTIMIZE
956 ph10 548 <a href="#execoptions">below.</a>
957 nigel 75 <a name="localesupport"></a></P>
958 ph10 691 <br><a name="SEC12" href="#TOC1">LOCALE SUPPORT</a><br>
959 nigel 63 <P>
960 ph10 142 PCRE handles caseless matching, and determines whether characters are letters,
961 nigel 75 digits, or whatever, by reference to a set of tables, indexed by character
962 nigel 77 value. When running in UTF-8 mode, this applies only to characters with codes
963 ph10 535 less than 128. By default, higher-valued codes never match escapes such as \w
964     or \d, but they can be tested with \p if PCRE is built with Unicode character
965     property support. Alternatively, the PCRE_UCP option can be set at compile
966     time; this causes \w and friends to use Unicode property support instead of
967     built-in tables. The use of locales with Unicode is discouraged. If you are
968     handling characters with codes greater than 128, you should either use UTF-8
969     and Unicode, or use locales, but not try to mix the two.
970 nigel 63 </P>
971     <P>
972 ph10 142 PCRE contains an internal set of tables that are used when the final argument
973     of <b>pcre_compile()</b> is NULL. These are sufficient for many applications.
974     Normally, the internal tables recognize only ASCII characters. However, when
975     PCRE is built, it is possible to cause the internal tables to be rebuilt in the
976     default "C" locale of the local system, which may cause them to be different.
977 nigel 63 </P>
978     <P>
979 ph10 142 The internal tables can always be overridden by tables supplied by the
980     application that calls PCRE. These may be created in a different locale from
981     the default. As more and more applications change to using Unicode, the need
982     for this locale support is expected to die away.
983     </P>
984     <P>
985 nigel 75 External tables are built by calling the <b>pcre_maketables()</b> function,
986     which has no arguments, in the relevant locale. The result can then be passed
987     to <b>pcre_compile()</b> or <b>pcre_exec()</b> as often as necessary. For
988     example, to build and use tables that are appropriate for the French locale
989     (where accented characters with values greater than 128 are treated as letters),
990     the following code could be used:
991 nigel 63 <pre>
992 nigel 75 setlocale(LC_CTYPE, "fr_FR");
993 nigel 63 tables = pcre_maketables();
994     re = pcre_compile(..., tables);
995 nigel 75 </pre>
996 ph10 142 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
997     are using Windows, the name for the French locale is "french".
998     </P>
999     <P>
1000 nigel 75 When <b>pcre_maketables()</b> runs, the tables are built in memory that is
1001     obtained via <b>pcre_malloc</b>. It is the caller's responsibility to ensure
1002     that the memory containing the tables remains available for as long as it is
1003     needed.
1004 nigel 63 </P>
1005     <P>
1006 nigel 75 The pointer that is passed to <b>pcre_compile()</b> is saved with the compiled
1007 nigel 63 pattern, and the same tables are used via this pointer by <b>pcre_study()</b>
1008 nigel 75 and normally also by <b>pcre_exec()</b>. Thus, by default, for any single
1009     pattern, compilation, studying and matching all happen in the same locale, but
1010     different patterns can be compiled in different locales.
1011 nigel 63 </P>
1012     <P>
1013 nigel 75 It is possible to pass a table pointer or NULL (indicating the use of the
1014     internal tables) to <b>pcre_exec()</b>. Although not intended for this purpose,
1015     this facility could be used to match a pattern in a different locale from the
1016     one in which it was compiled. Passing table pointers at run time is discussed
1017     below in the section on matching a pattern.
1018 ph10 654 <a name="infoaboutpattern"></a></P>
1019 ph10 691 <br><a name="SEC13" href="#TOC1">INFORMATION ABOUT A PATTERN</a><br>
1020 nigel 75 <P>
1021 nigel 63 <b>int pcre_fullinfo(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
1022     <b>int <i>what</i>, void *<i>where</i>);</b>
1023     </P>
1024     <P>
1025     The <b>pcre_fullinfo()</b> function returns information about a compiled
1026     pattern. It replaces the obsolete <b>pcre_info()</b> function, which is
1027     nevertheless retained for backwards compability (and is documented below).
1028     </P>
1029     <P>
1030     The first argument for <b>pcre_fullinfo()</b> is a pointer to the compiled
1031     pattern. The second argument is the result of <b>pcre_study()</b>, or NULL if
1032     the pattern was not studied. The third argument specifies which piece of
1033     information is required, and the fourth argument is a pointer to a variable
1034     to receive the data. The yield of the function is zero for success, or one of
1035     the following negative numbers:
1036     <pre>
1037     PCRE_ERROR_NULL the argument <i>code</i> was NULL
1038     the argument <i>where</i> was NULL
1039     PCRE_ERROR_BADMAGIC the "magic number" was not found
1040     PCRE_ERROR_BADOPTION the value of <i>what</i> was invalid
1041 nigel 75 </pre>
1042     The "magic number" is placed at the start of each compiled pattern as an simple
1043     check against passing an arbitrary memory pointer. Here is a typical call of
1044     <b>pcre_fullinfo()</b>, to obtain the length of the compiled pattern:
1045 nigel 63 <pre>
1046     int rc;
1047 nigel 91 size_t length;
1048 nigel 63 rc = pcre_fullinfo(
1049     re, /* result of pcre_compile() */
1050 ph10 691 sd, /* result of pcre_study(), or NULL */
1051 nigel 63 PCRE_INFO_SIZE, /* what is required */
1052     &length); /* where to put the data */
1053 nigel 75 </pre>
1054 nigel 63 The possible values for the third argument are defined in <b>pcre.h</b>, and are
1055     as follows:
1056     <pre>
1057     PCRE_INFO_BACKREFMAX
1058 nigel 75 </pre>
1059 nigel 63 Return the number of the highest back reference in the pattern. The fourth
1060     argument should point to an <b>int</b> variable. Zero is returned if there are
1061     no back references.
1062     <pre>
1063     PCRE_INFO_CAPTURECOUNT
1064 nigel 75 </pre>
1065 nigel 63 Return the number of capturing subpatterns in the pattern. The fourth argument
1066 nigel 75 should point to an <b>int</b> variable.
1067 nigel 63 <pre>
1068 nigel 77 PCRE_INFO_DEFAULT_TABLES
1069 nigel 75 </pre>
1070     Return a pointer to the internal default character tables within PCRE. The
1071     fourth argument should point to an <b>unsigned char *</b> variable. This
1072     information call is provided for internal use by the <b>pcre_study()</b>
1073     function. External callers can cause PCRE to use its internal tables by passing
1074     a NULL table pointer.
1075     <pre>
1076 nigel 63 PCRE_INFO_FIRSTBYTE
1077 nigel 75 </pre>
1078 nigel 63 Return information about the first byte of any matched string, for a
1079 nigel 91 non-anchored pattern. The fourth argument should point to an <b>int</b>
1080     variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
1081     still recognized for backwards compatibility.)
1082 nigel 63 </P>
1083     <P>
1084 nigel 75 If there is a fixed first byte, for example, from a pattern such as
1085 nigel 93 (cat|cow|coyote), its value is returned. Otherwise, if either
1086 nigel 75 <br>
1087     <br>
1088 nigel 63 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1089     starts with "^", or
1090 nigel 75 <br>
1091     <br>
1092 nigel 63 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1093     (if it were set, the pattern would be anchored),
1094 nigel 75 <br>
1095     <br>
1096 nigel 63 -1 is returned, indicating that the pattern matches only at the start of a
1097     subject string or after any newline within the string. Otherwise -2 is
1098     returned. For anchored patterns, -2 is returned.
1099     <pre>
1100     PCRE_INFO_FIRSTTABLE
1101 nigel 75 </pre>
1102 nigel 63 If the pattern was studied, and this resulted in the construction of a 256-bit
1103     table indicating a fixed set of bytes for the first byte in any matching
1104     string, a pointer to the table is returned. Otherwise NULL is returned. The
1105     fourth argument should point to an <b>unsigned char *</b> variable.
1106     <pre>
1107 ph10 227 PCRE_INFO_HASCRORLF
1108     </pre>
1109     Return 1 if the pattern contains any explicit matches for CR or LF characters,
1110 ph10 243 otherwise 0. The fourth argument should point to an <b>int</b> variable. An
1111     explicit match is either a literal CR or LF character, or \r or \n.
1112 ph10 227 <pre>
1113 ph10 172 PCRE_INFO_JCHANGED
1114     </pre>
1115 ph10 286 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1116     0. The fourth argument should point to an <b>int</b> variable. (?J) and
1117     (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1118 ph10 172 <pre>
1119 ph10 691 PCRE_INFO_JIT
1120     </pre>
1121     Return 1 if the pattern was studied with the PCRE_STUDY_JIT_COMPILE option, and
1122     just-in-time compiling was successful. The fourth argument should point to an
1123     <b>int</b> variable. A return value of 0 means that JIT support is not available
1124     in this version of PCRE, or that the pattern was not studied with the
1125     PCRE_STUDY_JIT_COMPILE option, or that the JIT compiler could not handle this
1126     particular pattern. See the
1127     <a href="pcrejit.html"><b>pcrejit</b></a>
1128     documentation for details of what can and cannot be handled.
1129     <pre>
1130 nigel 63 PCRE_INFO_LASTLITERAL
1131 nigel 75 </pre>
1132 nigel 65 Return the value of the rightmost literal byte that must exist in any matched
1133     string, other than at its start, if such a byte has been recorded. The fourth
1134     argument should point to an <b>int</b> variable. If there is no such byte, -1 is
1135     returned. For anchored patterns, a last literal byte is recorded only if it
1136     follows something of variable length. For example, for the pattern
1137     /^a\d+z\d+/ the returned value is "z", but for /^a\dz\d/ the returned value
1138     is -1.
1139 nigel 63 <pre>
1140 ph10 461 PCRE_INFO_MINLENGTH
1141     </pre>
1142     If the pattern was studied and a minimum length for matching subject strings
1143     was computed, its value is returned. Otherwise the returned value is -1. The
1144     value is a number of characters, not bytes (this may be relevant in UTF-8
1145     mode). The fourth argument should point to an <b>int</b> variable. A
1146     non-negative value is a lower bound to the length of any matching string. There
1147     may not be any strings of that length that do actually match, but every string
1148     that does match is at least that long.
1149     <pre>
1150 nigel 63 PCRE_INFO_NAMECOUNT
1151     PCRE_INFO_NAMEENTRYSIZE
1152     PCRE_INFO_NAMETABLE
1153 nigel 75 </pre>
1154 nigel 63 PCRE supports the use of named as well as numbered capturing parentheses. The
1155     names are just an additional way of identifying the parentheses, which still
1156 nigel 91 acquire numbers. Several convenience functions such as
1157     <b>pcre_get_named_substring()</b> are provided for extracting captured
1158     substrings by name. It is also possible to extract the data directly, by first
1159     converting the name to a number in order to access the correct pointers in the
1160     output vector (described with <b>pcre_exec()</b> below). To do the conversion,
1161     you need to use the name-to-number map, which is described by these three
1162     values.
1163 nigel 63 </P>
1164     <P>
1165     The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1166     the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1167     entry; both of these return an <b>int</b> value. The entry size depends on the
1168     length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1169     entry of the table (a pointer to <b>char</b>). The first two bytes of each entry
1170     are the number of the capturing parenthesis, most significant byte first. The
1171 ph10 461 rest of the entry is the corresponding name, zero terminated.
1172     </P>
1173     <P>
1174     The names are in alphabetical order. Duplicate names may appear if (?| is used
1175     to create multiple groups with the same number, as described in the
1176     <a href="pcrepattern.html#dupsubpatternnumber">section on duplicate subpattern numbers</a>
1177     in the
1178     <a href="pcrepattern.html"><b>pcrepattern</b></a>
1179     page. Duplicate names for subpatterns with different numbers are permitted only
1180     if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1181     table in the order in which they were found in the pattern. In the absence of
1182     (?| this is the order of increasing number; when (?| is used this is not
1183     necessarily the case because later subpatterns may have lower numbers.
1184     </P>
1185     <P>
1186     As a simple example of the name/number table, consider the following pattern
1187     (assume PCRE_EXTENDED is set, so white space - including newlines - is
1188     ignored):
1189 nigel 63 <pre>
1190 nigel 93 (?&#60;date&#62; (?&#60;year&#62;(\d\d)?\d\d) - (?&#60;month&#62;\d\d) - (?&#60;day&#62;\d\d) )
1191 nigel 75 </pre>
1192 nigel 63 There are four named subpatterns, so the table has four entries, and each entry
1193     in the table is eight bytes long. The table is as follows, with non-printing
1194 nigel 75 bytes shows in hexadecimal, and undefined bytes shown as ??:
1195 nigel 63 <pre>
1196     00 01 d a t e 00 ??
1197     00 05 d a y 00 ?? ??
1198     00 04 m o n t h 00
1199     00 02 y e a r 00 ??
1200 nigel 75 </pre>
1201     When writing code to extract data from named subpatterns using the
1202 nigel 91 name-to-number map, remember that the length of the entries is likely to be
1203 nigel 75 different for each compiled pattern.
1204 nigel 63 <pre>
1205 ph10 172 PCRE_INFO_OKPARTIAL
1206     </pre>
1207 ph10 453 Return 1 if the pattern can be used for partial matching with
1208     <b>pcre_exec()</b>, otherwise 0. The fourth argument should point to an
1209     <b>int</b> variable. From release 8.00, this always returns 1, because the
1210     restrictions that previously applied to partial matching have been lifted. The
1211 ph10 172 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1212 ph10 429 documentation gives details of partial matching.
1213 ph10 172 <pre>
1214 nigel 63 PCRE_INFO_OPTIONS
1215 nigel 75 </pre>
1216 nigel 63 Return a copy of the options with which the pattern was compiled. The fourth
1217     argument should point to an <b>unsigned long int</b> variable. These option bits
1218     are those specified in the call to <b>pcre_compile()</b>, modified by any
1219 ph10 197 top-level option settings at the start of the pattern itself. In other words,
1220     they are the options that will be in force when matching starts. For example,
1221     if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1222     result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1223 nigel 63 </P>
1224     <P>
1225     A pattern is automatically anchored by PCRE if all of its top-level
1226     alternatives begin with one of the following:
1227     <pre>
1228     ^ unless PCRE_MULTILINE is set
1229     \A always
1230     \G always
1231 nigel 75 .* if PCRE_DOTALL is set and there are no back references to the subpattern in which .* appears
1232     </pre>
1233 nigel 63 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1234     <b>pcre_fullinfo()</b>.
1235     <pre>
1236     PCRE_INFO_SIZE
1237 nigel 75 </pre>
1238 ph10 835 Return the size of the compiled pattern, that is, the value that was passed as
1239     the argument to <b>pcre_malloc()</b> when PCRE was getting memory in which to
1240     place the compiled data. The fourth argument should point to a <b>size_t</b>
1241     variable.
1242 nigel 63 <pre>
1243     PCRE_INFO_STUDYSIZE
1244 nigel 75 </pre>
1245 ph10 656 Return the size of the data block pointed to by the <i>study_data</i> field in a
1246     <b>pcre_extra</b> block. If <b>pcre_extra</b> is NULL, or there is no study data,
1247     zero is returned. The fourth argument should point to a <b>size_t</b> variable.
1248     The <i>study_data</i> field is set by <b>pcre_study()</b> to record information
1249     that will speed up matching (see the section entitled
1250     <a href="#studyingapattern">"Studying a pattern"</a>
1251     above). The format of the <i>study_data</i> block is private, but its length
1252     is made available via this option so that it can be saved and restored (see the
1253     <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1254     documentation for details).
1255 nigel 63 </P>
1256 ph10 691 <br><a name="SEC14" href="#TOC1">OBSOLETE INFO FUNCTION</a><br>
1257 nigel 63 <P>
1258     <b>int pcre_info(const pcre *<i>code</i>, int *<i>optptr</i>, int</b>
1259     <b>*<i>firstcharptr</i>);</b>
1260     </P>
1261     <P>
1262     The <b>pcre_info()</b> function is now obsolete because its interface is too
1263     restrictive to return all the available data about a compiled pattern. New
1264     programs should use <b>pcre_fullinfo()</b> instead. The yield of
1265     <b>pcre_info()</b> is the number of capturing subpatterns, or one of the
1266     following negative numbers:
1267     <pre>
1268     PCRE_ERROR_NULL the argument <i>code</i> was NULL
1269     PCRE_ERROR_BADMAGIC the "magic number" was not found
1270 nigel 75 </pre>
1271 nigel 63 If the <i>optptr</i> argument is not NULL, a copy of the options with which the
1272     pattern was compiled is placed in the integer it points to (see
1273     PCRE_INFO_OPTIONS above).
1274     </P>
1275     <P>
1276     If the pattern is not anchored and the <i>firstcharptr</i> argument is not NULL,
1277     it is used to pass back information about the first character of any matched
1278     string (see PCRE_INFO_FIRSTBYTE above).
1279     </P>
1280 ph10 691 <br><a name="SEC15" href="#TOC1">REFERENCE COUNTS</a><br>
1281 nigel 63 <P>
1282 nigel 77 <b>int pcre_refcount(pcre *<i>code</i>, int <i>adjust</i>);</b>
1283     </P>
1284     <P>
1285     The <b>pcre_refcount()</b> function is used to maintain a reference count in the
1286     data block that contains a compiled pattern. It is provided for the benefit of
1287     applications that operate in an object-oriented manner, where different parts
1288     of the application may be using the same compiled pattern, but you want to free
1289     the block when they are all done.
1290     </P>
1291     <P>
1292     When a pattern is compiled, the reference count field is initialized to zero.
1293     It is changed only by calling this function, whose action is to add the
1294     <i>adjust</i> value (which may be positive or negative) to it. The yield of the
1295     function is the new value. However, the value of the count is constrained to
1296     lie between 0 and 65535, inclusive. If the new value is outside these limits,
1297     it is forced to the appropriate limit value.
1298     </P>
1299     <P>
1300     Except when it is zero, the reference count is not correctly preserved if a
1301     pattern is compiled on one host and then transferred to a host whose byte-order
1302     is different. (This seems a highly unlikely scenario.)
1303     </P>
1304 ph10 691 <br><a name="SEC16" href="#TOC1">MATCHING A PATTERN: THE TRADITIONAL FUNCTION</a><br>
1305 nigel 77 <P>
1306 nigel 63 <b>int pcre_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
1307     <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
1308     <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>);</b>
1309     </P>
1310     <P>
1311     The function <b>pcre_exec()</b> is called to match a subject string against a
1312 nigel 75 compiled pattern, which is passed in the <i>code</i> argument. If the
1313 ph10 461 pattern was studied, the result of the study should be passed in the
1314 ph10 708 <i>extra</i> argument. You can call <b>pcre_exec()</b> with the same <i>code</i>
1315     and <i>extra</i> arguments as many times as you like, in order to match
1316     different subject strings with the same pattern.
1317     </P>
1318     <P>
1319     This function is the main matching facility of the library, and it operates in
1320     a Perl-like manner. For specialist use there is also an alternative matching
1321     function, which is described
1322 nigel 77 <a href="#dfamatch">below</a>
1323     in the section about the <b>pcre_dfa_exec()</b> function.
1324 nigel 63 </P>
1325     <P>
1326 nigel 75 In most applications, the pattern will have been compiled (and optionally
1327     studied) in the same process that calls <b>pcre_exec()</b>. However, it is
1328     possible to save compiled patterns and study data, and then use them later
1329     in different processes, possibly even on different hosts. For a discussion
1330     about this, see the
1331     <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1332     documentation.
1333 nigel 63 </P>
1334     <P>
1335 nigel 75 Here is an example of a simple call to <b>pcre_exec()</b>:
1336 nigel 63 <pre>
1337     int rc;
1338     int ovector[30];
1339     rc = pcre_exec(
1340     re, /* result of pcre_compile() */
1341     NULL, /* we didn't study the pattern */
1342     "some string", /* the subject string */
1343     11, /* the length of the subject string */
1344     0, /* start at offset 0 in the subject */
1345     0, /* default options */
1346 nigel 75 ovector, /* vector of integers for substring information */
1347 nigel 77 30); /* number of elements (NOT size in bytes) */
1348 nigel 75 <a name="extradata"></a></PRE>
1349 nigel 63 </P>
1350 nigel 75 <br><b>
1351     Extra data for <b>pcre_exec()</b>
1352     </b><br>
1353 nigel 63 <P>
1354     If the <i>extra</i> argument is not NULL, it must point to a <b>pcre_extra</b>
1355     data block. The <b>pcre_study()</b> function returns such a block (when it
1356     doesn't return NULL), but you can also create one for yourself, and pass
1357 nigel 87 additional information in it. The <b>pcre_extra</b> block contains the following
1358     fields (not necessarily in this order):
1359 nigel 63 <pre>
1360     unsigned long int <i>flags</i>;
1361     void *<i>study_data</i>;
1362 ph10 691 void *<i>executable_jit</i>;
1363 nigel 63 unsigned long int <i>match_limit</i>;
1364 nigel 87 unsigned long int <i>match_limit_recursion</i>;
1365 nigel 63 void *<i>callout_data</i>;
1366 nigel 75 const unsigned char *<i>tables</i>;
1367 ph10 512 unsigned char **<i>mark</i>;
1368 nigel 75 </pre>
1369 nigel 63 The <i>flags</i> field is a bitmap that specifies which of the other fields
1370     are set. The flag bits are:
1371     <pre>
1372     PCRE_EXTRA_STUDY_DATA
1373 ph10 691 PCRE_EXTRA_EXECUTABLE_JIT
1374 nigel 63 PCRE_EXTRA_MATCH_LIMIT
1375 nigel 87 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1376 nigel 63 PCRE_EXTRA_CALLOUT_DATA
1377 nigel 75 PCRE_EXTRA_TABLES
1378 ph10 512 PCRE_EXTRA_MARK
1379 nigel 75 </pre>
1380 ph10 691 Other flag bits should be set to zero. The <i>study_data</i> field and sometimes
1381     the <i>executable_jit</i> field are set in the <b>pcre_extra</b> block that is
1382     returned by <b>pcre_study()</b>, together with the appropriate flag bits. You
1383     should not set these yourself, but you may add to the block by setting the
1384     other fields and their corresponding flag bits.
1385 nigel 63 </P>
1386     <P>
1387     The <i>match_limit</i> field provides a means of preventing PCRE from using up a
1388     vast amount of resources when running patterns that are not going to match,
1389     but which have a very large number of possibilities in their search trees. The
1390 ph10 461 classic example is a pattern that uses nested unlimited repeats.
1391 nigel 63 </P>
1392     <P>
1393 ph10 691 Internally, <b>pcre_exec()</b> uses a function called <b>match()</b>, which it
1394     calls repeatedly (sometimes recursively). The limit set by <i>match_limit</i> is
1395     imposed on the number of times this function is called during a match, which
1396     has the effect of limiting the amount of backtracking that can take place. For
1397     patterns that are not anchored, the count restarts from zero for each position
1398     in the subject string.
1399 nigel 75 </P>
1400     <P>
1401 ph10 691 When <b>pcre_exec()</b> is called with a pattern that was successfully studied
1402     with the PCRE_STUDY_JIT_COMPILE option, the way that the matching is executed
1403     is entirely different. However, there is still the possibility of runaway
1404     matching that goes on for a very long time, and so the <i>match_limit</i> value
1405     is also used in this case (but in a different way) to limit how long the
1406     matching can continue.
1407     </P>
1408     <P>
1409 nigel 87 The default value for the limit can be set when PCRE is built; the default
1410 nigel 63 default is 10 million, which handles all but the most extreme cases. You can
1411 nigel 87 override the default by suppling <b>pcre_exec()</b> with a <b>pcre_extra</b>
1412     block in which <i>match_limit</i> is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1413     the <i>flags</i> field. If the limit is exceeded, <b>pcre_exec()</b> returns
1414     PCRE_ERROR_MATCHLIMIT.
1415 nigel 63 </P>
1416     <P>
1417 nigel 87 The <i>match_limit_recursion</i> field is similar to <i>match_limit</i>, but
1418     instead of limiting the total number of times that <b>match()</b> is called, it
1419     limits the depth of recursion. The recursion depth is a smaller number than the
1420     total number of calls, because not all calls to <b>match()</b> are recursive.
1421     This limit is of use only if it is set smaller than <i>match_limit</i>.
1422     </P>
1423     <P>
1424 ph10 691 Limiting the recursion depth limits the amount of machine stack that can be
1425     used, or, when PCRE has been compiled to use memory on the heap instead of the
1426     stack, the amount of heap memory that can be used. This limit is not relevant,
1427     and is ignored, if the pattern was successfully studied with
1428     PCRE_STUDY_JIT_COMPILE.
1429 nigel 87 </P>
1430     <P>
1431     The default value for <i>match_limit_recursion</i> can be set when PCRE is
1432     built; the default default is the same value as the default for
1433     <i>match_limit</i>. You can override the default by suppling <b>pcre_exec()</b>
1434     with a <b>pcre_extra</b> block in which <i>match_limit_recursion</i> is set, and
1435     PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the <i>flags</i> field. If the limit
1436     is exceeded, <b>pcre_exec()</b> returns PCRE_ERROR_RECURSIONLIMIT.
1437     </P>
1438     <P>
1439 ph10 453 The <i>callout_data</i> field is used in conjunction with the "callout" feature,
1440     and is described in the
1441 nigel 75 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1442     documentation.
1443 nigel 63 </P>
1444     <P>
1445 nigel 75 The <i>tables</i> field is used to pass a character tables pointer to
1446     <b>pcre_exec()</b>; this overrides the value that is stored with the compiled
1447     pattern. A non-NULL value is stored with the compiled pattern only if custom
1448     tables were supplied to <b>pcre_compile()</b> via its <i>tableptr</i> argument.
1449     If NULL is passed to <b>pcre_exec()</b> using this mechanism, it forces PCRE's
1450     internal tables to be used. This facility is helpful when re-using patterns
1451     that have been saved after compiling with an external set of tables, because
1452     the external tables might be at a different address when <b>pcre_exec()</b> is
1453     called. See the
1454     <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1455     documentation for a discussion of saving compiled patterns for later use.
1456 ph10 512 </P>
1457     <P>
1458     If PCRE_EXTRA_MARK is set in the <i>flags</i> field, the <i>mark</i> field must
1459     be set to point to a <b>char *</b> variable. If the pattern contains any
1460     backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1461     a name to pass back, a pointer to the name string (zero terminated) is placed
1462     in the variable pointed to by the <i>mark</i> field. The names are within the
1463     compiled pattern; if you wish to retain such a name you must copy it before
1464     freeing the memory of a compiled pattern. If there is no name to pass back, the
1465     variable pointed to by the <i>mark</i> field set to NULL. For details of the
1466     backtracking control verbs, see the section entitled
1467     <a href="pcrepattern#backtrackcontrol">"Backtracking control"</a>
1468     in the
1469     <a href="pcrepattern.html"><b>pcrepattern</b></a>
1470     documentation.
1471 ph10 227 <a name="execoptions"></a></P>
1472 nigel 75 <br><b>
1473     Option bits for <b>pcre_exec()</b>
1474     </b><br>
1475 nigel 63 <P>
1476 nigel 75 The unused bits of the <i>options</i> argument for <b>pcre_exec()</b> must be
1477 nigel 91 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
1478 ph10 453 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1479     PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_SOFT, and
1480     PCRE_PARTIAL_HARD.
1481 ph10 691 </P>
1482     <P>
1483     If the pattern was successfully studied with the PCRE_STUDY_JIT_COMPILE option,
1484     the only supported options for JIT execution are PCRE_NO_UTF8_CHECK,
1485     PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, and PCRE_NOTEMPTY_ATSTART. Note in
1486     particular that partial matching is not supported. If an unsupported option is
1487     used, JIT execution is disabled and the normal interpretive code in
1488     <b>pcre_exec()</b> is run.
1489 nigel 63 <pre>
1490 nigel 75 PCRE_ANCHORED
1491     </pre>
1492     The PCRE_ANCHORED option limits <b>pcre_exec()</b> to matching at the first
1493     matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1494     to be anchored by virtue of its contents, it cannot be made unachored at
1495     matching time.
1496     <pre>
1497 ph10 231 PCRE_BSR_ANYCRLF
1498     PCRE_BSR_UNICODE
1499     </pre>
1500     These options (which are mutually exclusive) control what the \R escape
1501     sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1502     match any Unicode newline sequence. These options override the choice that was
1503     made or defaulted when the pattern was compiled.
1504     <pre>
1505 nigel 91 PCRE_NEWLINE_CR
1506     PCRE_NEWLINE_LF
1507     PCRE_NEWLINE_CRLF
1508 ph10 150 PCRE_NEWLINE_ANYCRLF
1509 nigel 93 PCRE_NEWLINE_ANY
1510 nigel 91 </pre>
1511     These options override the newline definition that was chosen or defaulted when
1512 nigel 93 the pattern was compiled. For details, see the description of
1513     <b>pcre_compile()</b> above. During matching, the newline choice affects the
1514     behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1515     the way the match position is advanced after a match failure for an unanchored
1516 ph10 227 pattern.
1517     </P>
1518     <P>
1519     When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1520     match attempt for an unanchored pattern fails when the current position is at a
1521 ph10 231 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1522 ph10 227 characters, the match position is advanced by two characters instead of one, in
1523     other words, to after the CRLF.
1524     </P>
1525     <P>
1526     The above rule is a compromise that makes the most common cases work as
1527     expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1528     set), it does not match the string "\r\nA" because, after failing at the
1529     start, it skips both the CR and the LF before retrying. However, the pattern
1530     [\r\n]A does match that string, because it contains an explicit CR or LF
1531     reference, and so advances only by one character after the first failure.
1532     </P>
1533     <P>
1534 ph10 231 An explicit match for CR of LF is either a literal appearance of one of those
1535     characters, or one of the \r or \n escape sequences. Implicit matches such as
1536     [^X] do not count, nor does \s (which includes CR and LF in the characters
1537     that it matches).
1538     </P>
1539     <P>
1540 ph10 227 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1541     valid newline sequence and explicit \r or \n escapes appear in the pattern.
1542 nigel 91 <pre>
1543 nigel 63 PCRE_NOTBOL
1544 nigel 75 </pre>
1545     This option specifies that first character of the subject string is not the
1546     beginning of a line, so the circumflex metacharacter should not match before
1547     it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1548     never to match. This option affects only the behaviour of the circumflex
1549     metacharacter. It does not affect \A.
1550 nigel 63 <pre>
1551     PCRE_NOTEOL
1552 nigel 75 </pre>
1553     This option specifies that the end of the subject string is not the end of a
1554     line, so the dollar metacharacter should not match it nor (except in multiline
1555     mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1556     compile time) causes dollar never to match. This option affects only the
1557     behaviour of the dollar metacharacter. It does not affect \Z or \z.
1558 nigel 63 <pre>
1559     PCRE_NOTEMPTY
1560 nigel 75 </pre>
1561 nigel 63 An empty string is not considered to be a valid match if this option is set. If
1562     there are alternatives in the pattern, they are tried. If all the alternatives
1563     match the empty string, the entire match fails. For example, if the pattern
1564     <pre>
1565     a?b?
1566 nigel 75 </pre>
1567 ph10 453 is applied to a string not beginning with "a" or "b", it matches an empty
1568 nigel 63 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1569     valid, so PCRE searches further into the string for occurrences of "a" or "b".
1570 ph10 453 <pre>
1571     PCRE_NOTEMPTY_ATSTART
1572     </pre>
1573 ph10 461 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1574 ph10 453 the start of the subject is permitted. If the pattern is anchored, such a match
1575     can occur only if the pattern contains \K.
1576 nigel 63 </P>
1577     <P>
1578 ph10 453 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1579     does make a special case of a pattern match of the empty string within its
1580     <b>split()</b> function, and when using the /g modifier. It is possible to
1581     emulate Perl's behaviour after matching a null string by first trying the match
1582     again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1583     if that fails, by advancing the starting offset (see below) and trying an
1584     ordinary match again. There is some code that demonstrates how to do this in
1585     the
1586 ph10 429 <a href="pcredemo.html"><b>pcredemo</b></a>
1587 ph10 572 sample program. In the most general case, you have to check to see if the
1588     newline convention recognizes CRLF as a newline, and if so, and the current
1589 ph10 567 character is CR followed by LF, advance the starting offset by two characters
1590     instead of one.
1591 nigel 75 <pre>
1592 ph10 392 PCRE_NO_START_OPTIMIZE
1593     </pre>
1594     There are a number of optimizations that <b>pcre_exec()</b> uses at the start of
1595 ph10 545 a match, in order to speed up the process. For example, if it is known that an
1596     unanchored match must start with a specific character, it searches the subject
1597     for that character, and fails immediately if it cannot find it, without
1598     actually running the main matching function. This means that a special item
1599     such as (*COMMIT) at the start of a pattern is not considered until after a
1600 ph10 548 suitable starting point for the match has been found. When callouts or (*MARK)
1601     items are in use, these "start-up" optimizations can cause them to be skipped
1602     if the pattern is never actually used. The start-up optimizations are in effect
1603     a pre-scan of the subject that takes place before the pattern is run.
1604     </P>
1605     <P>
1606     The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1607     causing performance to suffer, but ensuring that in cases where the result is
1608     "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1609 ph10 579 are considered at every possible starting position in the subject string. If
1610     PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1611     time.
1612     </P>
1613     <P>
1614 ph10 548 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1615     Consider the pattern
1616 ph10 392 <pre>
1617 ph10 548 (*COMMIT)ABC
1618     </pre>
1619     When this is compiled, PCRE records the fact that a match must start with the
1620     character "A". Suppose the subject string is "DEFABC". The start-up
1621     optimization scans along the subject, finds "A" and runs the first match
1622     attempt from there. The (*COMMIT) item means that the pattern must match the
1623     current starting position, which in this case, it does. However, if the same
1624     match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1625     subject string does not happen. The first match attempt is run starting from
1626     "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1627     the overall result is "no match". If the pattern is studied, more start-up
1628     optimizations may be used. For example, a minimum length for the subject may be
1629     recorded. Consider the pattern
1630     <pre>
1631     (*MARK:A)(X|Y)
1632     </pre>
1633     The minimum length for a match is one character. If the subject is "ABC", there
1634     will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1635     If the pattern is studied, the final attempt does not take place, because PCRE
1636     knows that the subject is too short, and so the (*MARK) is never encountered.
1637     In this case, studying the pattern does not affect the overall match result,
1638     which is still "no match", but it does affect the auxiliary information that is
1639     returned.
1640     <pre>
1641 nigel 75 PCRE_NO_UTF8_CHECK
1642     </pre>
1643     When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1644     string is automatically checked when <b>pcre_exec()</b> is subsequently called.
1645     The value of <i>startoffset</i> is also checked to ensure that it points to the
1646 ph10 211 start of a UTF-8 character. There is a discussion about the validity of UTF-8
1647     strings in the
1648     <a href="pcre.html#utf8strings">section on UTF-8 support</a>
1649     in the main
1650     <a href="pcre.html"><b>pcre</b></a>
1651     page. If an invalid UTF-8 sequence of bytes is found, <b>pcre_exec()</b> returns
1652 ph10 572 the error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is
1653 ph10 654 a truncated UTF-8 character at the end of the subject, PCRE_ERROR_SHORTUTF8. In
1654     both cases, information about the precise nature of the error may also be
1655     returned (see the descriptions of these errors in the section entitled \fIError
1656     return values from\fP <b>pcre_exec()</b>
1657     <a href="#errorlist">below).</a>
1658     If <i>startoffset</i> contains a value that does not point to the start of a
1659     UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1660 ph10 572 returned.
1661 nigel 63 </P>
1662     <P>
1663 nigel 75 If you already know that your subject is valid, and you want to skip these
1664     checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1665     calling <b>pcre_exec()</b>. You might want to do this for the second and
1666     subsequent calls to <b>pcre_exec()</b> if you are making repeated calls to find
1667     all the matches in a single subject string. However, you should be sure that
1668 ph10 567 the value of <i>startoffset</i> points to the start of a UTF-8 character (or the
1669     end of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1670     invalid UTF-8 string as a subject or an invalid value of <i>startoffset</i> is
1671     undefined. Your program may crash.
1672 nigel 75 <pre>
1673 ph10 461 PCRE_PARTIAL_HARD
1674 ph10 429 PCRE_PARTIAL_SOFT
1675 nigel 75 </pre>
1676 ph10 429 These options turn on the partial matching feature. For backwards
1677     compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1678     occurs if the end of the subject string is reached successfully, but there are
1679     not enough subject characters to complete the match. If this happens when
1680 ph10 567 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1681     testing any remaining alternatives. Only if no complete match can be found is
1682     PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1683     PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1684     but only if no complete match can be found.
1685     </P>
1686     <P>
1687     If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1688     partial match is found, <b>pcre_exec()</b> immediately returns
1689     PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1690 ph10 572 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1691 ph10 567 important that an alternative complete match.
1692     </P>
1693     <P>
1694     In both cases, the portion of the string that was inspected when the partial
1695     match was found is set as the first matching string. There is a more detailed
1696     discussion of partial and multi-segment matching, with examples, in the
1697 nigel 75 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1698     documentation.
1699     </P>
1700     <br><b>
1701     The string to be matched by <b>pcre_exec()</b>
1702     </b><br>
1703     <P>
1704 nigel 63 The subject string is passed to <b>pcre_exec()</b> as a pointer in
1705 ph10 371 <i>subject</i>, a length (in bytes) in <i>length</i>, and a starting byte offset
1706 ph10 572 in <i>startoffset</i>. If this is negative or greater than the length of the
1707     subject, <b>pcre_exec()</b> returns PCRE_ERROR_BADOFFSET. When the starting
1708     offset is zero, the search for a match starts at the beginning of the subject,
1709     and this is by far the most common case. In UTF-8 mode, the byte offset must
1710     point to the start of a UTF-8 character (or the end of the subject). Unlike the
1711     pattern string, the subject may contain binary zero bytes.
1712 nigel 63 </P>
1713     <P>
1714     A non-zero starting offset is useful when searching for another match in the
1715     same subject by calling <b>pcre_exec()</b> again after a previous success.
1716     Setting <i>startoffset</i> differs from just passing over a shortened string and
1717     setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1718     lookbehind. For example, consider the pattern
1719     <pre>
1720     \Biss\B
1721 nigel 75 </pre>
1722 nigel 63 which finds occurrences of "iss" in the middle of words. (\B matches only if
1723     the current position in the subject is not a word boundary.) When applied to
1724     the string "Mississipi" the first call to <b>pcre_exec()</b> finds the first
1725     occurrence. If <b>pcre_exec()</b> is called again with just the remainder of the
1726     subject, namely "issipi", it does not match, because \B is always false at the
1727     start of the subject, which is deemed to be a word boundary. However, if
1728     <b>pcre_exec()</b> is passed the entire string again, but with <i>startoffset</i>
1729     set to 4, it finds the second occurrence of "iss" because it is able to look
1730     behind the starting point to discover that it is preceded by a letter.
1731     </P>
1732     <P>
1733 ph10 567 Finding all the matches in a subject is tricky when the pattern can match an
1734     empty string. It is possible to emulate Perl's /g behaviour by first trying the
1735     match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1736     PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1737     and trying an ordinary match again. There is some code that demonstrates how to
1738     do this in the
1739     <a href="pcredemo.html"><b>pcredemo</b></a>
1740 ph10 572 sample program. In the most general case, you have to check to see if the
1741     newline convention recognizes CRLF as a newline, and if so, and the current
1742 ph10 567 character is CR followed by LF, advance the starting offset by two characters
1743     instead of one.
1744     </P>
1745     <P>
1746 nigel 63 If a non-zero starting offset is passed when the pattern is anchored, one
1747 nigel 75 attempt to match at the given offset is made. This can only succeed if the
1748 nigel 63 pattern does not require the match to be at the start of the subject.
1749     </P>
1750 nigel 75 <br><b>
1751     How <b>pcre_exec()</b> returns captured substrings
1752     </b><br>
1753 nigel 63 <P>
1754     In general, a pattern matches a certain portion of the subject, and in
1755     addition, further substrings from the subject may be picked out by parts of the
1756     pattern. Following the usage in Jeffrey Friedl's book, this is called
1757     "capturing" in what follows, and the phrase "capturing subpattern" is used for
1758     a fragment of a pattern that picks out a substring. PCRE supports several other
1759     kinds of parenthesized subpattern that do not cause substrings to be captured.
1760     </P>
1761     <P>
1762 ph10 371 Captured substrings are returned to the caller via a vector of integers whose
1763     address is passed in <i>ovector</i>. The number of elements in the vector is
1764     passed in <i>ovecsize</i>, which must be a non-negative number. <b>Note</b>: this
1765     argument is NOT the size of <i>ovector</i> in bytes.
1766 nigel 63 </P>
1767     <P>
1768 nigel 75 The first two-thirds of the vector is used to pass back captured substrings,
1769     each substring using a pair of integers. The remaining third of the vector is
1770     used as workspace by <b>pcre_exec()</b> while matching capturing subpatterns,
1771 ph10 371 and is not available for passing back information. The number passed in
1772 nigel 75 <i>ovecsize</i> should always be a multiple of three. If it is not, it is
1773     rounded down.
1774     </P>
1775     <P>
1776     When a match is successful, information about captured substrings is returned
1777     in pairs of integers, starting at the beginning of <i>ovector</i>, and
1778 ph10 371 continuing up to two-thirds of its length at the most. The first element of
1779     each pair is set to the byte offset of the first character in a substring, and
1780     the second is set to the byte offset of the first character after the end of a
1781     substring. <b>Note</b>: these values are always byte offsets, even in UTF-8
1782     mode. They are not character counts.
1783 nigel 63 </P>
1784     <P>
1785 ph10 371 The first pair of integers, <i>ovector[0]</i> and <i>ovector[1]</i>, identify the
1786     portion of the subject string matched by the entire pattern. The next pair is
1787     used for the first capturing subpattern, and so on. The value returned by
1788     <b>pcre_exec()</b> is one more than the highest numbered pair that has been set.
1789     For example, if two substrings have been captured, the returned value is 3. If
1790     there are no capturing subpatterns, the return value from a successful match is
1791     1, indicating that just the first pair of offsets has been set.
1792     </P>
1793     <P>
1794 nigel 63 If a capturing subpattern is matched repeatedly, it is the last portion of the
1795 nigel 75 string that it matched that is returned.
1796 nigel 63 </P>
1797     <P>
1798 nigel 75 If the vector is too small to hold all the captured substring offsets, it is
1799     used as far as possible (up to two-thirds of its length), and the function
1800 ph10 691 returns a value of zero. If neither the actual string matched not any captured
1801     substrings are of interest, <b>pcre_exec()</b> may be called with <i>ovector</i>
1802     passed as NULL and <i>ovecsize</i> as zero. However, if the pattern contains
1803     back references and the <i>ovector</i> is not big enough to remember the related
1804     substrings, PCRE has to get additional memory for use during matching. Thus it
1805     is usually advisable to supply an <i>ovector</i> of reasonable size.
1806 nigel 63 </P>
1807     <P>
1808 ph10 691 There are some cases where zero is returned (indicating vector overflow) when
1809     in fact the vector is exactly the right size for the final match. For example,
1810     consider the pattern
1811     <pre>
1812     (a)(?:(b)c|bd)
1813     </pre>
1814     If a vector of 6 elements (allowing for only 1 captured substring) is given
1815     with subject string "abd", <b>pcre_exec()</b> will try to set the second
1816     captured string, thereby recording a vector overflow, before failing to match
1817     "c" and backing up to try the second alternative. The zero return, however,
1818     does correctly indicate that the maximum number of slots (namely 2) have been
1819     filled. In similar cases where there is temporary overflow, but the final
1820     number of used slots is actually less than the maximum, a non-zero value is
1821     returned.
1822     </P>
1823     <P>
1824 ph10 461 The <b>pcre_fullinfo()</b> function can be used to find out how many capturing
1825 nigel 63 subpatterns there are in a compiled pattern. The smallest size for
1826     <i>ovector</i> that will allow for <i>n</i> captured substrings, in addition to
1827     the offsets of the substring matched by the whole pattern, is (<i>n</i>+1)*3.
1828 nigel 91 </P>
1829     <P>
1830     It is possible for capturing subpattern number <i>n+1</i> to match some part of
1831     the subject when subpattern <i>n</i> has not been used at all. For example, if
1832     the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1833     function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1834     happens, both values in the offset pairs corresponding to unused subpatterns
1835     are set to -1.
1836     </P>
1837     <P>
1838     Offset values that correspond to unused subpatterns at the end of the
1839     expression are also set to -1. For example, if the string "abc" is matched
1840     against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1841     return from the function is 2, because the highest used capturing subpattern
1842 ph10 572 number is 1, and the offsets for for the second and third capturing subpatterns
1843     (assuming the vector is large enough, of course) are set to -1.
1844 nigel 91 </P>
1845     <P>
1846 ph10 691 <b>Note</b>: Elements in the first two-thirds of <i>ovector</i> that do not
1847     correspond to capturing parentheses in the pattern are never changed. That is,
1848     if a pattern contains <i>n</i> capturing parentheses, no more than
1849     <i>ovector[0]</i> to <i>ovector[2n+1]</i> are set by <b>pcre_exec()</b>. The other
1850     elements (in the first two-thirds) retain whatever values they previously had.
1851 ph10 572 </P>
1852     <P>
1853 nigel 91 Some convenience functions are provided for extracting the captured substrings
1854     as separate strings. These are described below.
1855 nigel 77 <a name="errorlist"></a></P>
1856 nigel 75 <br><b>
1857 nigel 91 Error return values from <b>pcre_exec()</b>
1858 nigel 75 </b><br>
1859 nigel 63 <P>
1860     If <b>pcre_exec()</b> fails, it returns a negative number. The following are
1861     defined in the header file:
1862     <pre>
1863     PCRE_ERROR_NOMATCH (-1)
1864 nigel 75 </pre>
1865 nigel 63 The subject string did not match the pattern.
1866     <pre>
1867     PCRE_ERROR_NULL (-2)
1868 nigel 75 </pre>
1869 nigel 63 Either <i>code</i> or <i>subject</i> was passed as NULL, or <i>ovector</i> was
1870     NULL and <i>ovecsize</i> was not zero.
1871     <pre>
1872     PCRE_ERROR_BADOPTION (-3)
1873 nigel 75 </pre>
1874 nigel 63 An unrecognized bit was set in the <i>options</i> argument.
1875     <pre>
1876     PCRE_ERROR_BADMAGIC (-4)
1877 nigel 75 </pre>
1878 nigel 63 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1879 nigel 75 the case when it is passed a junk pointer and to detect when a pattern that was
1880     compiled in an environment of one endianness is run in an environment with the
1881     other endianness. This is the error that PCRE gives when the magic number is
1882     not present.
1883 nigel 63 <pre>
1884 nigel 93 PCRE_ERROR_UNKNOWN_OPCODE (-5)
1885 nigel 75 </pre>
1886 nigel 63 While running the pattern match, an unknown item was encountered in the
1887     compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1888     of the compiled pattern.
1889     <pre>
1890     PCRE_ERROR_NOMEMORY (-6)
1891 nigel 75 </pre>
1892 nigel 63 If a pattern contains back references, but the <i>ovector</i> that is passed to
1893     <b>pcre_exec()</b> is not big enough to remember the referenced substrings, PCRE
1894     gets a block of memory at the start of matching to use for this purpose. If the
1895 nigel 75 call via <b>pcre_malloc()</b> fails, this error is given. The memory is
1896     automatically freed at the end of matching.
1897 ph10 535 </P>
1898     <P>
1899     This error is also given if <b>pcre_stack_malloc()</b> fails in
1900     <b>pcre_exec()</b>. This can happen only when PCRE has been compiled with
1901     <b>--disable-stack-for-recursion</b>.
1902 nigel 63 <pre>
1903     PCRE_ERROR_NOSUBSTRING (-7)
1904 nigel 75 </pre>
1905 nigel 63 This error is used by the <b>pcre_copy_substring()</b>,
1906     <b>pcre_get_substring()</b>, and <b>pcre_get_substring_list()</b> functions (see
1907     below). It is never returned by <b>pcre_exec()</b>.
1908     <pre>
1909     PCRE_ERROR_MATCHLIMIT (-8)
1910 nigel 75 </pre>
1911 nigel 87 The backtracking limit, as specified by the <i>match_limit</i> field in a
1912     <b>pcre_extra</b> structure (or defaulted) was reached. See the description
1913     above.
1914     <pre>
1915 nigel 63 PCRE_ERROR_CALLOUT (-9)
1916 nigel 75 </pre>
1917 nigel 63 This error is never generated by <b>pcre_exec()</b> itself. It is provided for
1918     use by callout functions that want to yield a distinctive error code. See the
1919 nigel 75 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1920     documentation for details.
1921 nigel 71 <pre>
1922 nigel 73 PCRE_ERROR_BADUTF8 (-10)
1923 nigel 75 </pre>
1924 ph10 654 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
1925     and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
1926     (<i>ovecsize</i>) is at least 2, the byte offset to the start of the the invalid
1927     UTF-8 character is placed in the first element, and a reason code is placed in
1928     the second element. The reason codes are listed in the
1929     <a href="#badutf8reasons">following section.</a>
1930     For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
1931     truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
1932     PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
1933 nigel 73 <pre>
1934     PCRE_ERROR_BADUTF8_OFFSET (-11)
1935 nigel 75 </pre>
1936 ph10 654 The UTF-8 byte sequence that was passed as a subject was checked and found to
1937     be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
1938     <i>startoffset</i> did not point to the beginning of a UTF-8 character or the
1939 ph10 572 end of the subject.
1940 nigel 75 <pre>
1941 nigel 77 PCRE_ERROR_PARTIAL (-12)
1942 nigel 75 </pre>
1943     The subject string did not match, but it did match partially. See the
1944     <a href="pcrepartial.html"><b>pcrepartial</b></a>
1945     documentation for details of partial matching.
1946     <pre>
1947 nigel 77 PCRE_ERROR_BADPARTIAL (-13)
1948 nigel 75 </pre>
1949 ph10 429 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1950     option was used with a compiled pattern containing items that were not
1951 ph10 461 supported for partial matching. From release 8.00 onwards, there are no
1952 ph10 429 restrictions on partial matching.
1953 nigel 75 <pre>
1954 nigel 77 PCRE_ERROR_INTERNAL (-14)
1955 nigel 75 </pre>
1956     An unexpected internal error has occurred. This error could be caused by a bug
1957     in PCRE or by overwriting of the compiled pattern.
1958     <pre>
1959 nigel 77 PCRE_ERROR_BADCOUNT (-15)
1960 nigel 75 </pre>
1961     This error is given if the value of the <i>ovecsize</i> argument is negative.
1962 nigel 93 <pre>
1963     PCRE_ERROR_RECURSIONLIMIT (-21)
1964     </pre>
1965     The internal recursion limit, as specified by the <i>match_limit_recursion</i>
1966     field in a <b>pcre_extra</b> structure (or defaulted) was reached. See the
1967     description above.
1968     <pre>
1969     PCRE_ERROR_BADNEWLINE (-23)
1970     </pre>
1971     An invalid combination of PCRE_NEWLINE_<i>xxx</i> options was given.
1972 ph10 567 <pre>
1973     PCRE_ERROR_BADOFFSET (-24)
1974     </pre>
1975 ph10 572 The value of <i>startoffset</i> was negative or greater than the length of the
1976 ph10 567 subject, that is, the value in <i>length</i>.
1977 ph10 572 <pre>
1978     PCRE_ERROR_SHORTUTF8 (-25)
1979     </pre>
1980 ph10 654 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
1981     ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
1982     Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
1983     fact sufficient to detect this case, but this special error code for
1984     PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
1985     retained for backwards compatibility.
1986     <pre>
1987     PCRE_ERROR_RECURSELOOP (-26)
1988     </pre>
1989     This error is returned when <b>pcre_exec()</b> detects a recursion loop within
1990     the pattern. Specifically, it means that either the whole pattern or a
1991     subpattern has been called recursively for the second time at the same position
1992     in the subject string. Some simple patterns that might do this are detected and
1993     faulted at compile time, but more complicated cases, in particular mutual
1994     recursions between two different subpatterns, cannot be detected until run
1995     time.
1996 ph10 691 <pre>
1997     PCRE_ERROR_JIT_STACKLIMIT (-27)
1998     </pre>
1999     This error is returned when a pattern that was successfully studied using the
2000     PCRE_STUDY_JIT_COMPILE option is being matched, but the memory available for
2001     the just-in-time processing stack is not large enough. See the
2002     <a href="pcrejit.html"><b>pcrejit</b></a>
2003     documentation for more details.
2004 nigel 73 </P>
2005 nigel 93 <P>
2006 ph10 197 Error numbers -16 to -20 and -22 are not used by <b>pcre_exec()</b>.
2007 ph10 654 <a name="badutf8reasons"></a></P>
2008     <br><b>
2009     Reason codes for invalid UTF-8 strings
2010     </b><br>
2011     <P>
2012     When <b>pcre_exec()</b> returns either PCRE_ERROR_BADUTF8 or
2013     PCRE_ERROR_SHORTUTF8, and the size of the output vector (<i>ovecsize</i>) is at
2014     least 2, the offset of the start of the invalid UTF-8 character is placed in
2015     the first output vector element (<i>ovector[0]</i>) and a reason code is placed
2016     in the second element (<i>ovector[1]</i>). The reason codes are given names in
2017     the <b>pcre.h</b> header file:
2018     <pre>
2019     PCRE_UTF8_ERR1
2020     PCRE_UTF8_ERR2
2021     PCRE_UTF8_ERR3
2022     PCRE_UTF8_ERR4
2023     PCRE_UTF8_ERR5
2024     </pre>
2025     The string ends with a truncated UTF-8 character; the code specifies how many
2026     bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2027     no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2028     allows for up to 6 bytes, and this is checked first; hence the possibility of
2029     4 or 5 missing bytes.
2030     <pre>
2031     PCRE_UTF8_ERR6
2032     PCRE_UTF8_ERR7
2033     PCRE_UTF8_ERR8
2034     PCRE_UTF8_ERR9
2035     PCRE_UTF8_ERR10
2036     </pre>
2037     The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2038     character do not have the binary value 0b10 (that is, either the most
2039     significant bit is 0, or the next bit is 1).
2040     <pre>
2041     PCRE_UTF8_ERR11
2042     PCRE_UTF8_ERR12
2043     </pre>
2044     A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2045     these code points are excluded by RFC 3629.
2046     <pre>
2047     PCRE_UTF8_ERR13
2048     </pre>
2049     A 4-byte character has a value greater than 0x10fff; these code points are
2050     excluded by RFC 3629.
2051     <pre>
2052     PCRE_UTF8_ERR14
2053     </pre>
2054     A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2055     code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2056     from UTF-8.
2057     <pre>
2058     PCRE_UTF8_ERR15
2059     PCRE_UTF8_ERR16
2060     PCRE_UTF8_ERR17
2061     PCRE_UTF8_ERR18
2062     PCRE_UTF8_ERR19
2063     </pre>
2064     A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2065     value that can be represented by fewer bytes, which is invalid. For example,
2066     the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2067     one byte.
2068     <pre>
2069     PCRE_UTF8_ERR20
2070     </pre>
2071     The two most significant bits of the first byte of a character have the binary
2072     value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2073     byte can only validly occur as the second or subsequent byte of a multi-byte
2074     character.
2075     <pre>
2076     PCRE_UTF8_ERR21
2077     </pre>
2078     The first byte of a character has the value 0xfe or 0xff. These values can
2079     never occur in a valid UTF-8 string.
2080 nigel 93 </P>
2081 ph10 691 <br><a name="SEC17" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NUMBER</a><br>
2082 nigel 63 <P>
2083     <b>int pcre_copy_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
2084     <b>int <i>stringcount</i>, int <i>stringnumber</i>, char *<i>buffer</i>,</b>
2085     <b>int <i>buffersize</i>);</b>
2086     </P>
2087     <P>
2088     <b>int pcre_get_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
2089     <b>int <i>stringcount</i>, int <i>stringnumber</i>,</b>
2090     <b>const char **<i>stringptr</i>);</b>
2091     </P>
2092     <P>
2093     <b>int pcre_get_substring_list(const char *<i>subject</i>,</b>
2094     <b>int *<i>ovector</i>, int <i>stringcount</i>, const char ***<i>listptr</i>);</b>
2095     </P>
2096     <P>
2097     Captured substrings can be accessed directly by using the offsets returned by
2098     <b>pcre_exec()</b> in <i>ovector</i>. For convenience, the functions
2099     <b>pcre_copy_substring()</b>, <b>pcre_get_substring()</b>, and
2100     <b>pcre_get_substring_list()</b> are provided for extracting captured substrings
2101     as new, separate, zero-terminated strings. These functions identify substrings
2102     by number. The next section describes functions for extracting named
2103 nigel 91 substrings.
2104 nigel 63 </P>
2105     <P>
2106 nigel 91 A substring that contains a binary zero is correctly extracted and has a
2107     further zero added on the end, but the result is not, of course, a C string.
2108     However, you can process such a string by referring to the length that is
2109     returned by <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>.
2110     Unfortunately, the interface to <b>pcre_get_substring_list()</b> is not adequate
2111     for handling strings containing binary zeros, because the end of the final
2112     string is not independently indicated.
2113     </P>
2114     <P>
2115 nigel 63 The first three arguments are the same for all three of these functions:
2116 nigel 75 <i>subject</i> is the subject string that has just been successfully matched,
2117 nigel 63 <i>ovector</i> is a pointer to the vector of integer offsets that was passed to
2118     <b>pcre_exec()</b>, and <i>stringcount</i> is the number of substrings that were
2119     captured by the match, including the substring that matched the entire regular
2120 nigel 75 expression. This is the value returned by <b>pcre_exec()</b> if it is greater
2121     than zero. If <b>pcre_exec()</b> returned zero, indicating that it ran out of
2122     space in <i>ovector</i>, the value passed as <i>stringcount</i> should be the
2123     number of elements in the vector divided by three.
2124 nigel 63 </P>
2125     <P>
2126     The functions <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>
2127     extract a single substring, whose number is given as <i>stringnumber</i>. A
2128 nigel 75 value of zero extracts the substring that matched the entire pattern, whereas
2129 nigel 63 higher values extract the captured substrings. For <b>pcre_copy_substring()</b>,
2130     the string is placed in <i>buffer</i>, whose length is given by
2131     <i>buffersize</i>, while for <b>pcre_get_substring()</b> a new block of memory is
2132     obtained via <b>pcre_malloc</b>, and its address is returned via
2133     <i>stringptr</i>. The yield of the function is the length of the string, not
2134 nigel 93 including the terminating zero, or one of these error codes:
2135 nigel 63 <pre>
2136     PCRE_ERROR_NOMEMORY (-6)
2137 nigel 75 </pre>
2138 nigel 63 The buffer was too small for <b>pcre_copy_substring()</b>, or the attempt to get
2139     memory failed for <b>pcre_get_substring()</b>.
2140     <pre>
2141     PCRE_ERROR_NOSUBSTRING (-7)
2142 nigel 75 </pre>
2143 nigel 63 There is no substring whose number is <i>stringnumber</i>.
2144     </P>
2145     <P>
2146     The <b>pcre_get_substring_list()</b> function extracts all available substrings
2147     and builds a list of pointers to them. All this is done in a single block of
2148 nigel 75 memory that is obtained via <b>pcre_malloc</b>. The address of the memory block
2149 nigel 63 is returned via <i>listptr</i>, which is also the start of the list of string
2150     pointers. The end of the list is marked by a NULL pointer. The yield of the
2151 nigel 93 function is zero if all went well, or the error code
2152 nigel 63 <pre>
2153     PCRE_ERROR_NOMEMORY (-6)
2154 nigel 75 </pre>
2155 nigel 63 if the attempt to get the memory block failed.
2156     </P>
2157     <P>
2158     When any of these functions encounter a substring that is unset, which can
2159     happen when capturing subpattern number <i>n+1</i> matches some part of the
2160     subject, but subpattern <i>n</i> has not been used at all, they return an empty
2161     string. This can be distinguished from a genuine zero-length substring by
2162     inspecting the appropriate offset in <i>ovector</i>, which is negative for unset
2163     substrings.
2164     </P>
2165     <P>
2166     The two convenience functions <b>pcre_free_substring()</b> and
2167     <b>pcre_free_substring_list()</b> can be used to free the memory returned by
2168     a previous call of <b>pcre_get_substring()</b> or
2169     <b>pcre_get_substring_list()</b>, respectively. They do nothing more than call
2170     the function pointed to by <b>pcre_free</b>, which of course could be called
2171     directly from a C program. However, PCRE is used in some situations where it is
2172 nigel 91 linked via a special interface to another programming language that cannot use
2173 nigel 63 <b>pcre_free</b> directly; it is for these cases that the functions are
2174     provided.
2175     </P>
2176 ph10 691 <br><a name="SEC18" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NAME</a><br>
2177 nigel 63 <P>
2178 nigel 75 <b>int pcre_get_stringnumber(const pcre *<i>code</i>,</b>
2179     <b>const char *<i>name</i>);</b>
2180     </P>
2181     <P>
2182 nigel 63 <b>int pcre_copy_named_substring(const pcre *<i>code</i>,</b>
2183     <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
2184     <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
2185     <b>char *<i>buffer</i>, int <i>buffersize</i>);</b>
2186     </P>
2187     <P>
2188     <b>int pcre_get_named_substring(const pcre *<i>code</i>,</b>
2189     <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
2190     <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
2191     <b>const char **<i>stringptr</i>);</b>
2192     </P>
2193     <P>
2194 nigel 75 To extract a substring by name, you first have to find associated number.
2195     For example, for this pattern
2196 nigel 63 <pre>
2197 nigel 93 (a+)b(?&#60;xxx&#62;\d+)...
2198 nigel 75 </pre>
2199 nigel 91 the number of the subpattern called "xxx" is 2. If the name is known to be
2200     unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2201     calling <b>pcre_get_stringnumber()</b>. The first argument is the compiled
2202     pattern, and the second is the name. The yield of the function is the
2203 nigel 75 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2204     that name.
2205 nigel 63 </P>
2206     <P>
2207 nigel 75 Given the number, you can extract the substring directly, or use one of the
2208     functions described in the previous section. For convenience, there are also
2209     two functions that do the whole job.
2210 nigel 63 </P>
2211     <P>
2212 nigel 91 Most of the arguments of <b>pcre_copy_named_substring()</b> and
2213     <b>pcre_get_named_substring()</b> are the same as those for the similarly named
2214 nigel 75 functions that extract by number. As these are described in the previous
2215     section, they are not re-described here. There are just two differences:
2216 nigel 63 </P>
2217     <P>
2218     First, instead of a substring number, a substring name is given. Second, there
2219     is an extra argument, given at the start, which is a pointer to the compiled
2220     pattern. This is needed in order to gain access to the name-to-number
2221     translation table.
2222     </P>
2223     <P>
2224     These functions call <b>pcre_get_stringnumber()</b>, and if it succeeds, they
2225 ph10 128 then call <b>pcre_copy_substring()</b> or <b>pcre_get_substring()</b>, as
2226     appropriate. <b>NOTE:</b> If PCRE_DUPNAMES is set and there are duplicate names,
2227     the behaviour may not be what you want (see the next section).
2228 nigel 63 </P>
2229 ph10 392 <P>
2230 ph10 461 <b>Warning:</b> If the pattern uses the (?| feature to set up multiple
2231     subpatterns with the same number, as described in the
2232     <a href="pcrepattern.html#dupsubpatternnumber">section on duplicate subpattern numbers</a>
2233     in the
2234     <a href="pcrepattern.html"><b>pcrepattern</b></a>
2235     page, you cannot use names to distinguish the different subpatterns, because
2236     names are not included in the compiled code. The matching process uses only
2237     numbers. For this reason, the use of different names for subpatterns of the
2238     same number causes an error at compile time.
2239 ph10 392 </P>
2240 ph10 691 <br><a name="SEC19" href="#TOC1">DUPLICATE SUBPATTERN NAMES</a><br>
2241 nigel 63 <P>
2242 nigel 91 <b>int pcre_get_stringtable_entries(const pcre *<i>code</i>,</b>
2243     <b>const char *<i>name</i>, char **<i>first</i>, char **<i>last</i>);</b>
2244     </P>
2245     <P>
2246     When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2247 ph10 461 are not required to be unique. (Duplicate names are always allowed for
2248     subpatterns with the same number, created by using the (?| feature. Indeed, if
2249     such subpatterns are named, they are required to use the same names.)
2250     </P>
2251     <P>
2252     Normally, patterns with duplicate names are such that in any one match, only
2253     one of the named subpatterns participates. An example is shown in the
2254 nigel 91 <a href="pcrepattern.html"><b>pcrepattern</b></a>
2255 ph10 208 documentation.
2256     </P>
2257     <P>
2258     When duplicates are present, <b>pcre_copy_named_substring()</b> and
2259     <b>pcre_get_named_substring()</b> return the first substring corresponding to
2260     the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2261     returned; no data is returned. The <b>pcre_get_stringnumber()</b> function
2262     returns one of the numbers that are associated with the name, but it is not
2263     defined which it is.
2264     </P>
2265     <P>
2266 nigel 91 If you want to get full details of all captured substrings for a given name,
2267     you must use the <b>pcre_get_stringtable_entries()</b> function. The first
2268     argument is the compiled pattern, and the second is the name. The third and
2269     fourth are pointers to variables which are updated by the function. After it
2270     has run, they point to the first and last entries in the name-to-number table
2271     for the given name. The function itself returns the length of each entry, or
2272 nigel 93 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2273 ph10 654 described above in the section entitled <i>Information about a pattern</i>
2274     <a href="#infoaboutpattern">above.</a>
2275 nigel 93 Given all the relevant entries for the name, you can extract each of their
2276     numbers, and hence the captured data, if any.
2277 nigel 91 </P>
2278 ph10 691 <br><a name="SEC20" href="#TOC1">FINDING ALL POSSIBLE MATCHES</a><br>
2279 nigel 91 <P>
2280 nigel 77 The traditional matching function uses a similar algorithm to Perl, which stops
2281     when it finds the first match, starting at a given point in the subject. If you
2282     want to find all possible matches, or the longest possible match, consider
2283     using the alternative matching function (see below) instead. If you cannot use
2284     the alternative function, but still need to find all possible matches, you
2285     can kludge it up by making use of the callout facility, which is described in
2286     the
2287     <a href="pcrecallout.html"><b>pcrecallout</b></a>
2288     documentation.
2289     </P>
2290     <P>
2291     What you have to do is to insert a callout right at the end of the pattern.
2292     When your callout function is called, extract and save the current matched
2293     substring. Then return 1, which forces <b>pcre_exec()</b> to backtrack and try
2294     other alternatives. Ultimately, when it runs out of matches, <b>pcre_exec()</b>
2295     will yield PCRE_ERROR_NOMATCH.
2296     <a name="dfamatch"></a></P>
2297 ph10 691 <br><a name="SEC21" href="#TOC1">MATCHING A PATTERN: THE ALTERNATIVE FUNCTION</a><br>
2298 nigel 77 <P>
2299     <b>int pcre_dfa_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
2300     <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
2301     <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>,</b>
2302     <b>int *<i>workspace</i>, int <i>wscount</i>);</b>
2303     </P>
2304     <P>
2305     The function <b>pcre_dfa_exec()</b> is called to match a subject string against
2306 nigel 93 a compiled pattern, using a matching algorithm that scans the subject string
2307     just once, and does not backtrack. This has different characteristics to the
2308     normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2309     patterns are not supported. Nevertheless, there are times when this kind of
2310 ph10 461 matching can be useful. For a discussion of the two matching algorithms, and a
2311 ph10 453 list of features that <b>pcre_dfa_exec()</b> does not support, see the
2312 nigel 77 <a href="pcrematching.html"><b>pcrematching</b></a>
2313     documentation.
2314     </P>
2315     <P>
2316     The arguments for the <b>pcre_dfa_exec()</b> function are the same as for
2317     <b>pcre_exec()</b>, plus two extras. The <i>ovector</i> argument is used in a
2318     different way, and this is described below. The other common arguments are used
2319     in the same way as for <b>pcre_exec()</b>, so their description is not repeated
2320     here.
2321     </P>
2322     <P>
2323     The two additional arguments provide workspace for the function. The workspace
2324     vector should contain at least 20 elements. It is used for keeping track of
2325     multiple paths through the pattern tree. More workspace will be needed for
2326 nigel 91 patterns and subjects where there are a lot of potential matches.
2327 nigel 77 </P>
2328     <P>
2329 nigel 87 Here is an example of a simple call to <b>pcre_dfa_exec()</b>:
2330 nigel 77 <pre>
2331     int rc;
2332     int ovector[10];
2333     int wspace[20];
2334 nigel 87 rc = pcre_dfa_exec(
2335 nigel 77 re, /* result of pcre_compile() */
2336     NULL, /* we didn't study the pattern */
2337     "some string", /* the subject string */
2338     11, /* the length of the subject string */
2339     0, /* start at offset 0 in the subject */
2340     0, /* default options */
2341     ovector, /* vector of integers for substring information */
2342     10, /* number of elements (NOT size in bytes) */
2343     wspace, /* working space vector */
2344     20); /* number of elements (NOT size in bytes) */
2345     </PRE>
2346     </P>
2347     <br><b>
2348     Option bits for <b>pcre_dfa_exec()</b>
2349     </b><br>
2350     <P>
2351     The unused bits of the <i>options</i> argument for <b>pcre_dfa_exec()</b> must be
2352 nigel 91 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
2353 ph10 453 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2354 ph10 545 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2355     PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2356     All but the last four of these are exactly the same as for <b>pcre_exec()</b>,
2357     so their description is not repeated here.
2358 nigel 77 <pre>
2359 ph10 429 PCRE_PARTIAL_HARD
2360 ph10 461 PCRE_PARTIAL_SOFT
2361 nigel 77 </pre>
2362 ph10 429 These have the same general effect as they do for <b>pcre_exec()</b>, but the
2363     details are slightly different. When PCRE_PARTIAL_HARD is set for
2364     <b>pcre_dfa_exec()</b>, it returns PCRE_ERROR_PARTIAL if the end of the subject
2365     is reached and there is still at least one matching possibility that requires
2366     additional characters. This happens even if some complete matches have also
2367     been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2368     is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2369     there have been no complete matches, but there is still at least one matching
2370 ph10 453 possibility. The portion of the string that was inspected when the longest
2371     partial match was found is set as the first matching string in both cases.
2372 ph10 567 There is a more detailed discussion of partial and multi-segment matching, with
2373     examples, in the
2374     <a href="pcrepartial.html"><b>pcrepartial</b></a>
2375     documentation.
2376 nigel 77 <pre>
2377     PCRE_DFA_SHORTEST
2378     </pre>
2379     Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2380 nigel 93 soon as it has found one match. Because of the way the alternative algorithm
2381     works, this is necessarily the shortest possible match at the first possible
2382     matching point in the subject string.
2383 nigel 77 <pre>
2384     PCRE_DFA_RESTART
2385     </pre>
2386 ph10 429 When <b>pcre_dfa_exec()</b> returns a partial match, it is possible to call it
2387     again, with additional subject characters, and have it continue with the same
2388     match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2389     <i>workspace</i> and <i>wscount</i> options must reference the same vector as
2390     before because data about the match so far is left in them after a partial
2391     match. There is more discussion of this facility in the
2392 nigel 77 <a href="pcrepartial.html"><b>pcrepartial</b></a>
2393     documentation.
2394     </P>
2395     <br><b>
2396     Successful returns from <b>pcre_dfa_exec()</b>
2397     </b><br>
2398     <P>
2399     When <b>pcre_dfa_exec()</b> succeeds, it may have matched more than one
2400     substring in the subject. Note, however, that all the matches from one run of
2401     the function start at the same point in the subject. The shorter matches are
2402     all initial substrings of the longer matches. For example, if the pattern
2403     <pre>
2404     &#60;.*&#62;
2405     </pre>
2406     is matched against the string
2407     <pre>
2408     This is &#60;something&#62; &#60;something else&#62; &#60;something further&#62; no more
2409     </pre>
2410     the three matched strings are
2411     <pre>
2412     &#60;something&#62;
2413     &#60;something&#62; &#60;something else&#62;
2414     &#60;something&#62; &#60;something else&#62; &#60;something further&#62;
2415     </pre>
2416     On success, the yield of the function is a number greater than zero, which is
2417     the number of matched substrings. The substrings themselves are returned in
2418     <i>ovector</i>. Each string uses two elements; the first is the offset to the
2419 nigel 93 start, and the second is the offset to the end. In fact, all the strings have
2420     the same start offset. (Space could have been saved by giving this only once,
2421     but it was decided to retain some compatibility with the way <b>pcre_exec()</b>
2422     returns data, even though the meaning of the strings is different.)
2423 nigel 77 </P>
2424     <P>
2425     The strings are returned in reverse order of length; that is, the longest
2426     matching string is given first. If there were too many matches to fit into
2427     <i>ovector</i>, the yield of the function is zero, and the vector is filled with
2428 ph10 691 the longest matches. Unlike <b>pcre_exec()</b>, <b>pcre_dfa_exec()</b> can use
2429     the entire <i>ovector</i> for returning matched strings.
2430 nigel 77 </P>
2431     <br><b>
2432     Error returns from <b>pcre_dfa_exec()</b>
2433     </b><br>
2434     <P>
2435     The <b>pcre_dfa_exec()</b> function returns a negative number when it fails.
2436     Many of the errors are the same as for <b>pcre_exec()</b>, and these are
2437     described
2438     <a href="#errorlist">above.</a>
2439     There are in addition the following errors that are specific to
2440     <b>pcre_dfa_exec()</b>:
2441     <pre>
2442     PCRE_ERROR_DFA_UITEM (-16)
2443     </pre>
2444     This return is given if <b>pcre_dfa_exec()</b> encounters an item in the pattern
2445     that it does not support, for instance, the use of \C or a back reference.
2446     <pre>
2447     PCRE_ERROR_DFA_UCOND (-17)
2448     </pre>
2449 nigel 93 This return is given if <b>pcre_dfa_exec()</b> encounters a condition item that
2450     uses a back reference for the condition, or a test for recursion in a specific
2451     group. These are not supported.
2452 nigel 77 <pre>
2453     PCRE_ERROR_DFA_UMLIMIT (-18)
2454     </pre>
2455     This return is given if <b>pcre_dfa_exec()</b> is called with an <i>extra</i>
2456 ph10 691 block that contains a setting of the <i>match_limit</i> or
2457     <i>match_limit_recursion</i> fields. This is not supported (these fields are
2458     meaningless for DFA matching).
2459 nigel 77 <pre>
2460     PCRE_ERROR_DFA_WSSIZE (-19)
2461     </pre>
2462     This return is given if <b>pcre_dfa_exec()</b> runs out of space in the
2463     <i>workspace</i> vector.
2464     <pre>
2465     PCRE_ERROR_DFA_RECURSE (-20)
2466     </pre>
2467     When a recursive subpattern is processed, the matching function calls itself
2468     recursively, using private vectors for <i>ovector</i> and <i>workspace</i>. This
2469     error is given if the output vector is not large enough. This should be
2470     extremely rare, as a vector of size 1000 is used.
2471     </P>
2472 ph10 691 <br><a name="SEC22" href="#TOC1">SEE ALSO</a><br>
2473 nigel 77 <P>
2474 nigel 93 <b>pcrebuild</b>(3), <b>pcrecallout</b>(3), <b>pcrecpp(3)</b>(3),
2475     <b>pcrematching</b>(3), <b>pcrepartial</b>(3), <b>pcreposix</b>(3),
2476     <b>pcreprecompile</b>(3), <b>pcresample</b>(3), <b>pcrestack</b>(3).
2477     </P>
2478 ph10 691 <br><a name="SEC23" href="#TOC1">AUTHOR</a><br>
2479 nigel 93 <P>
2480 ph10 99 Philip Hazel
2481 nigel 63 <br>
2482 ph10 99 University Computing Service
2483     <br>
2484     Cambridge CB2 3QH, England.
2485     <br>
2486     </P>
2487 ph10 691 <br><a name="SEC24" href="#TOC1">REVISION</a><br>
2488 ph10 99 <P>
2489 ph10 835 Last updated: 23 September 2011
2490 ph10 99 <br>
2491 ph10 654 Copyright &copy; 1997-2011 University of Cambridge.
2492 ph10 99 <br>
2493 nigel 75 <p>
2494     Return to the <a href="index.html">PCRE index page</a>.
2495     </p>

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12