ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log

Revision 548 - (hide annotations) (download) (as text)
Fri Jun 25 14:42:00 2010 UTC (4 years, 10 months ago) by ph10
File MIME type: text/html
File size: 99618 byte(s)
Final updates for 8.10 release.

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 nigel 75 <li><a name="TOC1" href="#SEC1">PCRE NATIVE API</a>
17     <li><a name="TOC2" href="#SEC2">PCRE API OVERVIEW</a>
18 nigel 91 <li><a name="TOC3" href="#SEC3">NEWLINES</a>
19     <li><a name="TOC4" href="#SEC4">MULTITHREADING</a>
20     <li><a name="TOC5" href="#SEC5">SAVING PRECOMPILED PATTERNS FOR LATER USE</a>
21     <li><a name="TOC6" href="#SEC6">CHECKING BUILD-TIME OPTIONS</a>
22     <li><a name="TOC7" href="#SEC7">COMPILING A PATTERN</a>
23     <li><a name="TOC8" href="#SEC8">COMPILATION ERROR CODES</a>
24     <li><a name="TOC9" href="#SEC9">STUDYING A PATTERN</a>
25     <li><a name="TOC10" href="#SEC10">LOCALE SUPPORT</a>
26     <li><a name="TOC11" href="#SEC11">INFORMATION ABOUT A PATTERN</a>
27     <li><a name="TOC12" href="#SEC12">OBSOLETE INFO FUNCTION</a>
28     <li><a name="TOC13" href="#SEC13">REFERENCE COUNTS</a>
29     <li><a name="TOC14" href="#SEC14">MATCHING A PATTERN: THE TRADITIONAL FUNCTION</a>
30     <li><a name="TOC15" href="#SEC15">EXTRACTING CAPTURED SUBSTRINGS BY NUMBER</a>
31     <li><a name="TOC16" href="#SEC16">EXTRACTING CAPTURED SUBSTRINGS BY NAME</a>
32     <li><a name="TOC17" href="#SEC17">DUPLICATE SUBPATTERN NAMES</a>
33     <li><a name="TOC18" href="#SEC18">FINDING ALL POSSIBLE MATCHES</a>
34     <li><a name="TOC19" href="#SEC19">MATCHING A PATTERN: THE ALTERNATIVE FUNCTION</a>
35 nigel 93 <li><a name="TOC20" href="#SEC20">SEE ALSO</a>
36 ph10 99 <li><a name="TOC21" href="#SEC21">AUTHOR</a>
37     <li><a name="TOC22" href="#SEC22">REVISION</a>
38 nigel 63 </ul>
39 nigel 75 <br><a name="SEC1" href="#TOC1">PCRE NATIVE API</a><br>
40 nigel 63 <P>
41     <b>#include &#60;pcre.h&#62;</b>
42     </P>
43     <P>
44     <b>pcre *pcre_compile(const char *<i>pattern</i>, int <i>options</i>,</b>
45     <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
46     <b>const unsigned char *<i>tableptr</i>);</b>
47     </P>
48     <P>
49 nigel 77 <b>pcre *pcre_compile2(const char *<i>pattern</i>, int <i>options</i>,</b>
50     <b>int *<i>errorcodeptr</i>,</b>
51     <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
52     <b>const unsigned char *<i>tableptr</i>);</b>
53     </P>
54     <P>
55 nigel 63 <b>pcre_extra *pcre_study(const pcre *<i>code</i>, int <i>options</i>,</b>
56     <b>const char **<i>errptr</i>);</b>
57     </P>
58     <P>
59     <b>int pcre_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
60     <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
61     <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>);</b>
62     </P>
63     <P>
64 nigel 77 <b>int pcre_dfa_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     <b>int *<i>workspace</i>, int <i>wscount</i>);</b>
68     </P>
69     <P>
70 nigel 63 <b>int pcre_copy_named_substring(const pcre *<i>code</i>,</b>
71     <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
72     <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
73     <b>char *<i>buffer</i>, int <i>buffersize</i>);</b>
74     </P>
75     <P>
76     <b>int pcre_copy_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
77     <b>int <i>stringcount</i>, int <i>stringnumber</i>, char *<i>buffer</i>,</b>
78     <b>int <i>buffersize</i>);</b>
79     </P>
80     <P>
81     <b>int pcre_get_named_substring(const pcre *<i>code</i>,</b>
82     <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
83     <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
84     <b>const char **<i>stringptr</i>);</b>
85     </P>
86     <P>
87     <b>int pcre_get_stringnumber(const pcre *<i>code</i>,</b>
88     <b>const char *<i>name</i>);</b>
89     </P>
90     <P>
91 nigel 91 <b>int pcre_get_stringtable_entries(const pcre *<i>code</i>,</b>
92     <b>const char *<i>name</i>, char **<i>first</i>, char **<i>last</i>);</b>
93     </P>
94     <P>
95 nigel 63 <b>int pcre_get_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
96     <b>int <i>stringcount</i>, int <i>stringnumber</i>,</b>
97     <b>const char **<i>stringptr</i>);</b>
98     </P>
99     <P>
100     <b>int pcre_get_substring_list(const char *<i>subject</i>,</b>
101     <b>int *<i>ovector</i>, int <i>stringcount</i>, const char ***<i>listptr</i>);</b>
102     </P>
103     <P>
104     <b>void pcre_free_substring(const char *<i>stringptr</i>);</b>
105     </P>
106     <P>
107     <b>void pcre_free_substring_list(const char **<i>stringptr</i>);</b>
108     </P>
109     <P>
110     <b>const unsigned char *pcre_maketables(void);</b>
111     </P>
112     <P>
113     <b>int pcre_fullinfo(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
114     <b>int <i>what</i>, void *<i>where</i>);</b>
115     </P>
116     <P>
117     <b>int pcre_info(const pcre *<i>code</i>, int *<i>optptr</i>, int</b>
118     <b>*<i>firstcharptr</i>);</b>
119     </P>
120     <P>
121 nigel 77 <b>int pcre_refcount(pcre *<i>code</i>, int <i>adjust</i>);</b>
122     </P>
123     <P>
124 nigel 63 <b>int pcre_config(int <i>what</i>, void *<i>where</i>);</b>
125     </P>
126     <P>
127     <b>char *pcre_version(void);</b>
128     </P>
129     <P>
130     <b>void *(*pcre_malloc)(size_t);</b>
131     </P>
132     <P>
133     <b>void (*pcre_free)(void *);</b>
134     </P>
135     <P>
136 nigel 73 <b>void *(*pcre_stack_malloc)(size_t);</b>
137     </P>
138     <P>
139     <b>void (*pcre_stack_free)(void *);</b>
140     </P>
141     <P>
142 nigel 63 <b>int (*pcre_callout)(pcre_callout_block *);</b>
143     </P>
144 nigel 75 <br><a name="SEC2" href="#TOC1">PCRE API OVERVIEW</a><br>
145 nigel 63 <P>
146 nigel 93 PCRE has its own native API, which is described in this document. There are
147     also some wrapper functions that correspond to the POSIX regular expression
148 nigel 77 API. These are described in the
149 nigel 75 <a href="pcreposix.html"><b>pcreposix</b></a>
150 nigel 77 documentation. Both of these APIs define a set of C function calls. A C++
151     wrapper is distributed with PCRE. It is documented in the
152     <a href="pcrecpp.html"><b>pcrecpp</b></a>
153     page.
154 nigel 63 </P>
155     <P>
156 nigel 77 The native API C function prototypes are defined in the header file
157     <b>pcre.h</b>, and on Unix systems the library itself is called <b>libpcre</b>.
158     It can normally be accessed by adding <b>-lpcre</b> to the command for linking
159     an application that uses PCRE. The header file defines the macros PCRE_MAJOR
160     and PCRE_MINOR to contain the major and minor release numbers for the library.
161 nigel 75 Applications can use these to include support for different releases of PCRE.
162 nigel 63 </P>
163     <P>
164 ph10 535 In a Windows environment, if you want to statically link an application program
165     against a non-dll <b>pcre.a</b> file, you must define PCRE_STATIC before
166     including <b>pcre.h</b> or <b>pcrecpp.h</b>, because otherwise the
167     <b>pcre_malloc()</b> and <b>pcre_free()</b> exported functions will be declared
168     <b>__declspec(dllimport)</b>, with unwanted results.
169     </P>
170     <P>
171 nigel 77 The functions <b>pcre_compile()</b>, <b>pcre_compile2()</b>, <b>pcre_study()</b>,
172     and <b>pcre_exec()</b> are used for compiling and matching regular expressions
173     in a Perl-compatible manner. A sample program that demonstrates the simplest
174 ph10 429 way of using them is provided in the file called <i>pcredemo.c</i> in the PCRE
175     source distribution. A listing of this program is given in the
176     <a href="pcredemo.html"><b>pcredemo</b></a>
177     documentation, and the
178 nigel 75 <a href="pcresample.html"><b>pcresample</b></a>
179 ph10 313 documentation describes how to compile and run it.
180 nigel 63 </P>
181     <P>
182 nigel 77 A second matching function, <b>pcre_dfa_exec()</b>, which is not
183     Perl-compatible, is also provided. This uses a different algorithm for the
184 nigel 91 matching. The alternative algorithm finds all possible matches (at a given
185 ph10 453 point in the subject), and scans the subject just once (unless there are
186     lookbehind assertions). However, this algorithm does not return captured
187     substrings. A description of the two matching algorithms and their advantages
188     and disadvantages is given in the
189 nigel 77 <a href="pcrematching.html"><b>pcrematching</b></a>
190     documentation.
191     </P>
192     <P>
193 nigel 75 In addition to the main compiling and matching functions, there are convenience
194 nigel 77 functions for extracting captured substrings from a subject string that is
195     matched by <b>pcre_exec()</b>. They are:
196 nigel 63 <pre>
197     <b>pcre_copy_substring()</b>
198     <b>pcre_copy_named_substring()</b>
199     <b>pcre_get_substring()</b>
200     <b>pcre_get_named_substring()</b>
201     <b>pcre_get_substring_list()</b>
202 nigel 75 <b>pcre_get_stringnumber()</b>
203 nigel 91 <b>pcre_get_stringtable_entries()</b>
204 nigel 75 </pre>
205 nigel 63 <b>pcre_free_substring()</b> and <b>pcre_free_substring_list()</b> are also
206     provided, to free the memory used for extracted strings.
207     </P>
208     <P>
209 nigel 75 The function <b>pcre_maketables()</b> is used to build a set of character tables
210 nigel 77 in the current locale for passing to <b>pcre_compile()</b>, <b>pcre_exec()</b>,
211     or <b>pcre_dfa_exec()</b>. This is an optional facility that is provided for
212     specialist use. Most commonly, no special tables are passed, in which case
213     internal tables that are generated when PCRE is built are used.
214 nigel 63 </P>
215     <P>
216     The function <b>pcre_fullinfo()</b> is used to find out information about a
217 nigel 75 compiled pattern; <b>pcre_info()</b> is an obsolete version that returns only
218 nigel 63 some of the available information, but is retained for backwards compatibility.
219     The function <b>pcre_version()</b> returns a pointer to a string containing the
220     version of PCRE and its date of release.
221     </P>
222     <P>
223 nigel 77 The function <b>pcre_refcount()</b> maintains a reference count in a data block
224     containing a compiled pattern. This is provided for the benefit of
225     object-oriented applications.
226     </P>
227     <P>
228 nigel 63 The global variables <b>pcre_malloc</b> and <b>pcre_free</b> initially contain
229 nigel 75 the entry points of the standard <b>malloc()</b> and <b>free()</b> functions,
230 nigel 63 respectively. PCRE calls the memory management functions via these variables,
231     so a calling program can replace them if it wishes to intercept the calls. This
232     should be done before calling any PCRE functions.
233     </P>
234     <P>
235 nigel 73 The global variables <b>pcre_stack_malloc</b> and <b>pcre_stack_free</b> are also
236     indirections to memory management functions. These special functions are used
237     only when PCRE is compiled to use the heap for remembering data, instead of
238 nigel 91 recursive function calls, when running the <b>pcre_exec()</b> function. See the
239     <a href="pcrebuild.html"><b>pcrebuild</b></a>
240     documentation for details of how to do this. It is a non-standard way of
241     building PCRE, for use in environments that have limited stacks. Because of the
242     greater use of memory management, it runs more slowly. Separate functions are
243     provided so that special-purpose external code can be used for this case. When
244     used, these functions are always called in a stack-like manner (last obtained,
245     first freed), and always for memory blocks of the same size. There is a
246     discussion about PCRE's stack usage in the
247     <a href="pcrestack.html"><b>pcrestack</b></a>
248     documentation.
249 nigel 73 </P>
250     <P>
251 nigel 63 The global variable <b>pcre_callout</b> initially contains NULL. It can be set
252     by the caller to a "callout" function, which PCRE will then call at specified
253 nigel 75 points during a matching operation. Details are given in the
254     <a href="pcrecallout.html"><b>pcrecallout</b></a>
255 nigel 63 documentation.
256 ph10 227 <a name="newlines"></a></P>
257 nigel 91 <br><a name="SEC3" href="#TOC1">NEWLINES</a><br>
258 nigel 63 <P>
259 ph10 150 PCRE supports five different conventions for indicating line breaks in
260 nigel 93 strings: a single CR (carriage return) character, a single LF (linefeed)
261 ph10 150 character, the two-character sequence CRLF, any of the three preceding, or any
262     Unicode newline sequence. The Unicode newline sequences are the three just
263     mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
264     U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
265     (paragraph separator, U+2029).
266 nigel 93 </P>
267     <P>
268     Each of the first three conventions is used by at least one operating system as
269     its standard newline sequence. When PCRE is built, a default can be specified.
270     The default default is LF, which is the Unix standard. When PCRE is run, the
271     default can be overridden, either when a pattern is compiled, or when it is
272     matched.
273     </P>
274     <P>
275 ph10 227 At compile time, the newline convention can be specified by the <i>options</i>
276     argument of <b>pcre_compile()</b>, or it can be specified by special text at the
277     start of the pattern itself; this overrides any other settings. See the
278     <a href="pcrepattern.html"><b>pcrepattern</b></a>
279     page for details of the special character sequences.
280     </P>
281     <P>
282 nigel 91 In the PCRE documentation the word "newline" is used to mean "the character or
283 nigel 93 pair of characters that indicate a line break". The choice of newline
284     convention affects the handling of the dot, circumflex, and dollar
285     metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
286     recognized line ending sequence, the match position advancement for a
287 ph10 227 non-anchored pattern. There is more detail about this in the
288     <a href="#execoptions">section on <b>pcre_exec()</b> options</a>
289 ph10 231 below.
290 nigel 91 </P>
291 ph10 231 <P>
292     The choice of newline convention does not affect the interpretation of
293     the \n or \r escape sequences, nor does it affect what \R matches, which is
294     controlled in a similar way, but by separate options.
295     </P>
296 nigel 91 <br><a name="SEC4" href="#TOC1">MULTITHREADING</a><br>
297     <P>
298 nigel 63 The PCRE functions can be used in multi-threading applications, with the
299 nigel 73 proviso that the memory management functions pointed to by <b>pcre_malloc</b>,
300     <b>pcre_free</b>, <b>pcre_stack_malloc</b>, and <b>pcre_stack_free</b>, and the
301     callout function pointed to by <b>pcre_callout</b>, are shared by all threads.
302 nigel 63 </P>
303     <P>
304     The compiled form of a regular expression is not altered during matching, so
305     the same compiled pattern can safely be used by several threads at once.
306     </P>
307 nigel 91 <br><a name="SEC5" href="#TOC1">SAVING PRECOMPILED PATTERNS FOR LATER USE</a><br>
308 nigel 63 <P>
309 nigel 75 The compiled form of a regular expression can be saved and re-used at a later
310     time, possibly by a different program, and even on a host other than the one on
311     which it was compiled. Details are given in the
312     <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
313 ph10 155 documentation. However, compiling a regular expression with one version of PCRE
314     for use with a different version is not guaranteed to work and may cause
315     crashes.
316 nigel 75 </P>
317 nigel 91 <br><a name="SEC6" href="#TOC1">CHECKING BUILD-TIME OPTIONS</a><br>
318 nigel 75 <P>
319 nigel 63 <b>int pcre_config(int <i>what</i>, void *<i>where</i>);</b>
320     </P>
321     <P>
322     The function <b>pcre_config()</b> makes it possible for a PCRE client to
323     discover which optional features have been compiled into the PCRE library. The
324     <a href="pcrebuild.html"><b>pcrebuild</b></a>
325     documentation has more details about these optional features.
326     </P>
327     <P>
328     The first argument for <b>pcre_config()</b> is an integer, specifying which
329     information is required; the second argument is a pointer to a variable into
330     which the information is placed. The following information is available:
331     <pre>
333 nigel 75 </pre>
334 nigel 63 The output is an integer that is set to one if UTF-8 support is available;
335     otherwise it is set to zero.
336     <pre>
338     </pre>
339     The output is an integer that is set to one if support for Unicode character
340     properties is available; otherwise it is set to zero.
341     <pre>
343 nigel 75 </pre>
344 nigel 91 The output is an integer whose value specifies the default character sequence
345 nigel 93 that is recognized as meaning "newline". The four values that are supported
346 ph10 392 are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF, and -1 for ANY.
347     Though they are derived from ASCII, the same values are returned in EBCDIC
348     environments. The default should normally correspond to the standard sequence
349     for your operating system.
350 nigel 63 <pre>
351 ph10 231 PCRE_CONFIG_BSR
352     </pre>
353     The output is an integer whose value indicates what character sequences the \R
354     escape sequence matches by default. A value of 0 means that \R matches any
355     Unicode line ending sequence; a value of 1 means that \R matches only CR, LF,
356     or CRLF. The default can be overridden when a pattern is compiled or matched.
357     <pre>
359 nigel 75 </pre>
360 nigel 63 The output is an integer that contains the number of bytes used for internal
361     linkage in compiled regular expressions. The value is 2, 3, or 4. Larger values
362     allow larger regular expressions to be compiled, at the expense of slower
363     matching. The default value of 2 is sufficient for all but the most massive
364     patterns, since it allows the compiled pattern to be up to 64K in size.
365     <pre>
367 nigel 75 </pre>
368 nigel 63 The output is an integer that contains the threshold above which the POSIX
369     interface uses <b>malloc()</b> for output vectors. Further details are given in
370 nigel 75 the
371     <a href="pcreposix.html"><b>pcreposix</b></a>
372     documentation.
373 nigel 63 <pre>
375 nigel 75 </pre>
376 ph10 392 The output is a long integer that gives the default limit for the number of
377 nigel 63 internal matching function calls in a <b>pcre_exec()</b> execution. Further
378     details are given with <b>pcre_exec()</b> below.
379 nigel 73 <pre>
381     </pre>
382 ph10 392 The output is a long integer that gives the default limit for the depth of
383 nigel 87 recursion when calling the internal matching function in a <b>pcre_exec()</b>
384     execution. Further details are given with <b>pcre_exec()</b> below.
385     <pre>
387 nigel 75 </pre>
388 nigel 77 The output is an integer that is set to one if internal recursion when running
389     <b>pcre_exec()</b> is implemented by recursive function calls that use the stack
390     to remember their state. This is the usual way that PCRE is compiled. The
391     output is zero if PCRE was compiled to use blocks of data on the heap instead
392     of recursive function calls. In this case, <b>pcre_stack_malloc</b> and
393     <b>pcre_stack_free</b> are called to manage memory blocks on the heap, thus
394     avoiding the use of the stack.
395 nigel 73 </P>
396 nigel 91 <br><a name="SEC7" href="#TOC1">COMPILING A PATTERN</a><br>
397 nigel 63 <P>
398     <b>pcre *pcre_compile(const char *<i>pattern</i>, int <i>options</i>,</b>
399     <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
400     <b>const unsigned char *<i>tableptr</i>);</b>
401 nigel 77 <b>pcre *pcre_compile2(const char *<i>pattern</i>, int <i>options</i>,</b>
402     <b>int *<i>errorcodeptr</i>,</b>
403     <b>const char **<i>errptr</i>, int *<i>erroffset</i>,</b>
404     <b>const unsigned char *<i>tableptr</i>);</b>
405 nigel 63 </P>
406     <P>
407 nigel 77 Either of the functions <b>pcre_compile()</b> or <b>pcre_compile2()</b> can be
408     called to compile a pattern into an internal form. The only difference between
409     the two interfaces is that <b>pcre_compile2()</b> has an additional argument,
410 ph10 461 <i>errorcodeptr</i>, via which a numerical error code can be returned. To avoid
411     too much repetition, we refer just to <b>pcre_compile()</b> below, but the
412     information applies equally to <b>pcre_compile2()</b>.
413 nigel 63 </P>
414     <P>
415 nigel 77 The pattern is a C string terminated by a binary zero, and is passed in the
416     <i>pattern</i> argument. A pointer to a single block of memory that is obtained
417     via <b>pcre_malloc</b> is returned. This contains the compiled code and related
418     data. The <b>pcre</b> type is defined for the returned block; this is a typedef
419     for a structure whose contents are not externally defined. It is up to the
420 nigel 91 caller to free the memory (via <b>pcre_free</b>) when it is no longer required.
421 nigel 77 </P>
422     <P>
423 nigel 63 Although the compiled code of a PCRE regex is relocatable, that is, it does not
424     depend on memory location, the complete <b>pcre</b> data block is not
425 nigel 75 fully relocatable, because it may contain a copy of the <i>tableptr</i>
426     argument, which is an address (see below).
427 nigel 63 </P>
428     <P>
429 nigel 93 The <i>options</i> argument contains various bit settings that affect the
430 nigel 75 compilation. It should be zero if no options are required. The available
431 ph10 416 options are described below. Some of them (in particular, those that are
432 ph10 461 compatible with Perl, but some others as well) can also be set and unset from
433 ph10 416 within the pattern (see the detailed description in the
434 nigel 75 <a href="pcrepattern.html"><b>pcrepattern</b></a>
435 ph10 416 documentation). For those options that can be different in different parts of
436 ph10 461 the pattern, the contents of the <i>options</i> argument specifies their
437     settings at the start of compilation and execution. The PCRE_ANCHORED,
438     PCRE_BSR_<i>xxx</i>, and PCRE_NEWLINE_<i>xxx</i> options can be set at the time
439     of matching as well as at compile time.
440 nigel 63 </P>
441     <P>
442     If <i>errptr</i> is NULL, <b>pcre_compile()</b> returns NULL immediately.
443     Otherwise, if compilation of a pattern fails, <b>pcre_compile()</b> returns
444     NULL, and sets the variable pointed to by <i>errptr</i> to point to a textual
445 nigel 87 error message. This is a static string that is part of the library. You must
446 ph10 454 not try to free it. The byte offset from the start of the pattern to the
447 ph10 461 character that was being processed when the error was discovered is placed in
448 ph10 454 the variable pointed to by <i>erroffset</i>, which must not be NULL. If it is,
449     an immediate error is given. Some errors are not detected until checks are
450     carried out when the whole pattern has been scanned; in this case the offset is
451     set to the end of the pattern.
452 nigel 63 </P>
453     <P>
454 nigel 77 If <b>pcre_compile2()</b> is used instead of <b>pcre_compile()</b>, and the
455     <i>errorcodeptr</i> argument is not NULL, a non-zero error code number is
456     returned via this argument in the event of an error. This is in addition to the
457     textual error message. Error codes and messages are listed below.
458     </P>
459     <P>
460 nigel 63 If the final argument, <i>tableptr</i>, is NULL, PCRE uses a default set of
461 nigel 75 character tables that are built when PCRE is compiled, using the default C
462     locale. Otherwise, <i>tableptr</i> must be an address that is the result of a
463     call to <b>pcre_maketables()</b>. This value is stored with the compiled
464     pattern, and used again by <b>pcre_exec()</b>, unless another table pointer is
465     passed to it. For more discussion, see the section on locale support below.
466 nigel 63 </P>
467     <P>
468     This code fragment shows a typical straightforward call to <b>pcre_compile()</b>:
469     <pre>
470     pcre *re;
471     const char *error;
472     int erroffset;
473     re = pcre_compile(
474     "^A.*Z", /* the pattern */
475     0, /* default options */
476     &error, /* for error message */
477     &erroffset, /* for error offset */
478     NULL); /* use default character tables */
479 nigel 75 </pre>
480     The following names for option bits are defined in the <b>pcre.h</b> header
481     file:
482 nigel 63 <pre>
484 nigel 75 </pre>
485 nigel 63 If this bit is set, the pattern is forced to be "anchored", that is, it is
486 nigel 75 constrained to match only at the first matching point in the string that is
487 nigel 63 being searched (the "subject string"). This effect can also be achieved by
488     appropriate constructs in the pattern itself, which is the only way to do it in
489     Perl.
490     <pre>
491 nigel 75 PCRE_AUTO_CALLOUT
492     </pre>
493     If this bit is set, <b>pcre_compile()</b> automatically inserts callout items,
494     all with number 255, before each pattern item. For discussion of the callout
495     facility, see the
496     <a href="pcrecallout.html"><b>pcrecallout</b></a>
497     documentation.
498     <pre>
499 ph10 231 PCRE_BSR_ANYCRLF
501     </pre>
502     These options (which are mutually exclusive) control what the \R escape
503     sequence matches. The choice is either to match only CR, LF, or CRLF, or to
504     match any Unicode newline sequence. The default is specified when PCRE is
505     built. It can be overridden from within the pattern, or by setting an option
506     when a compiled pattern is matched.
507     <pre>
508 nigel 63 PCRE_CASELESS
509 nigel 75 </pre>
510 nigel 63 If this bit is set, letters in the pattern match both upper and lower case
511     letters. It is equivalent to Perl's /i option, and it can be changed within a
512 nigel 77 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
513     concept of case for characters whose values are less than 128, so caseless
514     matching is always possible. For characters with higher values, the concept of
515     case is supported if PCRE is compiled with Unicode property support, but not
516     otherwise. If you want to use caseless matching for characters 128 and above,
517     you must ensure that PCRE is compiled with Unicode property support as well as
518     with UTF-8 support.
519 nigel 63 <pre>
521 nigel 75 </pre>
522 nigel 63 If this bit is set, a dollar metacharacter in the pattern matches only at the
523     end of the subject string. Without this option, a dollar also matches
524 nigel 91 immediately before a newline at the end of the string (but not before any other
525     newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
526     There is no equivalent to this option in Perl, and no way to set it within a
527     pattern.
528 nigel 63 <pre>
530 nigel 75 </pre>
531 nigel 63 If this bit is set, a dot metacharater in the pattern matches all characters,
532 nigel 91 including those that indicate newline. Without it, a dot does not match when
533     the current position is at a newline. This option is equivalent to Perl's /s
534     option, and it can be changed within a pattern by a (?s) option setting. A
535 nigel 93 negative class such as [^a] always matches newline characters, independent of
536     the setting of this option.
537 nigel 63 <pre>
538 nigel 91 PCRE_DUPNAMES
539     </pre>
540     If this bit is set, names used to identify capturing subpatterns need not be
541     unique. This can be helpful for certain types of pattern when it is known that
542     only one instance of the named subpattern can ever be matched. There are more
543     details of named subpatterns below; see also the
544     <a href="pcrepattern.html"><b>pcrepattern</b></a>
545     documentation.
546     <pre>
547 nigel 63 PCRE_EXTENDED
548 nigel 75 </pre>
549 nigel 63 If this bit is set, whitespace data characters in the pattern are totally
550     ignored except when escaped or inside a character class. Whitespace does not
551     include the VT character (code 11). In addition, characters between an
552 nigel 91 unescaped # outside a character class and the next newline, inclusive, are also
553     ignored. This is equivalent to Perl's /x option, and it can be changed within a
554     pattern by a (?x) option setting.
555 nigel 63 </P>
556     <P>
557     This option makes it possible to include comments inside complicated patterns.
558     Note, however, that this applies only to data characters. Whitespace characters
559     may never appear within special character sequences in a pattern, for example
560     within the sequence (?( which introduces a conditional subpattern.
561     <pre>
562     PCRE_EXTRA
563 nigel 75 </pre>
564 nigel 63 This option was invented in order to turn on additional functionality of PCRE
565     that is incompatible with Perl, but it is currently of very little use. When
566     set, any backslash in a pattern that is followed by a letter that has no
567     special meaning causes an error, thus reserving these combinations for future
568     expansion. By default, as in Perl, a backslash followed by a letter with no
569 nigel 91 special meaning is treated as a literal. (Perl can, however, be persuaded to
570 ph10 518 give an error for this, by running it with the -w option.) There are at present
571     no other features controlled by this option. It can also be set by a (?X)
572     option setting within a pattern.
573 nigel 63 <pre>
574 nigel 77 PCRE_FIRSTLINE
575     </pre>
576     If this option is set, an unanchored pattern is required to match before or at
577 nigel 91 the first newline in the subject string, though the matched text may continue
578     over the newline.
579 nigel 77 <pre>
581     </pre>
582     If this option is set, PCRE's behaviour is changed in some ways so that it is
583     compatible with JavaScript rather than Perl. The changes are as follows:
584     </P>
585     <P>
586     (1) A lone closing square bracket in a pattern causes a compile-time error,
587     because this is illegal in JavaScript (by default it is treated as a data
588     character). Thus, the pattern AB]CD becomes illegal when this option is set.
589     </P>
590     <P>
591     (2) At run time, a back reference to an unset subpattern group matches an empty
592     string (by default this causes the current matching alternative to fail). A
593     pattern such as (\1)(a) succeeds when this option is set (assuming it can find
594     an "a" in the subject), whereas it fails by default, for Perl compatibility.
595     <pre>
596 nigel 63 PCRE_MULTILINE
597 nigel 75 </pre>
598     By default, PCRE treats the subject string as consisting of a single line of
599     characters (even if it actually contains newlines). The "start of line"
600 nigel 63 metacharacter (^) matches only at the start of the string, while the "end of
601     line" metacharacter ($) matches only at the end of the string, or before a
602     terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
603     Perl.
604     </P>
605     <P>
606     When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
607 nigel 91 match immediately following or immediately before internal newlines in the
608     subject string, respectively, as well as at the very start and end. This is
609     equivalent to Perl's /m option, and it can be changed within a pattern by a
610     (?m) option setting. If there are no newlines in a subject string, or no
611 nigel 63 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
612     <pre>
613 nigel 91 PCRE_NEWLINE_CR
617 nigel 93 PCRE_NEWLINE_ANY
618 nigel 91 </pre>
619     These options override the default newline definition that was chosen when PCRE
620     was built. Setting the first or the second specifies that a newline is
621 nigel 93 indicated by a single character (CR or LF, respectively). Setting
622     PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
623 ph10 150 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
624     preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
625     that any Unicode newline sequence should be recognized. The Unicode newline
626     sequences are the three just mentioned, plus the single characters VT (vertical
627     tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
628     separator, U+2028), and PS (paragraph separator, U+2029). The last two are
629     recognized only in UTF-8 mode.
630 nigel 91 </P>
631     <P>
632 nigel 93 The newline setting in the options word uses three bits that are treated
633 ph10 150 as a number, giving eight possibilities. Currently only six are used (default
634     plus the five values above). This means that if you set more than one newline
635 nigel 93 option, the combination may or may not be sensible. For example,
636     PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
637 ph10 150 other combinations may yield unused numbers and cause an error.
638 nigel 93 </P>
639     <P>
640     The only time that a line break is specially recognized when compiling a
641     pattern is if PCRE_EXTENDED is set, and an unescaped # outside a character
642     class is encountered. This indicates a comment that lasts until after the next
643     line break sequence. In other circumstances, line break sequences are treated
644     as literal data, except that in PCRE_EXTENDED mode, both CR and LF are treated
645     as whitespace characters and are therefore ignored.
646     </P>
647     <P>
648     The newline option that is set at compile time becomes the default that is used
649     for <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, but it can be overridden.
650 nigel 91 <pre>
652 nigel 75 </pre>
653 nigel 63 If this option is set, it disables the use of numbered capturing parentheses in
654     the pattern. Any opening parenthesis that is not followed by ? behaves as if it
655     were followed by ?: but named parentheses can still be used for capturing (and
656     they acquire numbers in the usual way). There is no equivalent of this option
657     in Perl.
658     <pre>
659 ph10 535 PCRE_UCP
660     </pre>
661     This option changes the way PCRE processes \b, \d, \s, \w, and some of the
662     POSIX character classes. By default, only ASCII characters are recognized, but
663     if PCRE_UCP is set, Unicode properties are used instead to classify characters.
664     More details are given in the section on
665     <a href="pcre.html#genericchartypes">generic character types</a>
666     in the
667     <a href="pcrepattern.html"><b>pcrepattern</b></a>
668     page. If you set PCRE_UCP, matching one of the items it affects takes much
669     longer. The option is available only if PCRE has been compiled with Unicode
670     property support.
671     <pre>
672 nigel 63 PCRE_UNGREEDY
673 nigel 75 </pre>
674 nigel 63 This option inverts the "greediness" of the quantifiers so that they are not
675     greedy by default, but become greedy if followed by "?". It is not compatible
676     with Perl. It can also be set by a (?U) option setting within the pattern.
677     <pre>
678     PCRE_UTF8
679 nigel 75 </pre>
680 nigel 63 This option causes PCRE to regard both the pattern and the subject as strings
681     of UTF-8 characters instead of single-byte character strings. However, it is
682 nigel 75 available only when PCRE is built to include UTF-8 support. If not, the use
683 nigel 63 of this option provokes an error. Details of how this option changes the
684     behaviour of PCRE are given in the
685     <a href="pcre.html#utf8support">section on UTF-8 support</a>
686     in the main
687     <a href="pcre.html"><b>pcre</b></a>
688     page.
689 nigel 71 <pre>
691 nigel 75 </pre>
692 nigel 71 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
693 ph10 211 automatically checked. There is a discussion about the
694     <a href="pcre.html#utf8strings">validity of UTF-8 strings</a>
695     in the main
696     <a href="pcre.html"><b>pcre</b></a>
697     page. If an invalid UTF-8 sequence of bytes is found, <b>pcre_compile()</b>
698     returns an error. If you already know that your pattern is valid, and you want
699     to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
700     option. When it is set, the effect of passing an invalid UTF-8 string as a
701     pattern is undefined. It may cause your program to crash. Note that this option
702     can also be passed to <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, to suppress
703     the UTF-8 validity checking of subject strings.
704 nigel 71 </P>
705 nigel 91 <br><a name="SEC8" href="#TOC1">COMPILATION ERROR CODES</a><br>
706 nigel 63 <P>
707 nigel 77 The following table lists the error codes than may be returned by
708     <b>pcre_compile2()</b>, along with the error messages that may be returned by
709 nigel 93 both compiling functions. As PCRE has developed, some error codes have fallen
710     out of use. To avoid confusion, they have not been re-used.
711 nigel 77 <pre>
712     0 no error
713     1 \ at end of pattern
714     2 \c at end of pattern
715     3 unrecognized character follows \
716     4 numbers out of order in {} quantifier
717     5 number too big in {} quantifier
718     6 missing terminating ] for character class
719     7 invalid escape sequence in character class
720     8 range out of order in character class
721     9 nothing to repeat
722 nigel 93 10 [this code is not in use]
723 nigel 77 11 internal error: unexpected repeat
724 ph10 292 12 unrecognized character after (? or (?-
725 nigel 77 13 POSIX named classes are supported only within a class
726     14 missing )
727     15 reference to non-existent subpattern
728     16 erroffset passed as NULL
729     17 unknown option bit(s) set
730     18 missing ) after comment
731 nigel 93 19 [this code is not in use]
732 ph10 292 20 regular expression is too large
733 nigel 77 21 failed to get memory
734     22 unmatched parentheses
735     23 internal error: code overflow
736     24 unrecognized character after (?&#60;
737     25 lookbehind assertion is not fixed length
738 nigel 91 26 malformed number or name after (?(
739 nigel 77 27 conditional group contains more than two branches
740     28 assertion expected after (?(
741 ph10 182 29 (?R or (?[+-]digits must be followed by )
742 nigel 77 30 unknown POSIX class name
743     31 POSIX collating elements are not supported
744     32 this version of PCRE is not compiled with PCRE_UTF8 support
745 nigel 93 33 [this code is not in use]
746 nigel 77 34 character value in \x{...} sequence is too large
747     35 invalid condition (?(0)
748     36 \C not allowed in lookbehind assertion
749     37 PCRE does not support \L, \l, \N, \U, or \u
750     38 number after (?C is &#62; 255
751     39 closing ) for (?C expected
752     40 recursive call could loop indefinitely
753     41 unrecognized character after (?P
754 nigel 93 42 syntax error in subpattern name (missing terminator)
755 nigel 91 43 two named subpatterns have the same name
756 nigel 77 44 invalid UTF-8 string
757     45 support for \P, \p, and \X has not been compiled
758     46 malformed \P or \p sequence
759     47 unknown property name after \P or \p
760 nigel 91 48 subpattern name is too long (maximum 32 characters)
761 ph10 292 49 too many named subpatterns (maximum 10000)
762 ph10 202 50 [this code is not in use]
763 nigel 91 51 octal value is greater than \377 (not in UTF-8 mode)
764 nigel 93 52 internal error: overran compiling workspace
765 ph10 548 53 internal error: previously-checked referenced subpattern
766     not found
767 nigel 93 54 DEFINE group contains more than one branch
768     55 repeating a DEFINE group is not allowed
769 ph10 231 56 inconsistent NEWLINE options
770 ph10 345 57 \g is not followed by a braced, angle-bracketed, or quoted
771     name/number or by a plain number
772     58 a numbered reference must not be zero
773 ph10 512 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
774 ph10 292 60 (*VERB) not recognized
775     61 number is too big
776     62 subpattern name expected
777     63 digit expected after (?+
778 ph10 345 64 ] is an invalid data character in JavaScript compatibility mode
779 ph10 548 65 different names for subpatterns of the same number are
780     not allowed
781 ph10 512 66 (*MARK) must have an argument
782 ph10 535 67 this version of PCRE is not compiled with PCRE_UCP support
783 ph10 292 </pre>
784     The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
785     be used if the limits were changed when PCRE was built.
786 nigel 77 </P>
787 nigel 91 <br><a name="SEC9" href="#TOC1">STUDYING A PATTERN</a><br>
788 nigel 77 <P>
789     <b>pcre_extra *pcre_study(const pcre *<i>code</i>, int <i>options</i></b>
790 nigel 63 <b>const char **<i>errptr</i>);</b>
791     </P>
792     <P>
793 nigel 75 If a compiled pattern is going to be used several times, it is worth spending
794     more time analyzing it in order to speed up the time taken for matching. The
795 nigel 63 function <b>pcre_study()</b> takes a pointer to a compiled pattern as its first
796 nigel 75 argument. If studying the pattern produces additional information that will
797     help speed up matching, <b>pcre_study()</b> returns a pointer to a
798     <b>pcre_extra</b> block, in which the <i>study_data</i> field points to the
799     results of the study.
800 nigel 63 </P>
801     <P>
802 nigel 75 The returned value from <b>pcre_study()</b> can be passed directly to
803 ph10 461 <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b>. However, a <b>pcre_extra</b> block
804     also contains other fields that can be set by the caller before the block is
805     passed; these are described
806 nigel 75 <a href="#extradata">below</a>
807     in the section on matching a pattern.
808 nigel 63 </P>
809     <P>
810 ph10 461 If studying the pattern does not produce any useful information,
811 nigel 75 <b>pcre_study()</b> returns NULL. In that circumstance, if the calling program
812 ph10 461 wants to pass any of the other fields to <b>pcre_exec()</b> or
813     <b>pcre_dfa_exec()</b>, it must set up its own <b>pcre_extra</b> block.
814 nigel 63 </P>
815     <P>
816 nigel 75 The second argument of <b>pcre_study()</b> contains option bits. At present, no
817     options are defined, and this argument should always be zero.
818     </P>
819     <P>
820 nigel 63 The third argument for <b>pcre_study()</b> is a pointer for an error message. If
821     studying succeeds (even if no data is returned), the variable it points to is
822 nigel 87 set to NULL. Otherwise it is set to point to a textual error message. This is a
823     static string that is part of the library. You must not try to free it. You
824     should test the error pointer for NULL after calling <b>pcre_study()</b>, to be
825     sure that it has run successfully.
826 nigel 63 </P>
827     <P>
828     This is a typical call to <b>pcre_study</b>():
829     <pre>
830     pcre_extra *pe;
831     pe = pcre_study(
832     re, /* result of pcre_compile() */
833     0, /* no options exist */
834     &error); /* set to NULL or points to a message */
835 nigel 75 </pre>
836 ph10 461 Studying a pattern does two things: first, a lower bound for the length of
837     subject string that is needed to match the pattern is computed. This does not
838     mean that there are any strings of that length that match, but it does
839     guarantee that no shorter strings match. The value is used by
840     <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b> to avoid wasting time by trying to
841     match strings that are shorter than the lower bound. You can find out the value
842     in a calling program via the <b>pcre_fullinfo()</b> function.
843     </P>
844     <P>
845     Studying a pattern is also useful for non-anchored patterns that do not have a
846     single fixed starting character. A bitmap of possible starting bytes is
847     created. This speeds up finding a position in the subject at which to start
848     matching.
849 ph10 548 </P>
850     <P>
851     The two optimizations just described can be disabled by setting the
852     PCRE_NO_START_OPTIMIZE option when calling <b>pcre_exec()</b> or
853     <b>pcre_dfa_exec()</b>. You might want to do this if your pattern contains
854     callouts, or make use of (*MARK), and you make use of these in cases where
855     matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
856     <a href="#execoptions">below.</a>
857 nigel 75 <a name="localesupport"></a></P>
858 nigel 91 <br><a name="SEC10" href="#TOC1">LOCALE SUPPORT</a><br>
859 nigel 63 <P>
860 ph10 142 PCRE handles caseless matching, and determines whether characters are letters,
861 nigel 75 digits, or whatever, by reference to a set of tables, indexed by character
862 nigel 77 value. When running in UTF-8 mode, this applies only to characters with codes
863 ph10 535 less than 128. By default, higher-valued codes never match escapes such as \w
864     or \d, but they can be tested with \p if PCRE is built with Unicode character
865     property support. Alternatively, the PCRE_UCP option can be set at compile
866     time; this causes \w and friends to use Unicode property support instead of
867     built-in tables. The use of locales with Unicode is discouraged. If you are
868     handling characters with codes greater than 128, you should either use UTF-8
869     and Unicode, or use locales, but not try to mix the two.
870 nigel 63 </P>
871     <P>
872 ph10 142 PCRE contains an internal set of tables that are used when the final argument
873     of <b>pcre_compile()</b> is NULL. These are sufficient for many applications.
874     Normally, the internal tables recognize only ASCII characters. However, when
875     PCRE is built, it is possible to cause the internal tables to be rebuilt in the
876     default "C" locale of the local system, which may cause them to be different.
877 nigel 63 </P>
878     <P>
879 ph10 142 The internal tables can always be overridden by tables supplied by the
880     application that calls PCRE. These may be created in a different locale from
881     the default. As more and more applications change to using Unicode, the need
882     for this locale support is expected to die away.
883     </P>
884     <P>
885 nigel 75 External tables are built by calling the <b>pcre_maketables()</b> function,
886     which has no arguments, in the relevant locale. The result can then be passed
887     to <b>pcre_compile()</b> or <b>pcre_exec()</b> as often as necessary. For
888     example, to build and use tables that are appropriate for the French locale
889     (where accented characters with values greater than 128 are treated as letters),
890     the following code could be used:
891 nigel 63 <pre>
892 nigel 75 setlocale(LC_CTYPE, "fr_FR");
893 nigel 63 tables = pcre_maketables();
894     re = pcre_compile(..., tables);
895 nigel 75 </pre>
896 ph10 142 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
897     are using Windows, the name for the French locale is "french".
898     </P>
899     <P>
900 nigel 75 When <b>pcre_maketables()</b> runs, the tables are built in memory that is
901     obtained via <b>pcre_malloc</b>. It is the caller's responsibility to ensure
902     that the memory containing the tables remains available for as long as it is
903     needed.
904 nigel 63 </P>
905     <P>
906 nigel 75 The pointer that is passed to <b>pcre_compile()</b> is saved with the compiled
907 nigel 63 pattern, and the same tables are used via this pointer by <b>pcre_study()</b>
908 nigel 75 and normally also by <b>pcre_exec()</b>. Thus, by default, for any single
909     pattern, compilation, studying and matching all happen in the same locale, but
910     different patterns can be compiled in different locales.
911 nigel 63 </P>
912     <P>
913 nigel 75 It is possible to pass a table pointer or NULL (indicating the use of the
914     internal tables) to <b>pcre_exec()</b>. Although not intended for this purpose,
915     this facility could be used to match a pattern in a different locale from the
916     one in which it was compiled. Passing table pointers at run time is discussed
917     below in the section on matching a pattern.
918     </P>
919 nigel 91 <br><a name="SEC11" href="#TOC1">INFORMATION ABOUT A PATTERN</a><br>
920 nigel 75 <P>
921 nigel 63 <b>int pcre_fullinfo(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
922     <b>int <i>what</i>, void *<i>where</i>);</b>
923     </P>
924     <P>
925     The <b>pcre_fullinfo()</b> function returns information about a compiled
926     pattern. It replaces the obsolete <b>pcre_info()</b> function, which is
927     nevertheless retained for backwards compability (and is documented below).
928     </P>
929     <P>
930     The first argument for <b>pcre_fullinfo()</b> is a pointer to the compiled
931     pattern. The second argument is the result of <b>pcre_study()</b>, or NULL if
932     the pattern was not studied. The third argument specifies which piece of
933     information is required, and the fourth argument is a pointer to a variable
934     to receive the data. The yield of the function is zero for success, or one of
935     the following negative numbers:
936     <pre>
937     PCRE_ERROR_NULL the argument <i>code</i> was NULL
938     the argument <i>where</i> was NULL
939     PCRE_ERROR_BADMAGIC the "magic number" was not found
940     PCRE_ERROR_BADOPTION the value of <i>what</i> was invalid
941 nigel 75 </pre>
942     The "magic number" is placed at the start of each compiled pattern as an simple
943     check against passing an arbitrary memory pointer. Here is a typical call of
944     <b>pcre_fullinfo()</b>, to obtain the length of the compiled pattern:
945 nigel 63 <pre>
946     int rc;
947 nigel 91 size_t length;
948 nigel 63 rc = pcre_fullinfo(
949     re, /* result of pcre_compile() */
950     pe, /* result of pcre_study(), or NULL */
951     PCRE_INFO_SIZE, /* what is required */
952     &length); /* where to put the data */
953 nigel 75 </pre>
954 nigel 63 The possible values for the third argument are defined in <b>pcre.h</b>, and are
955     as follows:
956     <pre>
958 nigel 75 </pre>
959 nigel 63 Return the number of the highest back reference in the pattern. The fourth
960     argument should point to an <b>int</b> variable. Zero is returned if there are
961     no back references.
962     <pre>
964 nigel 75 </pre>
965 nigel 63 Return the number of capturing subpatterns in the pattern. The fourth argument
966 nigel 75 should point to an <b>int</b> variable.
967 nigel 63 <pre>
969 nigel 75 </pre>
970     Return a pointer to the internal default character tables within PCRE. The
971     fourth argument should point to an <b>unsigned char *</b> variable. This
972     information call is provided for internal use by the <b>pcre_study()</b>
973     function. External callers can cause PCRE to use its internal tables by passing
974     a NULL table pointer.
975     <pre>
977 nigel 75 </pre>
978 nigel 63 Return information about the first byte of any matched string, for a
979 nigel 91 non-anchored pattern. The fourth argument should point to an <b>int</b>
980     variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
981     still recognized for backwards compatibility.)
982 nigel 63 </P>
983     <P>
984 nigel 75 If there is a fixed first byte, for example, from a pattern such as
985 nigel 93 (cat|cow|coyote), its value is returned. Otherwise, if either
986 nigel 75 <br>
987     <br>
988 nigel 63 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
989     starts with "^", or
990 nigel 75 <br>
991     <br>
992 nigel 63 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
993     (if it were set, the pattern would be anchored),
994 nigel 75 <br>
995     <br>
996 nigel 63 -1 is returned, indicating that the pattern matches only at the start of a
997     subject string or after any newline within the string. Otherwise -2 is
998     returned. For anchored patterns, -2 is returned.
999     <pre>
1001 nigel 75 </pre>
1002 nigel 63 If the pattern was studied, and this resulted in the construction of a 256-bit
1003     table indicating a fixed set of bytes for the first byte in any matching
1004     string, a pointer to the table is returned. Otherwise NULL is returned. The
1005     fourth argument should point to an <b>unsigned char *</b> variable.
1006     <pre>
1008     </pre>
1009     Return 1 if the pattern contains any explicit matches for CR or LF characters,
1010 ph10 243 otherwise 0. The fourth argument should point to an <b>int</b> variable. An
1011     explicit match is either a literal CR or LF character, or \r or \n.
1012 ph10 227 <pre>
1013 ph10 172 PCRE_INFO_JCHANGED
1014     </pre>
1015 ph10 286 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1016     0. The fourth argument should point to an <b>int</b> variable. (?J) and
1017     (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1018 ph10 172 <pre>
1020 nigel 75 </pre>
1021 nigel 65 Return the value of the rightmost literal byte that must exist in any matched
1022     string, other than at its start, if such a byte has been recorded. The fourth
1023     argument should point to an <b>int</b> variable. If there is no such byte, -1 is
1024     returned. For anchored patterns, a last literal byte is recorded only if it
1025     follows something of variable length. For example, for the pattern
1026     /^a\d+z\d+/ the returned value is "z", but for /^a\dz\d/ the returned value
1027     is -1.
1028 nigel 63 <pre>
1030     </pre>
1031     If the pattern was studied and a minimum length for matching subject strings
1032     was computed, its value is returned. Otherwise the returned value is -1. The
1033     value is a number of characters, not bytes (this may be relevant in UTF-8
1034     mode). The fourth argument should point to an <b>int</b> variable. A
1035     non-negative value is a lower bound to the length of any matching string. There
1036     may not be any strings of that length that do actually match, but every string
1037     that does match is at least that long.
1038     <pre>
1042 nigel 75 </pre>
1043 nigel 63 PCRE supports the use of named as well as numbered capturing parentheses. The
1044     names are just an additional way of identifying the parentheses, which still
1045 nigel 91 acquire numbers. Several convenience functions such as
1046     <b>pcre_get_named_substring()</b> are provided for extracting captured
1047     substrings by name. It is also possible to extract the data directly, by first
1048     converting the name to a number in order to access the correct pointers in the
1049     output vector (described with <b>pcre_exec()</b> below). To do the conversion,
1050     you need to use the name-to-number map, which is described by these three
1051     values.
1052 nigel 63 </P>
1053     <P>
1054     The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1055     the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1056     entry; both of these return an <b>int</b> value. The entry size depends on the
1057     length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1058     entry of the table (a pointer to <b>char</b>). The first two bytes of each entry
1059     are the number of the capturing parenthesis, most significant byte first. The
1060 ph10 461 rest of the entry is the corresponding name, zero terminated.
1061     </P>
1062     <P>
1063     The names are in alphabetical order. Duplicate names may appear if (?| is used
1064     to create multiple groups with the same number, as described in the
1065     <a href="pcrepattern.html#dupsubpatternnumber">section on duplicate subpattern numbers</a>
1066     in the
1067     <a href="pcrepattern.html"><b>pcrepattern</b></a>
1068     page. Duplicate names for subpatterns with different numbers are permitted only
1069     if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1070     table in the order in which they were found in the pattern. In the absence of
1071     (?| this is the order of increasing number; when (?| is used this is not
1072     necessarily the case because later subpatterns may have lower numbers.
1073     </P>
1074     <P>
1075     As a simple example of the name/number table, consider the following pattern
1076     (assume PCRE_EXTENDED is set, so white space - including newlines - is
1077     ignored):
1078 nigel 63 <pre>
1079 nigel 93 (?&#60;date&#62; (?&#60;year&#62;(\d\d)?\d\d) - (?&#60;month&#62;\d\d) - (?&#60;day&#62;\d\d) )
1080 nigel 75 </pre>
1081 nigel 63 There are four named subpatterns, so the table has four entries, and each entry
1082     in the table is eight bytes long. The table is as follows, with non-printing
1083 nigel 75 bytes shows in hexadecimal, and undefined bytes shown as ??:
1084 nigel 63 <pre>
1085     00 01 d a t e 00 ??
1086     00 05 d a y 00 ?? ??
1087     00 04 m o n t h 00
1088     00 02 y e a r 00 ??
1089 nigel 75 </pre>
1090     When writing code to extract data from named subpatterns using the
1091 nigel 91 name-to-number map, remember that the length of the entries is likely to be
1092 nigel 75 different for each compiled pattern.
1093 nigel 63 <pre>
1095     </pre>
1096 ph10 453 Return 1 if the pattern can be used for partial matching with
1097     <b>pcre_exec()</b>, otherwise 0. The fourth argument should point to an
1098     <b>int</b> variable. From release 8.00, this always returns 1, because the
1099     restrictions that previously applied to partial matching have been lifted. The
1100 ph10 172 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1101 ph10 429 documentation gives details of partial matching.
1102 ph10 172 <pre>
1103 nigel 63 PCRE_INFO_OPTIONS
1104 nigel 75 </pre>
1105 nigel 63 Return a copy of the options with which the pattern was compiled. The fourth
1106     argument should point to an <b>unsigned long int</b> variable. These option bits
1107     are those specified in the call to <b>pcre_compile()</b>, modified by any
1108 ph10 197 top-level option settings at the start of the pattern itself. In other words,
1109     they are the options that will be in force when matching starts. For example,
1110     if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1112 nigel 63 </P>
1113     <P>
1114     A pattern is automatically anchored by PCRE if all of its top-level
1115     alternatives begin with one of the following:
1116     <pre>
1117     ^ unless PCRE_MULTILINE is set
1118     \A always
1119     \G always
1120 nigel 75 .* if PCRE_DOTALL is set and there are no back references to the subpattern in which .* appears
1121     </pre>
1122 nigel 63 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1123     <b>pcre_fullinfo()</b>.
1124     <pre>
1126 nigel 75 </pre>
1127 nigel 63 Return the size of the compiled pattern, that is, the value that was passed as
1128     the argument to <b>pcre_malloc()</b> when PCRE was getting memory in which to
1129     place the compiled data. The fourth argument should point to a <b>size_t</b>
1130     variable.
1131     <pre>
1133 nigel 75 </pre>
1134     Return the size of the data block pointed to by the <i>study_data</i> field in
1135 nigel 63 a <b>pcre_extra</b> block. That is, it is the value that was passed to
1136     <b>pcre_malloc()</b> when PCRE was getting memory into which to place the data
1137 ph10 461 created by <b>pcre_study()</b>. If <b>pcre_extra</b> is NULL, or there is no
1138     study data, zero is returned. The fourth argument should point to a
1139 nigel 63 <b>size_t</b> variable.
1140     </P>
1141 nigel 91 <br><a name="SEC12" href="#TOC1">OBSOLETE INFO FUNCTION</a><br>
1142 nigel 63 <P>
1143     <b>int pcre_info(const pcre *<i>code</i>, int *<i>optptr</i>, int</b>
1144     <b>*<i>firstcharptr</i>);</b>
1145     </P>
1146     <P>
1147     The <b>pcre_info()</b> function is now obsolete because its interface is too
1148     restrictive to return all the available data about a compiled pattern. New
1149     programs should use <b>pcre_fullinfo()</b> instead. The yield of
1150     <b>pcre_info()</b> is the number of capturing subpatterns, or one of the
1151     following negative numbers:
1152     <pre>
1153     PCRE_ERROR_NULL the argument <i>code</i> was NULL
1154     PCRE_ERROR_BADMAGIC the "magic number" was not found
1155 nigel 75 </pre>
1156 nigel 63 If the <i>optptr</i> argument is not NULL, a copy of the options with which the
1157     pattern was compiled is placed in the integer it points to (see
1158     PCRE_INFO_OPTIONS above).
1159     </P>
1160     <P>
1161     If the pattern is not anchored and the <i>firstcharptr</i> argument is not NULL,
1162     it is used to pass back information about the first character of any matched
1163     string (see PCRE_INFO_FIRSTBYTE above).
1164     </P>
1165 nigel 91 <br><a name="SEC13" href="#TOC1">REFERENCE COUNTS</a><br>
1166 nigel 63 <P>
1167 nigel 77 <b>int pcre_refcount(pcre *<i>code</i>, int <i>adjust</i>);</b>
1168     </P>
1169     <P>
1170     The <b>pcre_refcount()</b> function is used to maintain a reference count in the
1171     data block that contains a compiled pattern. It is provided for the benefit of
1172     applications that operate in an object-oriented manner, where different parts
1173     of the application may be using the same compiled pattern, but you want to free
1174     the block when they are all done.
1175     </P>
1176     <P>
1177     When a pattern is compiled, the reference count field is initialized to zero.
1178     It is changed only by calling this function, whose action is to add the
1179     <i>adjust</i> value (which may be positive or negative) to it. The yield of the
1180     function is the new value. However, the value of the count is constrained to
1181     lie between 0 and 65535, inclusive. If the new value is outside these limits,
1182     it is forced to the appropriate limit value.
1183     </P>
1184     <P>
1185     Except when it is zero, the reference count is not correctly preserved if a
1186     pattern is compiled on one host and then transferred to a host whose byte-order
1187     is different. (This seems a highly unlikely scenario.)
1188     </P>
1189 nigel 91 <br><a name="SEC14" href="#TOC1">MATCHING A PATTERN: THE TRADITIONAL FUNCTION</a><br>
1190 nigel 77 <P>
1191 nigel 63 <b>int pcre_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
1192     <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
1193     <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>);</b>
1194     </P>
1195     <P>
1196     The function <b>pcre_exec()</b> is called to match a subject string against a
1197 nigel 75 compiled pattern, which is passed in the <i>code</i> argument. If the
1198 ph10 461 pattern was studied, the result of the study should be passed in the
1199 nigel 77 <i>extra</i> argument. This function is the main matching facility of the
1200     library, and it operates in a Perl-like manner. For specialist use there is
1201     also an alternative matching function, which is described
1202     <a href="#dfamatch">below</a>
1203     in the section about the <b>pcre_dfa_exec()</b> function.
1204 nigel 63 </P>
1205     <P>
1206 nigel 75 In most applications, the pattern will have been compiled (and optionally
1207     studied) in the same process that calls <b>pcre_exec()</b>. However, it is
1208     possible to save compiled patterns and study data, and then use them later
1209     in different processes, possibly even on different hosts. For a discussion
1210     about this, see the
1211     <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1212     documentation.
1213 nigel 63 </P>
1214     <P>
1215 nigel 75 Here is an example of a simple call to <b>pcre_exec()</b>:
1216 nigel 63 <pre>
1217     int rc;
1218     int ovector[30];
1219     rc = pcre_exec(
1220     re, /* result of pcre_compile() */
1221     NULL, /* we didn't study the pattern */
1222     "some string", /* the subject string */
1223     11, /* the length of the subject string */
1224     0, /* start at offset 0 in the subject */
1225     0, /* default options */
1226 nigel 75 ovector, /* vector of integers for substring information */
1227 nigel 77 30); /* number of elements (NOT size in bytes) */
1228 nigel 75 <a name="extradata"></a></PRE>
1229 nigel 63 </P>
1230 nigel 75 <br><b>
1231     Extra data for <b>pcre_exec()</b>
1232     </b><br>
1233 nigel 63 <P>
1234     If the <i>extra</i> argument is not NULL, it must point to a <b>pcre_extra</b>
1235     data block. The <b>pcre_study()</b> function returns such a block (when it
1236     doesn't return NULL), but you can also create one for yourself, and pass
1237 nigel 87 additional information in it. The <b>pcre_extra</b> block contains the following
1238     fields (not necessarily in this order):
1239 nigel 63 <pre>
1240     unsigned long int <i>flags</i>;
1241     void *<i>study_data</i>;
1242     unsigned long int <i>match_limit</i>;
1243 nigel 87 unsigned long int <i>match_limit_recursion</i>;
1244 nigel 63 void *<i>callout_data</i>;
1245 nigel 75 const unsigned char *<i>tables</i>;
1246 ph10 512 unsigned char **<i>mark</i>;
1247 nigel 75 </pre>
1248 nigel 63 The <i>flags</i> field is a bitmap that specifies which of the other fields
1249     are set. The flag bits are:
1250     <pre>
1255 nigel 75 PCRE_EXTRA_TABLES
1256 ph10 512 PCRE_EXTRA_MARK
1257 nigel 75 </pre>
1258 nigel 63 Other flag bits should be set to zero. The <i>study_data</i> field is set in the
1259     <b>pcre_extra</b> block that is returned by <b>pcre_study()</b>, together with
1260 nigel 75 the appropriate flag bit. You should not set this yourself, but you may add to
1261     the block by setting the other fields and their corresponding flag bits.
1262 nigel 63 </P>
1263     <P>
1264     The <i>match_limit</i> field provides a means of preventing PCRE from using up a
1265     vast amount of resources when running patterns that are not going to match,
1266     but which have a very large number of possibilities in their search trees. The
1267 ph10 461 classic example is a pattern that uses nested unlimited repeats.
1268 nigel 63 </P>
1269     <P>
1270 nigel 75 Internally, PCRE uses a function called <b>match()</b> which it calls repeatedly
1271 nigel 87 (sometimes recursively). The limit set by <i>match_limit</i> is imposed on the
1272     number of times this function is called during a match, which has the effect of
1273     limiting the amount of backtracking that can take place. For patterns that are
1274     not anchored, the count restarts from zero for each position in the subject
1275     string.
1276 nigel 75 </P>
1277     <P>
1278 nigel 87 The default value for the limit can be set when PCRE is built; the default
1279 nigel 63 default is 10 million, which handles all but the most extreme cases. You can
1280 nigel 87 override the default by suppling <b>pcre_exec()</b> with a <b>pcre_extra</b>
1281     block in which <i>match_limit</i> is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1282     the <i>flags</i> field. If the limit is exceeded, <b>pcre_exec()</b> returns
1284 nigel 63 </P>
1285     <P>
1286 nigel 87 The <i>match_limit_recursion</i> field is similar to <i>match_limit</i>, but
1287     instead of limiting the total number of times that <b>match()</b> is called, it
1288     limits the depth of recursion. The recursion depth is a smaller number than the
1289     total number of calls, because not all calls to <b>match()</b> are recursive.
1290     This limit is of use only if it is set smaller than <i>match_limit</i>.
1291     </P>
1292     <P>
1293     Limiting the recursion depth limits the amount of stack that can be used, or,
1294     when PCRE has been compiled to use memory on the heap instead of the stack, the
1295     amount of heap memory that can be used.
1296     </P>
1297     <P>
1298     The default value for <i>match_limit_recursion</i> can be set when PCRE is
1299     built; the default default is the same value as the default for
1300     <i>match_limit</i>. You can override the default by suppling <b>pcre_exec()</b>
1301     with a <b>pcre_extra</b> block in which <i>match_limit_recursion</i> is set, and
1302     PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the <i>flags</i> field. If the limit
1303     is exceeded, <b>pcre_exec()</b> returns PCRE_ERROR_RECURSIONLIMIT.
1304     </P>
1305     <P>
1306 ph10 453 The <i>callout_data</i> field is used in conjunction with the "callout" feature,
1307     and is described in the
1308 nigel 75 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1309     documentation.
1310 nigel 63 </P>
1311     <P>
1312 nigel 75 The <i>tables</i> field is used to pass a character tables pointer to
1313     <b>pcre_exec()</b>; this overrides the value that is stored with the compiled
1314     pattern. A non-NULL value is stored with the compiled pattern only if custom
1315     tables were supplied to <b>pcre_compile()</b> via its <i>tableptr</i> argument.
1316     If NULL is passed to <b>pcre_exec()</b> using this mechanism, it forces PCRE's
1317     internal tables to be used. This facility is helpful when re-using patterns
1318     that have been saved after compiling with an external set of tables, because
1319     the external tables might be at a different address when <b>pcre_exec()</b> is
1320     called. See the
1321     <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1322     documentation for a discussion of saving compiled patterns for later use.
1323 ph10 512 </P>
1324     <P>
1325     If PCRE_EXTRA_MARK is set in the <i>flags</i> field, the <i>mark</i> field must
1326     be set to point to a <b>char *</b> variable. If the pattern contains any
1327     backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1328     a name to pass back, a pointer to the name string (zero terminated) is placed
1329     in the variable pointed to by the <i>mark</i> field. The names are within the
1330     compiled pattern; if you wish to retain such a name you must copy it before
1331     freeing the memory of a compiled pattern. If there is no name to pass back, the
1332     variable pointed to by the <i>mark</i> field set to NULL. For details of the
1333     backtracking control verbs, see the section entitled
1334     <a href="pcrepattern#backtrackcontrol">"Backtracking control"</a>
1335     in the
1336     <a href="pcrepattern.html"><b>pcrepattern</b></a>
1337     documentation.
1338 ph10 227 <a name="execoptions"></a></P>
1339 nigel 75 <br><b>
1340     Option bits for <b>pcre_exec()</b>
1341     </b><br>
1342 nigel 63 <P>
1343 nigel 75 The unused bits of the <i>options</i> argument for <b>pcre_exec()</b> must be
1344 nigel 91 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
1348 nigel 63 <pre>
1349 nigel 75 PCRE_ANCHORED
1350     </pre>
1351     The PCRE_ANCHORED option limits <b>pcre_exec()</b> to matching at the first
1352     matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1353     to be anchored by virtue of its contents, it cannot be made unachored at
1354     matching time.
1355     <pre>
1356 ph10 231 PCRE_BSR_ANYCRLF
1358     </pre>
1359     These options (which are mutually exclusive) control what the \R escape
1360     sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1361     match any Unicode newline sequence. These options override the choice that was
1362     made or defaulted when the pattern was compiled.
1363     <pre>
1364 nigel 91 PCRE_NEWLINE_CR
1368 nigel 93 PCRE_NEWLINE_ANY
1369 nigel 91 </pre>
1370     These options override the newline definition that was chosen or defaulted when
1371 nigel 93 the pattern was compiled. For details, see the description of
1372     <b>pcre_compile()</b> above. During matching, the newline choice affects the
1373     behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1374     the way the match position is advanced after a match failure for an unanchored
1375 ph10 227 pattern.
1376     </P>
1377     <P>
1379     match attempt for an unanchored pattern fails when the current position is at a
1380 ph10 231 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1381 ph10 227 characters, the match position is advanced by two characters instead of one, in
1382     other words, to after the CRLF.
1383     </P>
1384     <P>
1385     The above rule is a compromise that makes the most common cases work as
1386     expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1387     set), it does not match the string "\r\nA" because, after failing at the
1388     start, it skips both the CR and the LF before retrying. However, the pattern
1389     [\r\n]A does match that string, because it contains an explicit CR or LF
1390     reference, and so advances only by one character after the first failure.
1391     </P>
1392     <P>
1393 ph10 231 An explicit match for CR of LF is either a literal appearance of one of those
1394     characters, or one of the \r or \n escape sequences. Implicit matches such as
1395     [^X] do not count, nor does \s (which includes CR and LF in the characters
1396     that it matches).
1397     </P>
1398     <P>
1399 ph10 227 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1400     valid newline sequence and explicit \r or \n escapes appear in the pattern.
1401 nigel 91 <pre>
1402 nigel 63 PCRE_NOTBOL
1403 nigel 75 </pre>
1404     This option specifies that first character of the subject string is not the
1405     beginning of a line, so the circumflex metacharacter should not match before
1406     it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1407     never to match. This option affects only the behaviour of the circumflex
1408     metacharacter. It does not affect \A.
1409 nigel 63 <pre>
1410     PCRE_NOTEOL
1411 nigel 75 </pre>
1412     This option specifies that the end of the subject string is not the end of a
1413     line, so the dollar metacharacter should not match it nor (except in multiline
1414     mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1415     compile time) causes dollar never to match. This option affects only the
1416     behaviour of the dollar metacharacter. It does not affect \Z or \z.
1417 nigel 63 <pre>
1419 nigel 75 </pre>
1420 nigel 63 An empty string is not considered to be a valid match if this option is set. If
1421     there are alternatives in the pattern, they are tried. If all the alternatives
1422     match the empty string, the entire match fails. For example, if the pattern
1423     <pre>
1424     a?b?
1425 nigel 75 </pre>
1426 ph10 453 is applied to a string not beginning with "a" or "b", it matches an empty
1427 nigel 63 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1428     valid, so PCRE searches further into the string for occurrences of "a" or "b".
1429 ph10 453 <pre>
1431     </pre>
1432 ph10 461 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1433 ph10 453 the start of the subject is permitted. If the pattern is anchored, such a match
1434     can occur only if the pattern contains \K.
1435 nigel 63 </P>
1436     <P>
1437 ph10 453 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1438     does make a special case of a pattern match of the empty string within its
1439     <b>split()</b> function, and when using the /g modifier. It is possible to
1440     emulate Perl's behaviour after matching a null string by first trying the match
1441     again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1442     if that fails, by advancing the starting offset (see below) and trying an
1443     ordinary match again. There is some code that demonstrates how to do this in
1444     the
1445 ph10 429 <a href="pcredemo.html"><b>pcredemo</b></a>
1446     sample program.
1447 nigel 75 <pre>
1449     </pre>
1450     There are a number of optimizations that <b>pcre_exec()</b> uses at the start of
1451 ph10 545 a match, in order to speed up the process. For example, if it is known that an
1452     unanchored match must start with a specific character, it searches the subject
1453     for that character, and fails immediately if it cannot find it, without
1454     actually running the main matching function. This means that a special item
1455     such as (*COMMIT) at the start of a pattern is not considered until after a
1456 ph10 548 suitable starting point for the match has been found. When callouts or (*MARK)
1457     items are in use, these "start-up" optimizations can cause them to be skipped
1458     if the pattern is never actually used. The start-up optimizations are in effect
1459     a pre-scan of the subject that takes place before the pattern is run.
1460     </P>
1461     <P>
1462     The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1463     causing performance to suffer, but ensuring that in cases where the result is
1464     "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1465     are considered at every possible starting position in the subject string.
1466     Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1467     Consider the pattern
1468 ph10 392 <pre>
1469 ph10 548 (*COMMIT)ABC
1470     </pre>
1471     When this is compiled, PCRE records the fact that a match must start with the
1472     character "A". Suppose the subject string is "DEFABC". The start-up
1473     optimization scans along the subject, finds "A" and runs the first match
1474     attempt from there. The (*COMMIT) item means that the pattern must match the
1475     current starting position, which in this case, it does. However, if the same
1476     match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1477     subject string does not happen. The first match attempt is run starting from
1478     "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1479     the overall result is "no match". If the pattern is studied, more start-up
1480     optimizations may be used. For example, a minimum length for the subject may be
1481     recorded. Consider the pattern
1482     <pre>
1483     (*MARK:A)(X|Y)
1484     </pre>
1485     The minimum length for a match is one character. If the subject is "ABC", there
1486     will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1487     If the pattern is studied, the final attempt does not take place, because PCRE
1488     knows that the subject is too short, and so the (*MARK) is never encountered.
1489     In this case, studying the pattern does not affect the overall match result,
1490     which is still "no match", but it does affect the auxiliary information that is
1491     returned.
1492     <pre>
1493 nigel 75 PCRE_NO_UTF8_CHECK
1494     </pre>
1495     When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1496     string is automatically checked when <b>pcre_exec()</b> is subsequently called.
1497     The value of <i>startoffset</i> is also checked to ensure that it points to the
1498 ph10 211 start of a UTF-8 character. There is a discussion about the validity of UTF-8
1499     strings in the
1500     <a href="pcre.html#utf8strings">section on UTF-8 support</a>
1501     in the main
1502     <a href="pcre.html"><b>pcre</b></a>
1503     page. If an invalid UTF-8 sequence of bytes is found, <b>pcre_exec()</b> returns
1504     the error PCRE_ERROR_BADUTF8. If <i>startoffset</i> contains an invalid value,
1505     PCRE_ERROR_BADUTF8_OFFSET is returned.
1506 nigel 63 </P>
1507     <P>
1508 nigel 75 If you already know that your subject is valid, and you want to skip these
1509     checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1510     calling <b>pcre_exec()</b>. You might want to do this for the second and
1511     subsequent calls to <b>pcre_exec()</b> if you are making repeated calls to find
1512     all the matches in a single subject string. However, you should be sure that
1513     the value of <i>startoffset</i> points to the start of a UTF-8 character. When
1514     PCRE_NO_UTF8_CHECK is set, the effect of passing an invalid UTF-8 string as a
1515     subject, or a value of <i>startoffset</i> that does not point to the start of a
1516     UTF-8 character, is undefined. Your program may crash.
1517     <pre>
1518 ph10 461 PCRE_PARTIAL_HARD
1519 ph10 429 PCRE_PARTIAL_SOFT
1520 nigel 75 </pre>
1521 ph10 429 These options turn on the partial matching feature. For backwards
1522     compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1523     occurs if the end of the subject string is reached successfully, but there are
1524     not enough subject characters to complete the match. If this happens when
1525     PCRE_PARTIAL_HARD is set, <b>pcre_exec()</b> immediately returns
1526     PCRE_ERROR_PARTIAL. Otherwise, if PCRE_PARTIAL_SOFT is set, matching continues
1527     by testing any other alternatives. Only if they all fail is PCRE_ERROR_PARTIAL
1528     returned (instead of PCRE_ERROR_NOMATCH). The portion of the string that
1529 ph10 453 was inspected when the partial match was found is set as the first matching
1530     string. There is a more detailed discussion in the
1531 nigel 75 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1532     documentation.
1533     </P>
1534     <br><b>
1535     The string to be matched by <b>pcre_exec()</b>
1536     </b><br>
1537     <P>
1538 nigel 63 The subject string is passed to <b>pcre_exec()</b> as a pointer in
1539 ph10 371 <i>subject</i>, a length (in bytes) in <i>length</i>, and a starting byte offset
1540     in <i>startoffset</i>. In UTF-8 mode, the byte offset must point to the start of
1541     a UTF-8 character. Unlike the pattern string, the subject may contain binary
1542     zero bytes. When the starting offset is zero, the search for a match starts at
1543     the beginning of the subject, and this is by far the most common case.
1544 nigel 63 </P>
1545     <P>
1546     A non-zero starting offset is useful when searching for another match in the
1547     same subject by calling <b>pcre_exec()</b> again after a previous success.
1548     Setting <i>startoffset</i> differs from just passing over a shortened string and
1549     setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1550     lookbehind. For example, consider the pattern
1551     <pre>
1552     \Biss\B
1553 nigel 75 </pre>
1554 nigel 63 which finds occurrences of "iss" in the middle of words. (\B matches only if
1555     the current position in the subject is not a word boundary.) When applied to
1556     the string "Mississipi" the first call to <b>pcre_exec()</b> finds the first
1557     occurrence. If <b>pcre_exec()</b> is called again with just the remainder of the
1558     subject, namely "issipi", it does not match, because \B is always false at the
1559     start of the subject, which is deemed to be a word boundary. However, if
1560     <b>pcre_exec()</b> is passed the entire string again, but with <i>startoffset</i>
1561     set to 4, it finds the second occurrence of "iss" because it is able to look
1562     behind the starting point to discover that it is preceded by a letter.
1563     </P>
1564     <P>
1565     If a non-zero starting offset is passed when the pattern is anchored, one
1566 nigel 75 attempt to match at the given offset is made. This can only succeed if the
1567 nigel 63 pattern does not require the match to be at the start of the subject.
1568     </P>
1569 nigel 75 <br><b>
1570     How <b>pcre_exec()</b> returns captured substrings
1571     </b><br>
1572 nigel 63 <P>
1573     In general, a pattern matches a certain portion of the subject, and in
1574     addition, further substrings from the subject may be picked out by parts of the
1575     pattern. Following the usage in Jeffrey Friedl's book, this is called
1576     "capturing" in what follows, and the phrase "capturing subpattern" is used for
1577     a fragment of a pattern that picks out a substring. PCRE supports several other
1578     kinds of parenthesized subpattern that do not cause substrings to be captured.
1579     </P>
1580     <P>
1581 ph10 371 Captured substrings are returned to the caller via a vector of integers whose
1582     address is passed in <i>ovector</i>. The number of elements in the vector is
1583     passed in <i>ovecsize</i>, which must be a non-negative number. <b>Note</b>: this
1584     argument is NOT the size of <i>ovector</i> in bytes.
1585 nigel 63 </P>
1586     <P>
1587 nigel 75 The first two-thirds of the vector is used to pass back captured substrings,
1588     each substring using a pair of integers. The remaining third of the vector is
1589     used as workspace by <b>pcre_exec()</b> while matching capturing subpatterns,
1590 ph10 371 and is not available for passing back information. The number passed in
1591 nigel 75 <i>ovecsize</i> should always be a multiple of three. If it is not, it is
1592     rounded down.
1593     </P>
1594     <P>
1595     When a match is successful, information about captured substrings is returned
1596     in pairs of integers, starting at the beginning of <i>ovector</i>, and
1597 ph10 371 continuing up to two-thirds of its length at the most. The first element of
1598     each pair is set to the byte offset of the first character in a substring, and
1599     the second is set to the byte offset of the first character after the end of a
1600     substring. <b>Note</b>: these values are always byte offsets, even in UTF-8
1601     mode. They are not character counts.
1602 nigel 63 </P>
1603     <P>
1604 ph10 371 The first pair of integers, <i>ovector[0]</i> and <i>ovector[1]</i>, identify the
1605     portion of the subject string matched by the entire pattern. The next pair is
1606     used for the first capturing subpattern, and so on. The value returned by
1607     <b>pcre_exec()</b> is one more than the highest numbered pair that has been set.
1608     For example, if two substrings have been captured, the returned value is 3. If
1609     there are no capturing subpatterns, the return value from a successful match is
1610     1, indicating that just the first pair of offsets has been set.
1611     </P>
1612     <P>
1613 nigel 63 If a capturing subpattern is matched repeatedly, it is the last portion of the
1614 nigel 75 string that it matched that is returned.
1615 nigel 63 </P>
1616     <P>
1617 nigel 75 If the vector is too small to hold all the captured substring offsets, it is
1618     used as far as possible (up to two-thirds of its length), and the function
1619 ph10 371 returns a value of zero. If the substring offsets are not of interest,
1620     <b>pcre_exec()</b> may be called with <i>ovector</i> passed as NULL and
1621 nigel 63 <i>ovecsize</i> as zero. However, if the pattern contains back references and
1622 nigel 75 the <i>ovector</i> is not big enough to remember the related substrings, PCRE
1623     has to get additional memory for use during matching. Thus it is usually
1624     advisable to supply an <i>ovector</i>.
1625 nigel 63 </P>
1626     <P>
1627 ph10 461 The <b>pcre_fullinfo()</b> function can be used to find out how many capturing
1628 nigel 63 subpatterns there are in a compiled pattern. The smallest size for
1629     <i>ovector</i> that will allow for <i>n</i> captured substrings, in addition to
1630     the offsets of the substring matched by the whole pattern, is (<i>n</i>+1)*3.
1631 nigel 91 </P>
1632     <P>
1633     It is possible for capturing subpattern number <i>n+1</i> to match some part of
1634     the subject when subpattern <i>n</i> has not been used at all. For example, if
1635     the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1636     function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1637     happens, both values in the offset pairs corresponding to unused subpatterns
1638     are set to -1.
1639     </P>
1640     <P>
1641     Offset values that correspond to unused subpatterns at the end of the
1642     expression are also set to -1. For example, if the string "abc" is matched
1643     against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1644     return from the function is 2, because the highest used capturing subpattern
1645     number is 1. However, you can refer to the offsets for the second and third
1646     capturing subpatterns if you wish (assuming the vector is large enough, of
1647     course).
1648     </P>
1649     <P>
1650     Some convenience functions are provided for extracting the captured substrings
1651     as separate strings. These are described below.
1652 nigel 77 <a name="errorlist"></a></P>
1653 nigel 75 <br><b>
1654 nigel 91 Error return values from <b>pcre_exec()</b>
1655 nigel 75 </b><br>
1656 nigel 63 <P>
1657     If <b>pcre_exec()</b> fails, it returns a negative number. The following are
1658     defined in the header file:
1659     <pre>
1660     PCRE_ERROR_NOMATCH (-1)
1661 nigel 75 </pre>
1662 nigel 63 The subject string did not match the pattern.
1663     <pre>
1664     PCRE_ERROR_NULL (-2)
1665 nigel 75 </pre>
1666 nigel 63 Either <i>code</i> or <i>subject</i> was passed as NULL, or <i>ovector</i> was
1667     NULL and <i>ovecsize</i> was not zero.
1668     <pre>
1670 nigel 75 </pre>
1671 nigel 63 An unrecognized bit was set in the <i>options</i> argument.
1672     <pre>
1674 nigel 75 </pre>
1675 nigel 63 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1676 nigel 75 the case when it is passed a junk pointer and to detect when a pattern that was
1677     compiled in an environment of one endianness is run in an environment with the
1678     other endianness. This is the error that PCRE gives when the magic number is
1679     not present.
1680 nigel 63 <pre>
1681 nigel 93 PCRE_ERROR_UNKNOWN_OPCODE (-5)
1682 nigel 75 </pre>
1683 nigel 63 While running the pattern match, an unknown item was encountered in the
1684     compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1685     of the compiled pattern.
1686     <pre>
1688 nigel 75 </pre>
1689 nigel 63 If a pattern contains back references, but the <i>ovector</i> that is passed to
1690     <b>pcre_exec()</b> is not big enough to remember the referenced substrings, PCRE
1691     gets a block of memory at the start of matching to use for this purpose. If the
1692 nigel 75 call via <b>pcre_malloc()</b> fails, this error is given. The memory is
1693     automatically freed at the end of matching.
1694 ph10 535 </P>
1695     <P>
1696     This error is also given if <b>pcre_stack_malloc()</b> fails in
1697     <b>pcre_exec()</b>. This can happen only when PCRE has been compiled with
1698     <b>--disable-stack-for-recursion</b>.
1699 nigel 63 <pre>
1701 nigel 75 </pre>
1702 nigel 63 This error is used by the <b>pcre_copy_substring()</b>,
1703     <b>pcre_get_substring()</b>, and <b>pcre_get_substring_list()</b> functions (see
1704     below). It is never returned by <b>pcre_exec()</b>.
1705     <pre>
1707 nigel 75 </pre>
1708 nigel 87 The backtracking limit, as specified by the <i>match_limit</i> field in a
1709     <b>pcre_extra</b> structure (or defaulted) was reached. See the description
1710     above.
1711     <pre>
1712 nigel 63 PCRE_ERROR_CALLOUT (-9)
1713 nigel 75 </pre>
1714 nigel 63 This error is never generated by <b>pcre_exec()</b> itself. It is provided for
1715     use by callout functions that want to yield a distinctive error code. See the
1716 nigel 75 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1717     documentation for details.
1718 nigel 71 <pre>
1719 nigel 73 PCRE_ERROR_BADUTF8 (-10)
1720 nigel 75 </pre>
1721 nigel 71 A string that contains an invalid UTF-8 byte sequence was passed as a subject.
1722 nigel 73 <pre>
1724 nigel 75 </pre>
1725 nigel 73 The UTF-8 byte sequence that was passed as a subject was valid, but the value
1726     of <i>startoffset</i> did not point to the beginning of a UTF-8 character.
1727 nigel 75 <pre>
1728 nigel 77 PCRE_ERROR_PARTIAL (-12)
1729 nigel 75 </pre>
1730     The subject string did not match, but it did match partially. See the
1731     <a href="pcrepartial.html"><b>pcrepartial</b></a>
1732     documentation for details of partial matching.
1733     <pre>
1734 nigel 77 PCRE_ERROR_BADPARTIAL (-13)
1735 nigel 75 </pre>
1736 ph10 429 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1737     option was used with a compiled pattern containing items that were not
1738 ph10 461 supported for partial matching. From release 8.00 onwards, there are no
1739 ph10 429 restrictions on partial matching.
1740 nigel 75 <pre>
1741 nigel 77 PCRE_ERROR_INTERNAL (-14)
1742 nigel 75 </pre>
1743     An unexpected internal error has occurred. This error could be caused by a bug
1744     in PCRE or by overwriting of the compiled pattern.
1745     <pre>
1746 nigel 77 PCRE_ERROR_BADCOUNT (-15)
1747 nigel 75 </pre>
1748     This error is given if the value of the <i>ovecsize</i> argument is negative.
1749 nigel 93 <pre>
1751     </pre>
1752     The internal recursion limit, as specified by the <i>match_limit_recursion</i>
1753     field in a <b>pcre_extra</b> structure (or defaulted) was reached. See the
1754     description above.
1755     <pre>
1757     </pre>
1758     An invalid combination of PCRE_NEWLINE_<i>xxx</i> options was given.
1759 nigel 73 </P>
1760 nigel 93 <P>
1761 ph10 197 Error numbers -16 to -20 and -22 are not used by <b>pcre_exec()</b>.
1762 nigel 93 </P>
1763 nigel 91 <br><a name="SEC15" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NUMBER</a><br>
1764 nigel 63 <P>
1765     <b>int pcre_copy_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
1766     <b>int <i>stringcount</i>, int <i>stringnumber</i>, char *<i>buffer</i>,</b>
1767     <b>int <i>buffersize</i>);</b>
1768     </P>
1769     <P>
1770     <b>int pcre_get_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
1771     <b>int <i>stringcount</i>, int <i>stringnumber</i>,</b>
1772     <b>const char **<i>stringptr</i>);</b>
1773     </P>
1774     <P>
1775     <b>int pcre_get_substring_list(const char *<i>subject</i>,</b>
1776     <b>int *<i>ovector</i>, int <i>stringcount</i>, const char ***<i>listptr</i>);</b>
1777     </P>
1778     <P>
1779     Captured substrings can be accessed directly by using the offsets returned by
1780     <b>pcre_exec()</b> in <i>ovector</i>. For convenience, the functions
1781     <b>pcre_copy_substring()</b>, <b>pcre_get_substring()</b>, and
1782     <b>pcre_get_substring_list()</b> are provided for extracting captured substrings
1783     as new, separate, zero-terminated strings. These functions identify substrings
1784     by number. The next section describes functions for extracting named
1785 nigel 91 substrings.
1786 nigel 63 </P>
1787     <P>
1788 nigel 91 A substring that contains a binary zero is correctly extracted and has a
1789     further zero added on the end, but the result is not, of course, a C string.
1790     However, you can process such a string by referring to the length that is
1791     returned by <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>.
1792     Unfortunately, the interface to <b>pcre_get_substring_list()</b> is not adequate
1793     for handling strings containing binary zeros, because the end of the final
1794     string is not independently indicated.
1795     </P>
1796     <P>
1797 nigel 63 The first three arguments are the same for all three of these functions:
1798 nigel 75 <i>subject</i> is the subject string that has just been successfully matched,
1799 nigel 63 <i>ovector</i> is a pointer to the vector of integer offsets that was passed to
1800     <b>pcre_exec()</b>, and <i>stringcount</i> is the number of substrings that were
1801     captured by the match, including the substring that matched the entire regular
1802 nigel 75 expression. This is the value returned by <b>pcre_exec()</b> if it is greater
1803     than zero. If <b>pcre_exec()</b> returned zero, indicating that it ran out of
1804     space in <i>ovector</i>, the value passed as <i>stringcount</i> should be the
1805     number of elements in the vector divided by three.
1806 nigel 63 </P>
1807     <P>
1808     The functions <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>
1809     extract a single substring, whose number is given as <i>stringnumber</i>. A
1810 nigel 75 value of zero extracts the substring that matched the entire pattern, whereas
1811 nigel 63 higher values extract the captured substrings. For <b>pcre_copy_substring()</b>,
1812     the string is placed in <i>buffer</i>, whose length is given by
1813     <i>buffersize</i>, while for <b>pcre_get_substring()</b> a new block of memory is
1814     obtained via <b>pcre_malloc</b>, and its address is returned via
1815     <i>stringptr</i>. The yield of the function is the length of the string, not
1816 nigel 93 including the terminating zero, or one of these error codes:
1817 nigel 63 <pre>
1819 nigel 75 </pre>
1820 nigel 63 The buffer was too small for <b>pcre_copy_substring()</b>, or the attempt to get
1821     memory failed for <b>pcre_get_substring()</b>.
1822     <pre>
1824 nigel 75 </pre>
1825 nigel 63 There is no substring whose number is <i>stringnumber</i>.
1826     </P>
1827     <P>
1828     The <b>pcre_get_substring_list()</b> function extracts all available substrings
1829     and builds a list of pointers to them. All this is done in a single block of
1830 nigel 75 memory that is obtained via <b>pcre_malloc</b>. The address of the memory block
1831 nigel 63 is returned via <i>listptr</i>, which is also the start of the list of string
1832     pointers. The end of the list is marked by a NULL pointer. The yield of the
1833 nigel 93 function is zero if all went well, or the error code
1834 nigel 63 <pre>
1836 nigel 75 </pre>
1837 nigel 63 if the attempt to get the memory block failed.
1838     </P>
1839     <P>
1840     When any of these functions encounter a substring that is unset, which can
1841     happen when capturing subpattern number <i>n+1</i> matches some part of the
1842     subject, but subpattern <i>n</i> has not been used at all, they return an empty
1843     string. This can be distinguished from a genuine zero-length substring by
1844     inspecting the appropriate offset in <i>ovector</i>, which is negative for unset
1845     substrings.
1846     </P>
1847     <P>
1848     The two convenience functions <b>pcre_free_substring()</b> and
1849     <b>pcre_free_substring_list()</b> can be used to free the memory returned by
1850     a previous call of <b>pcre_get_substring()</b> or
1851     <b>pcre_get_substring_list()</b>, respectively. They do nothing more than call
1852     the function pointed to by <b>pcre_free</b>, which of course could be called
1853     directly from a C program. However, PCRE is used in some situations where it is
1854 nigel 91 linked via a special interface to another programming language that cannot use
1855 nigel 63 <b>pcre_free</b> directly; it is for these cases that the functions are
1856     provided.
1857     </P>
1858 nigel 91 <br><a name="SEC16" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NAME</a><br>
1859 nigel 63 <P>
1860 nigel 75 <b>int pcre_get_stringnumber(const pcre *<i>code</i>,</b>
1861     <b>const char *<i>name</i>);</b>
1862     </P>
1863     <P>
1864 nigel 63 <b>int pcre_copy_named_substring(const pcre *<i>code</i>,</b>
1865     <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
1866     <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
1867     <b>char *<i>buffer</i>, int <i>buffersize</i>);</b>
1868     </P>
1869     <P>
1870     <b>int pcre_get_named_substring(const pcre *<i>code</i>,</b>
1871     <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
1872     <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
1873     <b>const char **<i>stringptr</i>);</b>
1874     </P>
1875     <P>
1876 nigel 75 To extract a substring by name, you first have to find associated number.
1877     For example, for this pattern
1878 nigel 63 <pre>
1879 nigel 93 (a+)b(?&#60;xxx&#62;\d+)...
1880 nigel 75 </pre>
1881 nigel 91 the number of the subpattern called "xxx" is 2. If the name is known to be
1882     unique (PCRE_DUPNAMES was not set), you can find the number from the name by
1883     calling <b>pcre_get_stringnumber()</b>. The first argument is the compiled
1884     pattern, and the second is the name. The yield of the function is the
1885 nigel 75 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
1886     that name.
1887 nigel 63 </P>
1888     <P>
1889 nigel 75 Given the number, you can extract the substring directly, or use one of the
1890     functions described in the previous section. For convenience, there are also
1891     two functions that do the whole job.
1892 nigel 63 </P>
1893     <P>
1894 nigel 91 Most of the arguments of <b>pcre_copy_named_substring()</b> and
1895     <b>pcre_get_named_substring()</b> are the same as those for the similarly named
1896 nigel 75 functions that extract by number. As these are described in the previous
1897     section, they are not re-described here. There are just two differences:
1898 nigel 63 </P>
1899     <P>
1900     First, instead of a substring number, a substring name is given. Second, there
1901     is an extra argument, given at the start, which is a pointer to the compiled
1902     pattern. This is needed in order to gain access to the name-to-number
1903     translation table.
1904     </P>
1905     <P>
1906     These functions call <b>pcre_get_stringnumber()</b>, and if it succeeds, they
1907 ph10 128 then call <b>pcre_copy_substring()</b> or <b>pcre_get_substring()</b>, as
1908     appropriate. <b>NOTE:</b> If PCRE_DUPNAMES is set and there are duplicate names,
1909     the behaviour may not be what you want (see the next section).
1910 nigel 63 </P>
1911 ph10 392 <P>
1912 ph10 461 <b>Warning:</b> If the pattern uses the (?| feature to set up multiple
1913     subpatterns with the same number, as described in the
1914     <a href="pcrepattern.html#dupsubpatternnumber">section on duplicate subpattern numbers</a>
1915     in the
1916     <a href="pcrepattern.html"><b>pcrepattern</b></a>
1917     page, you cannot use names to distinguish the different subpatterns, because
1918     names are not included in the compiled code. The matching process uses only
1919     numbers. For this reason, the use of different names for subpatterns of the
1920     same number causes an error at compile time.
1921 ph10 392 </P>
1922 nigel 91 <br><a name="SEC17" href="#TOC1">DUPLICATE SUBPATTERN NAMES</a><br>
1923 nigel 63 <P>
1924 nigel 91 <b>int pcre_get_stringtable_entries(const pcre *<i>code</i>,</b>
1925     <b>const char *<i>name</i>, char **<i>first</i>, char **<i>last</i>);</b>
1926     </P>
1927     <P>
1928     When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
1929 ph10 461 are not required to be unique. (Duplicate names are always allowed for
1930     subpatterns with the same number, created by using the (?| feature. Indeed, if
1931     such subpatterns are named, they are required to use the same names.)
1932     </P>
1933     <P>
1934     Normally, patterns with duplicate names are such that in any one match, only
1935     one of the named subpatterns participates. An example is shown in the
1936 nigel 91 <a href="pcrepattern.html"><b>pcrepattern</b></a>
1937 ph10 208 documentation.
1938     </P>
1939     <P>
1940     When duplicates are present, <b>pcre_copy_named_substring()</b> and
1941     <b>pcre_get_named_substring()</b> return the first substring corresponding to
1942     the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
1943     returned; no data is returned. The <b>pcre_get_stringnumber()</b> function
1944     returns one of the numbers that are associated with the name, but it is not
1945     defined which it is.
1946     </P>
1947     <P>
1948 nigel 91 If you want to get full details of all captured substrings for a given name,
1949     you must use the <b>pcre_get_stringtable_entries()</b> function. The first
1950     argument is the compiled pattern, and the second is the name. The third and
1951     fourth are pointers to variables which are updated by the function. After it
1952     has run, they point to the first and last entries in the name-to-number table
1953     for the given name. The function itself returns the length of each entry, or
1954 nigel 93 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
1955     described above in the section entitled <i>Information about a pattern</i>.
1956     Given all the relevant entries for the name, you can extract each of their
1957     numbers, and hence the captured data, if any.
1958 nigel 91 </P>
1959     <br><a name="SEC18" href="#TOC1">FINDING ALL POSSIBLE MATCHES</a><br>
1960     <P>
1961 nigel 77 The traditional matching function uses a similar algorithm to Perl, which stops
1962     when it finds the first match, starting at a given point in the subject. If you
1963     want to find all possible matches, or the longest possible match, consider
1964     using the alternative matching function (see below) instead. If you cannot use
1965     the alternative function, but still need to find all possible matches, you
1966     can kludge it up by making use of the callout facility, which is described in
1967     the
1968     <a href="pcrecallout.html"><b>pcrecallout</b></a>
1969     documentation.
1970     </P>
1971     <P>
1972     What you have to do is to insert a callout right at the end of the pattern.
1973     When your callout function is called, extract and save the current matched
1974     substring. Then return 1, which forces <b>pcre_exec()</b> to backtrack and try
1975     other alternatives. Ultimately, when it runs out of matches, <b>pcre_exec()</b>
1976     will yield PCRE_ERROR_NOMATCH.
1977     <a name="dfamatch"></a></P>
1978 nigel 91 <br><a name="SEC19" href="#TOC1">MATCHING A PATTERN: THE ALTERNATIVE FUNCTION</a><br>
1979 nigel 77 <P>
1980     <b>int pcre_dfa_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
1981     <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
1982     <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>,</b>
1983     <b>int *<i>workspace</i>, int <i>wscount</i>);</b>
1984     </P>
1985     <P>
1986     The function <b>pcre_dfa_exec()</b> is called to match a subject string against
1987 nigel 93 a compiled pattern, using a matching algorithm that scans the subject string
1988     just once, and does not backtrack. This has different characteristics to the
1989     normal algorithm, and is not compatible with Perl. Some of the features of PCRE
1990     patterns are not supported. Nevertheless, there are times when this kind of
1991 ph10 461 matching can be useful. For a discussion of the two matching algorithms, and a
1992 ph10 453 list of features that <b>pcre_dfa_exec()</b> does not support, see the
1993 nigel 77 <a href="pcrematching.html"><b>pcrematching</b></a>
1994     documentation.
1995     </P>
1996     <P>
1997     The arguments for the <b>pcre_dfa_exec()</b> function are the same as for
1998     <b>pcre_exec()</b>, plus two extras. The <i>ovector</i> argument is used in a
1999     different way, and this is described below. The other common arguments are used
2000     in the same way as for <b>pcre_exec()</b>, so their description is not repeated
2001     here.
2002     </P>
2003     <P>
2004     The two additional arguments provide workspace for the function. The workspace
2005     vector should contain at least 20 elements. It is used for keeping track of
2006     multiple paths through the pattern tree. More workspace will be needed for
2007 nigel 91 patterns and subjects where there are a lot of potential matches.
2008 nigel 77 </P>
2009     <P>
2010 nigel 87 Here is an example of a simple call to <b>pcre_dfa_exec()</b>:
2011 nigel 77 <pre>
2012     int rc;
2013     int ovector[10];
2014     int wspace[20];
2015 nigel 87 rc = pcre_dfa_exec(
2016 nigel 77 re, /* result of pcre_compile() */
2017     NULL, /* we didn't study the pattern */
2018     "some string", /* the subject string */
2019     11, /* the length of the subject string */
2020     0, /* start at offset 0 in the subject */
2021     0, /* default options */
2022     ovector, /* vector of integers for substring information */
2023     10, /* number of elements (NOT size in bytes) */
2024     wspace, /* working space vector */
2025     20); /* number of elements (NOT size in bytes) */
2026     </PRE>
2027     </P>
2028     <br><b>
2029     Option bits for <b>pcre_dfa_exec()</b>
2030     </b><br>
2031     <P>
2032     The unused bits of the <i>options</i> argument for <b>pcre_dfa_exec()</b> must be
2033 nigel 91 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
2037     All but the last four of these are exactly the same as for <b>pcre_exec()</b>,
2038     so their description is not repeated here.
2039 nigel 77 <pre>
2040 ph10 429 PCRE_PARTIAL_HARD
2041 ph10 461 PCRE_PARTIAL_SOFT
2042 nigel 77 </pre>
2043 ph10 429 These have the same general effect as they do for <b>pcre_exec()</b>, but the
2044     details are slightly different. When PCRE_PARTIAL_HARD is set for
2045     <b>pcre_dfa_exec()</b>, it returns PCRE_ERROR_PARTIAL if the end of the subject
2046     is reached and there is still at least one matching possibility that requires
2047     additional characters. This happens even if some complete matches have also
2048     been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2049     is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2050     there have been no complete matches, but there is still at least one matching
2051 ph10 453 possibility. The portion of the string that was inspected when the longest
2052     partial match was found is set as the first matching string in both cases.
2053 nigel 77 <pre>
2055     </pre>
2056     Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2057 nigel 93 soon as it has found one match. Because of the way the alternative algorithm
2058     works, this is necessarily the shortest possible match at the first possible
2059     matching point in the subject string.
2060 nigel 77 <pre>
2062     </pre>
2063 ph10 429 When <b>pcre_dfa_exec()</b> returns a partial match, it is possible to call it
2064     again, with additional subject characters, and have it continue with the same
2065     match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2066     <i>workspace</i> and <i>wscount</i> options must reference the same vector as
2067     before because data about the match so far is left in them after a partial
2068     match. There is more discussion of this facility in the
2069 nigel 77 <a href="pcrepartial.html"><b>pcrepartial</b></a>
2070     documentation.
2071     </P>
2072     <br><b>
2073     Successful returns from <b>pcre_dfa_exec()</b>
2074     </b><br>
2075     <P>
2076     When <b>pcre_dfa_exec()</b> succeeds, it may have matched more than one
2077     substring in the subject. Note, however, that all the matches from one run of
2078     the function start at the same point in the subject. The shorter matches are
2079     all initial substrings of the longer matches. For example, if the pattern
2080     <pre>
2081     &#60;.*&#62;
2082     </pre>
2083     is matched against the string
2084     <pre>
2085     This is &#60;something&#62; &#60;something else&#62; &#60;something further&#62; no more
2086     </pre>
2087     the three matched strings are
2088     <pre>
2089     &#60;something&#62;
2090     &#60;something&#62; &#60;something else&#62;
2091     &#60;something&#62; &#60;something else&#62; &#60;something further&#62;
2092     </pre>
2093     On success, the yield of the function is a number greater than zero, which is
2094     the number of matched substrings. The substrings themselves are returned in
2095     <i>ovector</i>. Each string uses two elements; the first is the offset to the
2096 nigel 93 start, and the second is the offset to the end. In fact, all the strings have
2097     the same start offset. (Space could have been saved by giving this only once,
2098     but it was decided to retain some compatibility with the way <b>pcre_exec()</b>
2099     returns data, even though the meaning of the strings is different.)
2100 nigel 77 </P>
2101     <P>
2102     The strings are returned in reverse order of length; that is, the longest
2103     matching string is given first. If there were too many matches to fit into
2104     <i>ovector</i>, the yield of the function is zero, and the vector is filled with
2105     the longest matches.
2106     </P>
2107     <br><b>
2108     Error returns from <b>pcre_dfa_exec()</b>
2109     </b><br>
2110     <P>
2111     The <b>pcre_dfa_exec()</b> function returns a negative number when it fails.
2112     Many of the errors are the same as for <b>pcre_exec()</b>, and these are
2113     described
2114     <a href="#errorlist">above.</a>
2115     There are in addition the following errors that are specific to
2116     <b>pcre_dfa_exec()</b>:
2117     <pre>
2118     PCRE_ERROR_DFA_UITEM (-16)
2119     </pre>
2120     This return is given if <b>pcre_dfa_exec()</b> encounters an item in the pattern
2121     that it does not support, for instance, the use of \C or a back reference.
2122     <pre>
2123     PCRE_ERROR_DFA_UCOND (-17)
2124     </pre>
2125 nigel 93 This return is given if <b>pcre_dfa_exec()</b> encounters a condition item that
2126     uses a back reference for the condition, or a test for recursion in a specific
2127     group. These are not supported.
2128 nigel 77 <pre>
2130     </pre>
2131     This return is given if <b>pcre_dfa_exec()</b> is called with an <i>extra</i>
2132     block that contains a setting of the <i>match_limit</i> field. This is not
2133     supported (it is meaningless).
2134     <pre>
2135     PCRE_ERROR_DFA_WSSIZE (-19)
2136     </pre>
2137     This return is given if <b>pcre_dfa_exec()</b> runs out of space in the
2138     <i>workspace</i> vector.
2139     <pre>
2141     </pre>
2142     When a recursive subpattern is processed, the matching function calls itself
2143     recursively, using private vectors for <i>ovector</i> and <i>workspace</i>. This
2144     error is given if the output vector is not large enough. This should be
2145     extremely rare, as a vector of size 1000 is used.
2146     </P>
2147 nigel 93 <br><a name="SEC20" href="#TOC1">SEE ALSO</a><br>
2148 nigel 77 <P>
2149 nigel 93 <b>pcrebuild</b>(3), <b>pcrecallout</b>(3), <b>pcrecpp(3)</b>(3),
2150     <b>pcrematching</b>(3), <b>pcrepartial</b>(3), <b>pcreposix</b>(3),
2151     <b>pcreprecompile</b>(3), <b>pcresample</b>(3), <b>pcrestack</b>(3).
2152     </P>
2153 ph10 99 <br><a name="SEC21" href="#TOC1">AUTHOR</a><br>
2154 nigel 93 <P>
2155 ph10 99 Philip Hazel
2156 nigel 63 <br>
2157 ph10 99 University Computing Service
2158     <br>
2159     Cambridge CB2 3QH, England.
2160     <br>
2161     </P>
2162     <br><a name="SEC22" href="#TOC1">REVISION</a><br>
2163     <P>
2164 ph10 548 Last updated: 21 June 2010
2165 ph10 99 <br>
2166 ph10 512 Copyright &copy; 1997-2010 University of Cambridge.
2167 ph10 99 <br>
2168 nigel 75 <p>
2169     Return to the <a href="index.html">PCRE index page</a>.
2170     </p>


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

ViewVC Help
Powered by ViewVC 1.1.12