ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log

Revision 572 - (hide annotations) (download) (as text)
Wed Nov 17 17:55:57 2010 UTC (4 years, 5 months ago) by ph10
File MIME type: text/html
File size: 103189 byte(s)
Documentation updates and tidies.

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 572 not try to free it. The offset from the start of the pattern to the byte that
447     was being processed when the error was discovered is placed in the variable
448     pointed to by <i>erroffset</i>, which must not be NULL. If it is, an immediate
449     error is given. Some errors are not detected until checks are carried out when
450     the whole pattern has been scanned; in this case the offset is set to the end
451     of the pattern.
452 nigel 63 </P>
453     <P>
454 ph10 572 Note that the offset is in bytes, not characters, even in UTF-8 mode. It may
455     point into the middle of a UTF-8 character (for example, when
456     PCRE_ERROR_BADUTF8 is returned for an invalid UTF-8 string).
457     </P>
458     <P>
459 nigel 77 If <b>pcre_compile2()</b> is used instead of <b>pcre_compile()</b>, and the
460     <i>errorcodeptr</i> argument is not NULL, a non-zero error code number is
461     returned via this argument in the event of an error. This is in addition to the
462     textual error message. Error codes and messages are listed below.
463     </P>
464     <P>
465 nigel 63 If the final argument, <i>tableptr</i>, is NULL, PCRE uses a default set of
466 nigel 75 character tables that are built when PCRE is compiled, using the default C
467     locale. Otherwise, <i>tableptr</i> must be an address that is the result of a
468     call to <b>pcre_maketables()</b>. This value is stored with the compiled
469     pattern, and used again by <b>pcre_exec()</b>, unless another table pointer is
470     passed to it. For more discussion, see the section on locale support below.
471 nigel 63 </P>
472     <P>
473     This code fragment shows a typical straightforward call to <b>pcre_compile()</b>:
474     <pre>
475     pcre *re;
476     const char *error;
477     int erroffset;
478     re = pcre_compile(
479     "^A.*Z", /* the pattern */
480     0, /* default options */
481     &error, /* for error message */
482     &erroffset, /* for error offset */
483     NULL); /* use default character tables */
484 nigel 75 </pre>
485     The following names for option bits are defined in the <b>pcre.h</b> header
486     file:
487 nigel 63 <pre>
489 nigel 75 </pre>
490 nigel 63 If this bit is set, the pattern is forced to be "anchored", that is, it is
491 nigel 75 constrained to match only at the first matching point in the string that is
492 nigel 63 being searched (the "subject string"). This effect can also be achieved by
493     appropriate constructs in the pattern itself, which is the only way to do it in
494     Perl.
495     <pre>
496 nigel 75 PCRE_AUTO_CALLOUT
497     </pre>
498     If this bit is set, <b>pcre_compile()</b> automatically inserts callout items,
499     all with number 255, before each pattern item. For discussion of the callout
500     facility, see the
501     <a href="pcrecallout.html"><b>pcrecallout</b></a>
502     documentation.
503     <pre>
504 ph10 231 PCRE_BSR_ANYCRLF
506     </pre>
507     These options (which are mutually exclusive) control what the \R escape
508     sequence matches. The choice is either to match only CR, LF, or CRLF, or to
509     match any Unicode newline sequence. The default is specified when PCRE is
510     built. It can be overridden from within the pattern, or by setting an option
511     when a compiled pattern is matched.
512     <pre>
513 nigel 63 PCRE_CASELESS
514 nigel 75 </pre>
515 nigel 63 If this bit is set, letters in the pattern match both upper and lower case
516     letters. It is equivalent to Perl's /i option, and it can be changed within a
517 nigel 77 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
518     concept of case for characters whose values are less than 128, so caseless
519     matching is always possible. For characters with higher values, the concept of
520     case is supported if PCRE is compiled with Unicode property support, but not
521     otherwise. If you want to use caseless matching for characters 128 and above,
522     you must ensure that PCRE is compiled with Unicode property support as well as
523     with UTF-8 support.
524 nigel 63 <pre>
526 nigel 75 </pre>
527 nigel 63 If this bit is set, a dollar metacharacter in the pattern matches only at the
528     end of the subject string. Without this option, a dollar also matches
529 nigel 91 immediately before a newline at the end of the string (but not before any other
530     newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
531     There is no equivalent to this option in Perl, and no way to set it within a
532     pattern.
533 nigel 63 <pre>
535 nigel 75 </pre>
536 ph10 572 If this bit is set, a dot metacharacter in the pattern matches a character of
537     any value, including one that indicates a newline. However, it only ever
538     matches one character, even if newlines are coded as CRLF. Without this option,
539     a dot does not match when the current position is at a newline. This option is
540     equivalent to Perl's /s option, and it can be changed within a pattern by a
541     (?s) option setting. A negative class such as [^a] always matches newline
542     characters, independent of the setting of this option.
543 nigel 63 <pre>
544 nigel 91 PCRE_DUPNAMES
545     </pre>
546     If this bit is set, names used to identify capturing subpatterns need not be
547     unique. This can be helpful for certain types of pattern when it is known that
548     only one instance of the named subpattern can ever be matched. There are more
549     details of named subpatterns below; see also the
550     <a href="pcrepattern.html"><b>pcrepattern</b></a>
551     documentation.
552     <pre>
553 nigel 63 PCRE_EXTENDED
554 nigel 75 </pre>
555 nigel 63 If this bit is set, whitespace data characters in the pattern are totally
556     ignored except when escaped or inside a character class. Whitespace does not
557     include the VT character (code 11). In addition, characters between an
558 nigel 91 unescaped # outside a character class and the next newline, inclusive, are also
559     ignored. This is equivalent to Perl's /x option, and it can be changed within a
560     pattern by a (?x) option setting.
561 nigel 63 </P>
562     <P>
563 ph10 572 Which characters are interpreted as newlines
564     is controlled by the options passed to <b>pcre_compile()</b> or by a special
565     sequence at the start of the pattern, as described in the section entitled
566     <a href="pcrepattern.html#newlines">"Newline conventions"</a>
567     in the <b>pcrepattern</b> documentation. Note that the end of this type of
568     comment is a literal newline sequence in the pattern; escape sequences that
569     happen to represent a newline do not count.
570     </P>
571     <P>
572 nigel 63 This option makes it possible to include comments inside complicated patterns.
573     Note, however, that this applies only to data characters. Whitespace characters
574     may never appear within special character sequences in a pattern, for example
575 ph10 572 within the sequence (?( that introduces a conditional subpattern.
576 nigel 63 <pre>
577     PCRE_EXTRA
578 nigel 75 </pre>
579 nigel 63 This option was invented in order to turn on additional functionality of PCRE
580     that is incompatible with Perl, but it is currently of very little use. When
581     set, any backslash in a pattern that is followed by a letter that has no
582     special meaning causes an error, thus reserving these combinations for future
583     expansion. By default, as in Perl, a backslash followed by a letter with no
584 nigel 91 special meaning is treated as a literal. (Perl can, however, be persuaded to
585 ph10 518 give an error for this, by running it with the -w option.) There are at present
586     no other features controlled by this option. It can also be set by a (?X)
587     option setting within a pattern.
588 nigel 63 <pre>
589 nigel 77 PCRE_FIRSTLINE
590     </pre>
591     If this option is set, an unanchored pattern is required to match before or at
592 nigel 91 the first newline in the subject string, though the matched text may continue
593     over the newline.
594 nigel 77 <pre>
596     </pre>
597     If this option is set, PCRE's behaviour is changed in some ways so that it is
598     compatible with JavaScript rather than Perl. The changes are as follows:
599     </P>
600     <P>
601     (1) A lone closing square bracket in a pattern causes a compile-time error,
602     because this is illegal in JavaScript (by default it is treated as a data
603     character). Thus, the pattern AB]CD becomes illegal when this option is set.
604     </P>
605     <P>
606     (2) At run time, a back reference to an unset subpattern group matches an empty
607     string (by default this causes the current matching alternative to fail). A
608     pattern such as (\1)(a) succeeds when this option is set (assuming it can find
609     an "a" in the subject), whereas it fails by default, for Perl compatibility.
610     <pre>
611 nigel 63 PCRE_MULTILINE
612 nigel 75 </pre>
613     By default, PCRE treats the subject string as consisting of a single line of
614     characters (even if it actually contains newlines). The "start of line"
615 nigel 63 metacharacter (^) matches only at the start of the string, while the "end of
616     line" metacharacter ($) matches only at the end of the string, or before a
617     terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
618     Perl.
619     </P>
620     <P>
621     When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
622 nigel 91 match immediately following or immediately before internal newlines in the
623     subject string, respectively, as well as at the very start and end. This is
624     equivalent to Perl's /m option, and it can be changed within a pattern by a
625     (?m) option setting. If there are no newlines in a subject string, or no
626 nigel 63 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
627     <pre>
628 nigel 91 PCRE_NEWLINE_CR
632 nigel 93 PCRE_NEWLINE_ANY
633 nigel 91 </pre>
634     These options override the default newline definition that was chosen when PCRE
635     was built. Setting the first or the second specifies that a newline is
636 nigel 93 indicated by a single character (CR or LF, respectively). Setting
637     PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
638 ph10 150 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
639     preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
640     that any Unicode newline sequence should be recognized. The Unicode newline
641     sequences are the three just mentioned, plus the single characters VT (vertical
642     tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
643     separator, U+2028), and PS (paragraph separator, U+2029). The last two are
644     recognized only in UTF-8 mode.
645 nigel 91 </P>
646     <P>
647 nigel 93 The newline setting in the options word uses three bits that are treated
648 ph10 150 as a number, giving eight possibilities. Currently only six are used (default
649     plus the five values above). This means that if you set more than one newline
650 nigel 93 option, the combination may or may not be sensible. For example,
651     PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
652 ph10 150 other combinations may yield unused numbers and cause an error.
653 nigel 93 </P>
654     <P>
655 ph10 572 The only time that a line break in a pattern is specially recognized when
656     compiling is when PCRE_EXTENDED is set. CR and LF are whitespace characters,
657     and so are ignored in this mode. Also, an unescaped # outside a character class
658     indicates a comment that lasts until after the next line break sequence. In
659     other circumstances, line break sequences in patterns are treated as literal
660     data.
661 nigel 93 </P>
662     <P>
663     The newline option that is set at compile time becomes the default that is used
664     for <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, but it can be overridden.
665 nigel 91 <pre>
667 nigel 75 </pre>
668 nigel 63 If this option is set, it disables the use of numbered capturing parentheses in
669     the pattern. Any opening parenthesis that is not followed by ? behaves as if it
670     were followed by ?: but named parentheses can still be used for capturing (and
671     they acquire numbers in the usual way). There is no equivalent of this option
672     in Perl.
673     <pre>
674 ph10 535 PCRE_UCP
675     </pre>
676 ph10 572 This option changes the way PCRE processes \B, \b, \D, \d, \S, \s, \W,
677     \w, and some of the POSIX character classes. By default, only ASCII characters
678     are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
679     classify characters. More details are given in the section on
680 ph10 535 <a href="pcre.html#genericchartypes">generic character types</a>
681     in the
682     <a href="pcrepattern.html"><b>pcrepattern</b></a>
683     page. If you set PCRE_UCP, matching one of the items it affects takes much
684     longer. The option is available only if PCRE has been compiled with Unicode
685     property support.
686     <pre>
687 nigel 63 PCRE_UNGREEDY
688 nigel 75 </pre>
689 nigel 63 This option inverts the "greediness" of the quantifiers so that they are not
690     greedy by default, but become greedy if followed by "?". It is not compatible
691     with Perl. It can also be set by a (?U) option setting within the pattern.
692     <pre>
693     PCRE_UTF8
694 nigel 75 </pre>
695 nigel 63 This option causes PCRE to regard both the pattern and the subject as strings
696     of UTF-8 characters instead of single-byte character strings. However, it is
697 nigel 75 available only when PCRE is built to include UTF-8 support. If not, the use
698 nigel 63 of this option provokes an error. Details of how this option changes the
699     behaviour of PCRE are given in the
700     <a href="pcre.html#utf8support">section on UTF-8 support</a>
701     in the main
702     <a href="pcre.html"><b>pcre</b></a>
703     page.
704 nigel 71 <pre>
706 nigel 75 </pre>
707 nigel 71 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
708 ph10 211 automatically checked. There is a discussion about the
709     <a href="pcre.html#utf8strings">validity of UTF-8 strings</a>
710     in the main
711     <a href="pcre.html"><b>pcre</b></a>
712     page. If an invalid UTF-8 sequence of bytes is found, <b>pcre_compile()</b>
713     returns an error. If you already know that your pattern is valid, and you want
714     to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
715     option. When it is set, the effect of passing an invalid UTF-8 string as a
716     pattern is undefined. It may cause your program to crash. Note that this option
717     can also be passed to <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b>, to suppress
718     the UTF-8 validity checking of subject strings.
719 nigel 71 </P>
720 nigel 91 <br><a name="SEC8" href="#TOC1">COMPILATION ERROR CODES</a><br>
721 nigel 63 <P>
722 nigel 77 The following table lists the error codes than may be returned by
723     <b>pcre_compile2()</b>, along with the error messages that may be returned by
724 nigel 93 both compiling functions. As PCRE has developed, some error codes have fallen
725     out of use. To avoid confusion, they have not been re-used.
726 nigel 77 <pre>
727     0 no error
728     1 \ at end of pattern
729     2 \c at end of pattern
730     3 unrecognized character follows \
731     4 numbers out of order in {} quantifier
732     5 number too big in {} quantifier
733     6 missing terminating ] for character class
734     7 invalid escape sequence in character class
735     8 range out of order in character class
736     9 nothing to repeat
737 nigel 93 10 [this code is not in use]
738 nigel 77 11 internal error: unexpected repeat
739 ph10 292 12 unrecognized character after (? or (?-
740 nigel 77 13 POSIX named classes are supported only within a class
741     14 missing )
742     15 reference to non-existent subpattern
743     16 erroffset passed as NULL
744     17 unknown option bit(s) set
745     18 missing ) after comment
746 nigel 93 19 [this code is not in use]
747 ph10 292 20 regular expression is too large
748 nigel 77 21 failed to get memory
749     22 unmatched parentheses
750     23 internal error: code overflow
751     24 unrecognized character after (?&#60;
752     25 lookbehind assertion is not fixed length
753 nigel 91 26 malformed number or name after (?(
754 nigel 77 27 conditional group contains more than two branches
755     28 assertion expected after (?(
756 ph10 182 29 (?R or (?[+-]digits must be followed by )
757 nigel 77 30 unknown POSIX class name
758     31 POSIX collating elements are not supported
759     32 this version of PCRE is not compiled with PCRE_UTF8 support
760 nigel 93 33 [this code is not in use]
761 nigel 77 34 character value in \x{...} sequence is too large
762     35 invalid condition (?(0)
763     36 \C not allowed in lookbehind assertion
764     37 PCRE does not support \L, \l, \N, \U, or \u
765     38 number after (?C is &#62; 255
766     39 closing ) for (?C expected
767     40 recursive call could loop indefinitely
768     41 unrecognized character after (?P
769 nigel 93 42 syntax error in subpattern name (missing terminator)
770 nigel 91 43 two named subpatterns have the same name
771 nigel 77 44 invalid UTF-8 string
772     45 support for \P, \p, and \X has not been compiled
773     46 malformed \P or \p sequence
774     47 unknown property name after \P or \p
775 nigel 91 48 subpattern name is too long (maximum 32 characters)
776 ph10 292 49 too many named subpatterns (maximum 10000)
777 ph10 202 50 [this code is not in use]
778 nigel 91 51 octal value is greater than \377 (not in UTF-8 mode)
779 nigel 93 52 internal error: overran compiling workspace
780 ph10 548 53 internal error: previously-checked referenced subpattern
781     not found
782 nigel 93 54 DEFINE group contains more than one branch
783     55 repeating a DEFINE group is not allowed
784 ph10 231 56 inconsistent NEWLINE options
785 ph10 345 57 \g is not followed by a braced, angle-bracketed, or quoted
786     name/number or by a plain number
787     58 a numbered reference must not be zero
788 ph10 512 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
789 ph10 292 60 (*VERB) not recognized
790     61 number is too big
791     62 subpattern name expected
792     63 digit expected after (?+
793 ph10 345 64 ] is an invalid data character in JavaScript compatibility mode
794 ph10 548 65 different names for subpatterns of the same number are
795     not allowed
796 ph10 512 66 (*MARK) must have an argument
797 ph10 535 67 this version of PCRE is not compiled with PCRE_UCP support
798 ph10 292 </pre>
799     The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
800     be used if the limits were changed when PCRE was built.
801 nigel 77 </P>
802 nigel 91 <br><a name="SEC9" href="#TOC1">STUDYING A PATTERN</a><br>
803 nigel 77 <P>
804     <b>pcre_extra *pcre_study(const pcre *<i>code</i>, int <i>options</i></b>
805 nigel 63 <b>const char **<i>errptr</i>);</b>
806     </P>
807     <P>
808 nigel 75 If a compiled pattern is going to be used several times, it is worth spending
809     more time analyzing it in order to speed up the time taken for matching. The
810 nigel 63 function <b>pcre_study()</b> takes a pointer to a compiled pattern as its first
811 nigel 75 argument. If studying the pattern produces additional information that will
812     help speed up matching, <b>pcre_study()</b> returns a pointer to a
813     <b>pcre_extra</b> block, in which the <i>study_data</i> field points to the
814     results of the study.
815 nigel 63 </P>
816     <P>
817 nigel 75 The returned value from <b>pcre_study()</b> can be passed directly to
818 ph10 461 <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b>. However, a <b>pcre_extra</b> block
819     also contains other fields that can be set by the caller before the block is
820     passed; these are described
821 nigel 75 <a href="#extradata">below</a>
822     in the section on matching a pattern.
823 nigel 63 </P>
824     <P>
825 ph10 461 If studying the pattern does not produce any useful information,
826 nigel 75 <b>pcre_study()</b> returns NULL. In that circumstance, if the calling program
827 ph10 461 wants to pass any of the other fields to <b>pcre_exec()</b> or
828     <b>pcre_dfa_exec()</b>, it must set up its own <b>pcre_extra</b> block.
829 nigel 63 </P>
830     <P>
831 nigel 75 The second argument of <b>pcre_study()</b> contains option bits. At present, no
832     options are defined, and this argument should always be zero.
833     </P>
834     <P>
835 nigel 63 The third argument for <b>pcre_study()</b> is a pointer for an error message. If
836     studying succeeds (even if no data is returned), the variable it points to is
837 nigel 87 set to NULL. Otherwise it is set to point to a textual error message. This is a
838     static string that is part of the library. You must not try to free it. You
839     should test the error pointer for NULL after calling <b>pcre_study()</b>, to be
840     sure that it has run successfully.
841 nigel 63 </P>
842     <P>
843     This is a typical call to <b>pcre_study</b>():
844     <pre>
845     pcre_extra *pe;
846     pe = pcre_study(
847     re, /* result of pcre_compile() */
848     0, /* no options exist */
849     &error); /* set to NULL or points to a message */
850 nigel 75 </pre>
851 ph10 461 Studying a pattern does two things: first, a lower bound for the length of
852     subject string that is needed to match the pattern is computed. This does not
853     mean that there are any strings of that length that match, but it does
854     guarantee that no shorter strings match. The value is used by
855     <b>pcre_exec()</b> and <b>pcre_dfa_exec()</b> to avoid wasting time by trying to
856     match strings that are shorter than the lower bound. You can find out the value
857     in a calling program via the <b>pcre_fullinfo()</b> function.
858     </P>
859     <P>
860     Studying a pattern is also useful for non-anchored patterns that do not have a
861     single fixed starting character. A bitmap of possible starting bytes is
862     created. This speeds up finding a position in the subject at which to start
863     matching.
864 ph10 548 </P>
865     <P>
866     The two optimizations just described can be disabled by setting the
867     PCRE_NO_START_OPTIMIZE option when calling <b>pcre_exec()</b> or
868     <b>pcre_dfa_exec()</b>. You might want to do this if your pattern contains
869 ph10 572 callouts or (*MARK), and you want to make use of these facilities in cases
870     where matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
871 ph10 548 <a href="#execoptions">below.</a>
872 nigel 75 <a name="localesupport"></a></P>
873 nigel 91 <br><a name="SEC10" href="#TOC1">LOCALE SUPPORT</a><br>
874 nigel 63 <P>
875 ph10 142 PCRE handles caseless matching, and determines whether characters are letters,
876 nigel 75 digits, or whatever, by reference to a set of tables, indexed by character
877 nigel 77 value. When running in UTF-8 mode, this applies only to characters with codes
878 ph10 535 less than 128. By default, higher-valued codes never match escapes such as \w
879     or \d, but they can be tested with \p if PCRE is built with Unicode character
880     property support. Alternatively, the PCRE_UCP option can be set at compile
881     time; this causes \w and friends to use Unicode property support instead of
882     built-in tables. The use of locales with Unicode is discouraged. If you are
883     handling characters with codes greater than 128, you should either use UTF-8
884     and Unicode, or use locales, but not try to mix the two.
885 nigel 63 </P>
886     <P>
887 ph10 142 PCRE contains an internal set of tables that are used when the final argument
888     of <b>pcre_compile()</b> is NULL. These are sufficient for many applications.
889     Normally, the internal tables recognize only ASCII characters. However, when
890     PCRE is built, it is possible to cause the internal tables to be rebuilt in the
891     default "C" locale of the local system, which may cause them to be different.
892 nigel 63 </P>
893     <P>
894 ph10 142 The internal tables can always be overridden by tables supplied by the
895     application that calls PCRE. These may be created in a different locale from
896     the default. As more and more applications change to using Unicode, the need
897     for this locale support is expected to die away.
898     </P>
899     <P>
900 nigel 75 External tables are built by calling the <b>pcre_maketables()</b> function,
901     which has no arguments, in the relevant locale. The result can then be passed
902     to <b>pcre_compile()</b> or <b>pcre_exec()</b> as often as necessary. For
903     example, to build and use tables that are appropriate for the French locale
904     (where accented characters with values greater than 128 are treated as letters),
905     the following code could be used:
906 nigel 63 <pre>
907 nigel 75 setlocale(LC_CTYPE, "fr_FR");
908 nigel 63 tables = pcre_maketables();
909     re = pcre_compile(..., tables);
910 nigel 75 </pre>
911 ph10 142 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
912     are using Windows, the name for the French locale is "french".
913     </P>
914     <P>
915 nigel 75 When <b>pcre_maketables()</b> runs, the tables are built in memory that is
916     obtained via <b>pcre_malloc</b>. It is the caller's responsibility to ensure
917     that the memory containing the tables remains available for as long as it is
918     needed.
919 nigel 63 </P>
920     <P>
921 nigel 75 The pointer that is passed to <b>pcre_compile()</b> is saved with the compiled
922 nigel 63 pattern, and the same tables are used via this pointer by <b>pcre_study()</b>
923 nigel 75 and normally also by <b>pcre_exec()</b>. Thus, by default, for any single
924     pattern, compilation, studying and matching all happen in the same locale, but
925     different patterns can be compiled in different locales.
926 nigel 63 </P>
927     <P>
928 nigel 75 It is possible to pass a table pointer or NULL (indicating the use of the
929     internal tables) to <b>pcre_exec()</b>. Although not intended for this purpose,
930     this facility could be used to match a pattern in a different locale from the
931     one in which it was compiled. Passing table pointers at run time is discussed
932     below in the section on matching a pattern.
933     </P>
934 nigel 91 <br><a name="SEC11" href="#TOC1">INFORMATION ABOUT A PATTERN</a><br>
935 nigel 75 <P>
936 nigel 63 <b>int pcre_fullinfo(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
937     <b>int <i>what</i>, void *<i>where</i>);</b>
938     </P>
939     <P>
940     The <b>pcre_fullinfo()</b> function returns information about a compiled
941     pattern. It replaces the obsolete <b>pcre_info()</b> function, which is
942     nevertheless retained for backwards compability (and is documented below).
943     </P>
944     <P>
945     The first argument for <b>pcre_fullinfo()</b> is a pointer to the compiled
946     pattern. The second argument is the result of <b>pcre_study()</b>, or NULL if
947     the pattern was not studied. The third argument specifies which piece of
948     information is required, and the fourth argument is a pointer to a variable
949     to receive the data. The yield of the function is zero for success, or one of
950     the following negative numbers:
951     <pre>
952     PCRE_ERROR_NULL the argument <i>code</i> was NULL
953     the argument <i>where</i> was NULL
954     PCRE_ERROR_BADMAGIC the "magic number" was not found
955     PCRE_ERROR_BADOPTION the value of <i>what</i> was invalid
956 nigel 75 </pre>
957     The "magic number" is placed at the start of each compiled pattern as an simple
958     check against passing an arbitrary memory pointer. Here is a typical call of
959     <b>pcre_fullinfo()</b>, to obtain the length of the compiled pattern:
960 nigel 63 <pre>
961     int rc;
962 nigel 91 size_t length;
963 nigel 63 rc = pcre_fullinfo(
964     re, /* result of pcre_compile() */
965     pe, /* result of pcre_study(), or NULL */
966     PCRE_INFO_SIZE, /* what is required */
967     &length); /* where to put the data */
968 nigel 75 </pre>
969 nigel 63 The possible values for the third argument are defined in <b>pcre.h</b>, and are
970     as follows:
971     <pre>
973 nigel 75 </pre>
974 nigel 63 Return the number of the highest back reference in the pattern. The fourth
975     argument should point to an <b>int</b> variable. Zero is returned if there are
976     no back references.
977     <pre>
979 nigel 75 </pre>
980 nigel 63 Return the number of capturing subpatterns in the pattern. The fourth argument
981 nigel 75 should point to an <b>int</b> variable.
982 nigel 63 <pre>
984 nigel 75 </pre>
985     Return a pointer to the internal default character tables within PCRE. The
986     fourth argument should point to an <b>unsigned char *</b> variable. This
987     information call is provided for internal use by the <b>pcre_study()</b>
988     function. External callers can cause PCRE to use its internal tables by passing
989     a NULL table pointer.
990     <pre>
992 nigel 75 </pre>
993 nigel 63 Return information about the first byte of any matched string, for a
994 nigel 91 non-anchored pattern. The fourth argument should point to an <b>int</b>
995     variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
996     still recognized for backwards compatibility.)
997 nigel 63 </P>
998     <P>
999 nigel 75 If there is a fixed first byte, for example, from a pattern such as
1000 nigel 93 (cat|cow|coyote), its value is returned. Otherwise, if either
1001 nigel 75 <br>
1002     <br>
1003 nigel 63 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1004     starts with "^", or
1005 nigel 75 <br>
1006     <br>
1007 nigel 63 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1008     (if it were set, the pattern would be anchored),
1009 nigel 75 <br>
1010     <br>
1011 nigel 63 -1 is returned, indicating that the pattern matches only at the start of a
1012     subject string or after any newline within the string. Otherwise -2 is
1013     returned. For anchored patterns, -2 is returned.
1014     <pre>
1016 nigel 75 </pre>
1017 nigel 63 If the pattern was studied, and this resulted in the construction of a 256-bit
1018     table indicating a fixed set of bytes for the first byte in any matching
1019     string, a pointer to the table is returned. Otherwise NULL is returned. The
1020     fourth argument should point to an <b>unsigned char *</b> variable.
1021     <pre>
1023     </pre>
1024     Return 1 if the pattern contains any explicit matches for CR or LF characters,
1025 ph10 243 otherwise 0. The fourth argument should point to an <b>int</b> variable. An
1026     explicit match is either a literal CR or LF character, or \r or \n.
1027 ph10 227 <pre>
1028 ph10 172 PCRE_INFO_JCHANGED
1029     </pre>
1030 ph10 286 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1031     0. The fourth argument should point to an <b>int</b> variable. (?J) and
1032     (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1033 ph10 172 <pre>
1035 nigel 75 </pre>
1036 nigel 65 Return the value of the rightmost literal byte that must exist in any matched
1037     string, other than at its start, if such a byte has been recorded. The fourth
1038     argument should point to an <b>int</b> variable. If there is no such byte, -1 is
1039     returned. For anchored patterns, a last literal byte is recorded only if it
1040     follows something of variable length. For example, for the pattern
1041     /^a\d+z\d+/ the returned value is "z", but for /^a\dz\d/ the returned value
1042     is -1.
1043 nigel 63 <pre>
1045     </pre>
1046     If the pattern was studied and a minimum length for matching subject strings
1047     was computed, its value is returned. Otherwise the returned value is -1. The
1048     value is a number of characters, not bytes (this may be relevant in UTF-8
1049     mode). The fourth argument should point to an <b>int</b> variable. A
1050     non-negative value is a lower bound to the length of any matching string. There
1051     may not be any strings of that length that do actually match, but every string
1052     that does match is at least that long.
1053     <pre>
1057 nigel 75 </pre>
1058 nigel 63 PCRE supports the use of named as well as numbered capturing parentheses. The
1059     names are just an additional way of identifying the parentheses, which still
1060 nigel 91 acquire numbers. Several convenience functions such as
1061     <b>pcre_get_named_substring()</b> are provided for extracting captured
1062     substrings by name. It is also possible to extract the data directly, by first
1063     converting the name to a number in order to access the correct pointers in the
1064     output vector (described with <b>pcre_exec()</b> below). To do the conversion,
1065     you need to use the name-to-number map, which is described by these three
1066     values.
1067 nigel 63 </P>
1068     <P>
1069     The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1070     the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1071     entry; both of these return an <b>int</b> value. The entry size depends on the
1072     length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1073     entry of the table (a pointer to <b>char</b>). The first two bytes of each entry
1074     are the number of the capturing parenthesis, most significant byte first. The
1075 ph10 461 rest of the entry is the corresponding name, zero terminated.
1076     </P>
1077     <P>
1078     The names are in alphabetical order. Duplicate names may appear if (?| is used
1079     to create multiple groups with the same number, as described in the
1080     <a href="pcrepattern.html#dupsubpatternnumber">section on duplicate subpattern numbers</a>
1081     in the
1082     <a href="pcrepattern.html"><b>pcrepattern</b></a>
1083     page. Duplicate names for subpatterns with different numbers are permitted only
1084     if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1085     table in the order in which they were found in the pattern. In the absence of
1086     (?| this is the order of increasing number; when (?| is used this is not
1087     necessarily the case because later subpatterns may have lower numbers.
1088     </P>
1089     <P>
1090     As a simple example of the name/number table, consider the following pattern
1091     (assume PCRE_EXTENDED is set, so white space - including newlines - is
1092     ignored):
1093 nigel 63 <pre>
1094 nigel 93 (?&#60;date&#62; (?&#60;year&#62;(\d\d)?\d\d) - (?&#60;month&#62;\d\d) - (?&#60;day&#62;\d\d) )
1095 nigel 75 </pre>
1096 nigel 63 There are four named subpatterns, so the table has four entries, and each entry
1097     in the table is eight bytes long. The table is as follows, with non-printing
1098 nigel 75 bytes shows in hexadecimal, and undefined bytes shown as ??:
1099 nigel 63 <pre>
1100     00 01 d a t e 00 ??
1101     00 05 d a y 00 ?? ??
1102     00 04 m o n t h 00
1103     00 02 y e a r 00 ??
1104 nigel 75 </pre>
1105     When writing code to extract data from named subpatterns using the
1106 nigel 91 name-to-number map, remember that the length of the entries is likely to be
1107 nigel 75 different for each compiled pattern.
1108 nigel 63 <pre>
1110     </pre>
1111 ph10 453 Return 1 if the pattern can be used for partial matching with
1112     <b>pcre_exec()</b>, otherwise 0. The fourth argument should point to an
1113     <b>int</b> variable. From release 8.00, this always returns 1, because the
1114     restrictions that previously applied to partial matching have been lifted. The
1115 ph10 172 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1116 ph10 429 documentation gives details of partial matching.
1117 ph10 172 <pre>
1118 nigel 63 PCRE_INFO_OPTIONS
1119 nigel 75 </pre>
1120 nigel 63 Return a copy of the options with which the pattern was compiled. The fourth
1121     argument should point to an <b>unsigned long int</b> variable. These option bits
1122     are those specified in the call to <b>pcre_compile()</b>, modified by any
1123 ph10 197 top-level option settings at the start of the pattern itself. In other words,
1124     they are the options that will be in force when matching starts. For example,
1125     if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1127 nigel 63 </P>
1128     <P>
1129     A pattern is automatically anchored by PCRE if all of its top-level
1130     alternatives begin with one of the following:
1131     <pre>
1132     ^ unless PCRE_MULTILINE is set
1133     \A always
1134     \G always
1135 nigel 75 .* if PCRE_DOTALL is set and there are no back references to the subpattern in which .* appears
1136     </pre>
1137 nigel 63 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1138     <b>pcre_fullinfo()</b>.
1139     <pre>
1141 nigel 75 </pre>
1142 nigel 63 Return the size of the compiled pattern, that is, the value that was passed as
1143     the argument to <b>pcre_malloc()</b> when PCRE was getting memory in which to
1144     place the compiled data. The fourth argument should point to a <b>size_t</b>
1145     variable.
1146     <pre>
1148 nigel 75 </pre>
1149     Return the size of the data block pointed to by the <i>study_data</i> field in
1150 nigel 63 a <b>pcre_extra</b> block. That is, it is the value that was passed to
1151     <b>pcre_malloc()</b> when PCRE was getting memory into which to place the data
1152 ph10 461 created by <b>pcre_study()</b>. If <b>pcre_extra</b> is NULL, or there is no
1153     study data, zero is returned. The fourth argument should point to a
1154 nigel 63 <b>size_t</b> variable.
1155     </P>
1156 nigel 91 <br><a name="SEC12" href="#TOC1">OBSOLETE INFO FUNCTION</a><br>
1157 nigel 63 <P>
1158     <b>int pcre_info(const pcre *<i>code</i>, int *<i>optptr</i>, int</b>
1159     <b>*<i>firstcharptr</i>);</b>
1160     </P>
1161     <P>
1162     The <b>pcre_info()</b> function is now obsolete because its interface is too
1163     restrictive to return all the available data about a compiled pattern. New
1164     programs should use <b>pcre_fullinfo()</b> instead. The yield of
1165     <b>pcre_info()</b> is the number of capturing subpatterns, or one of the
1166     following negative numbers:
1167     <pre>
1168     PCRE_ERROR_NULL the argument <i>code</i> was NULL
1169     PCRE_ERROR_BADMAGIC the "magic number" was not found
1170 nigel 75 </pre>
1171 nigel 63 If the <i>optptr</i> argument is not NULL, a copy of the options with which the
1172     pattern was compiled is placed in the integer it points to (see
1173     PCRE_INFO_OPTIONS above).
1174     </P>
1175     <P>
1176     If the pattern is not anchored and the <i>firstcharptr</i> argument is not NULL,
1177     it is used to pass back information about the first character of any matched
1178     string (see PCRE_INFO_FIRSTBYTE above).
1179     </P>
1180 nigel 91 <br><a name="SEC13" href="#TOC1">REFERENCE COUNTS</a><br>
1181 nigel 63 <P>
1182 nigel 77 <b>int pcre_refcount(pcre *<i>code</i>, int <i>adjust</i>);</b>
1183     </P>
1184     <P>
1185     The <b>pcre_refcount()</b> function is used to maintain a reference count in the
1186     data block that contains a compiled pattern. It is provided for the benefit of
1187     applications that operate in an object-oriented manner, where different parts
1188     of the application may be using the same compiled pattern, but you want to free
1189     the block when they are all done.
1190     </P>
1191     <P>
1192     When a pattern is compiled, the reference count field is initialized to zero.
1193     It is changed only by calling this function, whose action is to add the
1194     <i>adjust</i> value (which may be positive or negative) to it. The yield of the
1195     function is the new value. However, the value of the count is constrained to
1196     lie between 0 and 65535, inclusive. If the new value is outside these limits,
1197     it is forced to the appropriate limit value.
1198     </P>
1199     <P>
1200     Except when it is zero, the reference count is not correctly preserved if a
1201     pattern is compiled on one host and then transferred to a host whose byte-order
1202     is different. (This seems a highly unlikely scenario.)
1203     </P>
1204 nigel 91 <br><a name="SEC14" href="#TOC1">MATCHING A PATTERN: THE TRADITIONAL FUNCTION</a><br>
1205 nigel 77 <P>
1206 nigel 63 <b>int pcre_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
1207     <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
1208     <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>);</b>
1209     </P>
1210     <P>
1211     The function <b>pcre_exec()</b> is called to match a subject string against a
1212 nigel 75 compiled pattern, which is passed in the <i>code</i> argument. If the
1213 ph10 461 pattern was studied, the result of the study should be passed in the
1214 nigel 77 <i>extra</i> argument. This function is the main matching facility of the
1215     library, and it operates in a Perl-like manner. For specialist use there is
1216     also an alternative matching function, which is described
1217     <a href="#dfamatch">below</a>
1218     in the section about the <b>pcre_dfa_exec()</b> function.
1219 nigel 63 </P>
1220     <P>
1221 nigel 75 In most applications, the pattern will have been compiled (and optionally
1222     studied) in the same process that calls <b>pcre_exec()</b>. However, it is
1223     possible to save compiled patterns and study data, and then use them later
1224     in different processes, possibly even on different hosts. For a discussion
1225     about this, see the
1226     <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1227     documentation.
1228 nigel 63 </P>
1229     <P>
1230 nigel 75 Here is an example of a simple call to <b>pcre_exec()</b>:
1231 nigel 63 <pre>
1232     int rc;
1233     int ovector[30];
1234     rc = pcre_exec(
1235     re, /* result of pcre_compile() */
1236     NULL, /* we didn't study the pattern */
1237     "some string", /* the subject string */
1238     11, /* the length of the subject string */
1239     0, /* start at offset 0 in the subject */
1240     0, /* default options */
1241 nigel 75 ovector, /* vector of integers for substring information */
1242 nigel 77 30); /* number of elements (NOT size in bytes) */
1243 nigel 75 <a name="extradata"></a></PRE>
1244 nigel 63 </P>
1245 nigel 75 <br><b>
1246     Extra data for <b>pcre_exec()</b>
1247     </b><br>
1248 nigel 63 <P>
1249     If the <i>extra</i> argument is not NULL, it must point to a <b>pcre_extra</b>
1250     data block. The <b>pcre_study()</b> function returns such a block (when it
1251     doesn't return NULL), but you can also create one for yourself, and pass
1252 nigel 87 additional information in it. The <b>pcre_extra</b> block contains the following
1253     fields (not necessarily in this order):
1254 nigel 63 <pre>
1255     unsigned long int <i>flags</i>;
1256     void *<i>study_data</i>;
1257     unsigned long int <i>match_limit</i>;
1258 nigel 87 unsigned long int <i>match_limit_recursion</i>;
1259 nigel 63 void *<i>callout_data</i>;
1260 nigel 75 const unsigned char *<i>tables</i>;
1261 ph10 512 unsigned char **<i>mark</i>;
1262 nigel 75 </pre>
1263 nigel 63 The <i>flags</i> field is a bitmap that specifies which of the other fields
1264     are set. The flag bits are:
1265     <pre>
1270 nigel 75 PCRE_EXTRA_TABLES
1271 ph10 512 PCRE_EXTRA_MARK
1272 nigel 75 </pre>
1273 nigel 63 Other flag bits should be set to zero. The <i>study_data</i> field is set in the
1274     <b>pcre_extra</b> block that is returned by <b>pcre_study()</b>, together with
1275 nigel 75 the appropriate flag bit. You should not set this yourself, but you may add to
1276     the block by setting the other fields and their corresponding flag bits.
1277 nigel 63 </P>
1278     <P>
1279     The <i>match_limit</i> field provides a means of preventing PCRE from using up a
1280     vast amount of resources when running patterns that are not going to match,
1281     but which have a very large number of possibilities in their search trees. The
1282 ph10 461 classic example is a pattern that uses nested unlimited repeats.
1283 nigel 63 </P>
1284     <P>
1285 nigel 75 Internally, PCRE uses a function called <b>match()</b> which it calls repeatedly
1286 nigel 87 (sometimes recursively). The limit set by <i>match_limit</i> is imposed on the
1287     number of times this function is called during a match, which has the effect of
1288     limiting the amount of backtracking that can take place. For patterns that are
1289     not anchored, the count restarts from zero for each position in the subject
1290     string.
1291 nigel 75 </P>
1292     <P>
1293 nigel 87 The default value for the limit can be set when PCRE is built; the default
1294 nigel 63 default is 10 million, which handles all but the most extreme cases. You can
1295 nigel 87 override the default by suppling <b>pcre_exec()</b> with a <b>pcre_extra</b>
1296     block in which <i>match_limit</i> is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1297     the <i>flags</i> field. If the limit is exceeded, <b>pcre_exec()</b> returns
1299 nigel 63 </P>
1300     <P>
1301 nigel 87 The <i>match_limit_recursion</i> field is similar to <i>match_limit</i>, but
1302     instead of limiting the total number of times that <b>match()</b> is called, it
1303     limits the depth of recursion. The recursion depth is a smaller number than the
1304     total number of calls, because not all calls to <b>match()</b> are recursive.
1305     This limit is of use only if it is set smaller than <i>match_limit</i>.
1306     </P>
1307     <P>
1308     Limiting the recursion depth limits the amount of stack that can be used, or,
1309     when PCRE has been compiled to use memory on the heap instead of the stack, the
1310     amount of heap memory that can be used.
1311     </P>
1312     <P>
1313     The default value for <i>match_limit_recursion</i> can be set when PCRE is
1314     built; the default default is the same value as the default for
1315     <i>match_limit</i>. You can override the default by suppling <b>pcre_exec()</b>
1316     with a <b>pcre_extra</b> block in which <i>match_limit_recursion</i> is set, and
1317     PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the <i>flags</i> field. If the limit
1318     is exceeded, <b>pcre_exec()</b> returns PCRE_ERROR_RECURSIONLIMIT.
1319     </P>
1320     <P>
1321 ph10 453 The <i>callout_data</i> field is used in conjunction with the "callout" feature,
1322     and is described in the
1323 nigel 75 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1324     documentation.
1325 nigel 63 </P>
1326     <P>
1327 nigel 75 The <i>tables</i> field is used to pass a character tables pointer to
1328     <b>pcre_exec()</b>; this overrides the value that is stored with the compiled
1329     pattern. A non-NULL value is stored with the compiled pattern only if custom
1330     tables were supplied to <b>pcre_compile()</b> via its <i>tableptr</i> argument.
1331     If NULL is passed to <b>pcre_exec()</b> using this mechanism, it forces PCRE's
1332     internal tables to be used. This facility is helpful when re-using patterns
1333     that have been saved after compiling with an external set of tables, because
1334     the external tables might be at a different address when <b>pcre_exec()</b> is
1335     called. See the
1336     <a href="pcreprecompile.html"><b>pcreprecompile</b></a>
1337     documentation for a discussion of saving compiled patterns for later use.
1338 ph10 512 </P>
1339     <P>
1340     If PCRE_EXTRA_MARK is set in the <i>flags</i> field, the <i>mark</i> field must
1341     be set to point to a <b>char *</b> variable. If the pattern contains any
1342     backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1343     a name to pass back, a pointer to the name string (zero terminated) is placed
1344     in the variable pointed to by the <i>mark</i> field. The names are within the
1345     compiled pattern; if you wish to retain such a name you must copy it before
1346     freeing the memory of a compiled pattern. If there is no name to pass back, the
1347     variable pointed to by the <i>mark</i> field set to NULL. For details of the
1348     backtracking control verbs, see the section entitled
1349     <a href="pcrepattern#backtrackcontrol">"Backtracking control"</a>
1350     in the
1351     <a href="pcrepattern.html"><b>pcrepattern</b></a>
1352     documentation.
1353 ph10 227 <a name="execoptions"></a></P>
1354 nigel 75 <br><b>
1355     Option bits for <b>pcre_exec()</b>
1356     </b><br>
1357 nigel 63 <P>
1358 nigel 75 The unused bits of the <i>options</i> argument for <b>pcre_exec()</b> must be
1359 nigel 91 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
1363 nigel 63 <pre>
1364 nigel 75 PCRE_ANCHORED
1365     </pre>
1366     The PCRE_ANCHORED option limits <b>pcre_exec()</b> to matching at the first
1367     matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1368     to be anchored by virtue of its contents, it cannot be made unachored at
1369     matching time.
1370     <pre>
1371 ph10 231 PCRE_BSR_ANYCRLF
1373     </pre>
1374     These options (which are mutually exclusive) control what the \R escape
1375     sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1376     match any Unicode newline sequence. These options override the choice that was
1377     made or defaulted when the pattern was compiled.
1378     <pre>
1379 nigel 91 PCRE_NEWLINE_CR
1383 nigel 93 PCRE_NEWLINE_ANY
1384 nigel 91 </pre>
1385     These options override the newline definition that was chosen or defaulted when
1386 nigel 93 the pattern was compiled. For details, see the description of
1387     <b>pcre_compile()</b> above. During matching, the newline choice affects the
1388     behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1389     the way the match position is advanced after a match failure for an unanchored
1390 ph10 227 pattern.
1391     </P>
1392     <P>
1394     match attempt for an unanchored pattern fails when the current position is at a
1395 ph10 231 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1396 ph10 227 characters, the match position is advanced by two characters instead of one, in
1397     other words, to after the CRLF.
1398     </P>
1399     <P>
1400     The above rule is a compromise that makes the most common cases work as
1401     expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1402     set), it does not match the string "\r\nA" because, after failing at the
1403     start, it skips both the CR and the LF before retrying. However, the pattern
1404     [\r\n]A does match that string, because it contains an explicit CR or LF
1405     reference, and so advances only by one character after the first failure.
1406     </P>
1407     <P>
1408 ph10 231 An explicit match for CR of LF is either a literal appearance of one of those
1409     characters, or one of the \r or \n escape sequences. Implicit matches such as
1410     [^X] do not count, nor does \s (which includes CR and LF in the characters
1411     that it matches).
1412     </P>
1413     <P>
1414 ph10 227 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1415     valid newline sequence and explicit \r or \n escapes appear in the pattern.
1416 nigel 91 <pre>
1417 nigel 63 PCRE_NOTBOL
1418 nigel 75 </pre>
1419     This option specifies that first character of the subject string is not the
1420     beginning of a line, so the circumflex metacharacter should not match before
1421     it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1422     never to match. This option affects only the behaviour of the circumflex
1423     metacharacter. It does not affect \A.
1424 nigel 63 <pre>
1425     PCRE_NOTEOL
1426 nigel 75 </pre>
1427     This option specifies that the end of the subject string is not the end of a
1428     line, so the dollar metacharacter should not match it nor (except in multiline
1429     mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1430     compile time) causes dollar never to match. This option affects only the
1431     behaviour of the dollar metacharacter. It does not affect \Z or \z.
1432 nigel 63 <pre>
1434 nigel 75 </pre>
1435 nigel 63 An empty string is not considered to be a valid match if this option is set. If
1436     there are alternatives in the pattern, they are tried. If all the alternatives
1437     match the empty string, the entire match fails. For example, if the pattern
1438     <pre>
1439     a?b?
1440 nigel 75 </pre>
1441 ph10 453 is applied to a string not beginning with "a" or "b", it matches an empty
1442 nigel 63 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1443     valid, so PCRE searches further into the string for occurrences of "a" or "b".
1444 ph10 453 <pre>
1446     </pre>
1447 ph10 461 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1448 ph10 453 the start of the subject is permitted. If the pattern is anchored, such a match
1449     can occur only if the pattern contains \K.
1450 nigel 63 </P>
1451     <P>
1452 ph10 453 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1453     does make a special case of a pattern match of the empty string within its
1454     <b>split()</b> function, and when using the /g modifier. It is possible to
1455     emulate Perl's behaviour after matching a null string by first trying the match
1456     again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1457     if that fails, by advancing the starting offset (see below) and trying an
1458     ordinary match again. There is some code that demonstrates how to do this in
1459     the
1460 ph10 429 <a href="pcredemo.html"><b>pcredemo</b></a>
1461 ph10 572 sample program. In the most general case, you have to check to see if the
1462     newline convention recognizes CRLF as a newline, and if so, and the current
1463 ph10 567 character is CR followed by LF, advance the starting offset by two characters
1464     instead of one.
1465 nigel 75 <pre>
1467     </pre>
1468     There are a number of optimizations that <b>pcre_exec()</b> uses at the start of
1469 ph10 545 a match, in order to speed up the process. For example, if it is known that an
1470     unanchored match must start with a specific character, it searches the subject
1471     for that character, and fails immediately if it cannot find it, without
1472     actually running the main matching function. This means that a special item
1473     such as (*COMMIT) at the start of a pattern is not considered until after a
1474 ph10 548 suitable starting point for the match has been found. When callouts or (*MARK)
1475     items are in use, these "start-up" optimizations can cause them to be skipped
1476     if the pattern is never actually used. The start-up optimizations are in effect
1477     a pre-scan of the subject that takes place before the pattern is run.
1478     </P>
1479     <P>
1480     The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1481     causing performance to suffer, but ensuring that in cases where the result is
1482     "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1483     are considered at every possible starting position in the subject string.
1484     Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1485     Consider the pattern
1486 ph10 392 <pre>
1487 ph10 548 (*COMMIT)ABC
1488     </pre>
1489     When this is compiled, PCRE records the fact that a match must start with the
1490     character "A". Suppose the subject string is "DEFABC". The start-up
1491     optimization scans along the subject, finds "A" and runs the first match
1492     attempt from there. The (*COMMIT) item means that the pattern must match the
1493     current starting position, which in this case, it does. However, if the same
1494     match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1495     subject string does not happen. The first match attempt is run starting from
1496     "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1497     the overall result is "no match". If the pattern is studied, more start-up
1498     optimizations may be used. For example, a minimum length for the subject may be
1499     recorded. Consider the pattern
1500     <pre>
1501     (*MARK:A)(X|Y)
1502     </pre>
1503     The minimum length for a match is one character. If the subject is "ABC", there
1504     will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1505     If the pattern is studied, the final attempt does not take place, because PCRE
1506     knows that the subject is too short, and so the (*MARK) is never encountered.
1507     In this case, studying the pattern does not affect the overall match result,
1508     which is still "no match", but it does affect the auxiliary information that is
1509     returned.
1510     <pre>
1511 nigel 75 PCRE_NO_UTF8_CHECK
1512     </pre>
1513     When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1514     string is automatically checked when <b>pcre_exec()</b> is subsequently called.
1515     The value of <i>startoffset</i> is also checked to ensure that it points to the
1516 ph10 211 start of a UTF-8 character. There is a discussion about the validity of UTF-8
1517     strings in the
1518     <a href="pcre.html#utf8strings">section on UTF-8 support</a>
1519     in the main
1520     <a href="pcre.html"><b>pcre</b></a>
1521     page. If an invalid UTF-8 sequence of bytes is found, <b>pcre_exec()</b> returns
1522 ph10 572 the error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is
1523     a truncated UTF-8 character at the end of the subject, PCRE_ERROR_SHORTUTF8. If
1524     <i>startoffset</i> contains a value that does not point to the start of a UTF-8
1525     character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1526     returned.
1527 nigel 63 </P>
1528     <P>
1529 nigel 75 If you already know that your subject is valid, and you want to skip these
1530     checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1531     calling <b>pcre_exec()</b>. You might want to do this for the second and
1532     subsequent calls to <b>pcre_exec()</b> if you are making repeated calls to find
1533     all the matches in a single subject string. However, you should be sure that
1534 ph10 567 the value of <i>startoffset</i> points to the start of a UTF-8 character (or the
1535     end of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1536     invalid UTF-8 string as a subject or an invalid value of <i>startoffset</i> is
1537     undefined. Your program may crash.
1538 nigel 75 <pre>
1539 ph10 461 PCRE_PARTIAL_HARD
1540 ph10 429 PCRE_PARTIAL_SOFT
1541 nigel 75 </pre>
1542 ph10 429 These options turn on the partial matching feature. For backwards
1543     compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1544     occurs if the end of the subject string is reached successfully, but there are
1545     not enough subject characters to complete the match. If this happens when
1546 ph10 567 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1547     testing any remaining alternatives. Only if no complete match can be found is
1548     PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1549     PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1550     but only if no complete match can be found.
1551     </P>
1552     <P>
1553     If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1554     partial match is found, <b>pcre_exec()</b> immediately returns
1555     PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1556 ph10 572 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1557 ph10 567 important that an alternative complete match.
1558     </P>
1559     <P>
1560     In both cases, the portion of the string that was inspected when the partial
1561     match was found is set as the first matching string. There is a more detailed
1562     discussion of partial and multi-segment matching, with examples, in the
1563 nigel 75 <a href="pcrepartial.html"><b>pcrepartial</b></a>
1564     documentation.
1565     </P>
1566     <br><b>
1567     The string to be matched by <b>pcre_exec()</b>
1568     </b><br>
1569     <P>
1570 nigel 63 The subject string is passed to <b>pcre_exec()</b> as a pointer in
1571 ph10 371 <i>subject</i>, a length (in bytes) in <i>length</i>, and a starting byte offset
1572 ph10 572 in <i>startoffset</i>. If this is negative or greater than the length of the
1573     subject, <b>pcre_exec()</b> returns PCRE_ERROR_BADOFFSET. When the starting
1574     offset is zero, the search for a match starts at the beginning of the subject,
1575     and this is by far the most common case. In UTF-8 mode, the byte offset must
1576     point to the start of a UTF-8 character (or the end of the subject). Unlike the
1577     pattern string, the subject may contain binary zero bytes.
1578 nigel 63 </P>
1579     <P>
1580     A non-zero starting offset is useful when searching for another match in the
1581     same subject by calling <b>pcre_exec()</b> again after a previous success.
1582     Setting <i>startoffset</i> differs from just passing over a shortened string and
1583     setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1584     lookbehind. For example, consider the pattern
1585     <pre>
1586     \Biss\B
1587 nigel 75 </pre>
1588 nigel 63 which finds occurrences of "iss" in the middle of words. (\B matches only if
1589     the current position in the subject is not a word boundary.) When applied to
1590     the string "Mississipi" the first call to <b>pcre_exec()</b> finds the first
1591     occurrence. If <b>pcre_exec()</b> is called again with just the remainder of the
1592     subject, namely "issipi", it does not match, because \B is always false at the
1593     start of the subject, which is deemed to be a word boundary. However, if
1594     <b>pcre_exec()</b> is passed the entire string again, but with <i>startoffset</i>
1595     set to 4, it finds the second occurrence of "iss" because it is able to look
1596     behind the starting point to discover that it is preceded by a letter.
1597     </P>
1598     <P>
1599 ph10 567 Finding all the matches in a subject is tricky when the pattern can match an
1600     empty string. It is possible to emulate Perl's /g behaviour by first trying the
1601     match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1602     PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1603     and trying an ordinary match again. There is some code that demonstrates how to
1604     do this in the
1605     <a href="pcredemo.html"><b>pcredemo</b></a>
1606 ph10 572 sample program. In the most general case, you have to check to see if the
1607     newline convention recognizes CRLF as a newline, and if so, and the current
1608 ph10 567 character is CR followed by LF, advance the starting offset by two characters
1609     instead of one.
1610     </P>
1611     <P>
1612 nigel 63 If a non-zero starting offset is passed when the pattern is anchored, one
1613 nigel 75 attempt to match at the given offset is made. This can only succeed if the
1614 nigel 63 pattern does not require the match to be at the start of the subject.
1615     </P>
1616 nigel 75 <br><b>
1617     How <b>pcre_exec()</b> returns captured substrings
1618     </b><br>
1619 nigel 63 <P>
1620     In general, a pattern matches a certain portion of the subject, and in
1621     addition, further substrings from the subject may be picked out by parts of the
1622     pattern. Following the usage in Jeffrey Friedl's book, this is called
1623     "capturing" in what follows, and the phrase "capturing subpattern" is used for
1624     a fragment of a pattern that picks out a substring. PCRE supports several other
1625     kinds of parenthesized subpattern that do not cause substrings to be captured.
1626     </P>
1627     <P>
1628 ph10 371 Captured substrings are returned to the caller via a vector of integers whose
1629     address is passed in <i>ovector</i>. The number of elements in the vector is
1630     passed in <i>ovecsize</i>, which must be a non-negative number. <b>Note</b>: this
1631     argument is NOT the size of <i>ovector</i> in bytes.
1632 nigel 63 </P>
1633     <P>
1634 nigel 75 The first two-thirds of the vector is used to pass back captured substrings,
1635     each substring using a pair of integers. The remaining third of the vector is
1636     used as workspace by <b>pcre_exec()</b> while matching capturing subpatterns,
1637 ph10 371 and is not available for passing back information. The number passed in
1638 nigel 75 <i>ovecsize</i> should always be a multiple of three. If it is not, it is
1639     rounded down.
1640     </P>
1641     <P>
1642     When a match is successful, information about captured substrings is returned
1643     in pairs of integers, starting at the beginning of <i>ovector</i>, and
1644 ph10 371 continuing up to two-thirds of its length at the most. The first element of
1645     each pair is set to the byte offset of the first character in a substring, and
1646     the second is set to the byte offset of the first character after the end of a
1647     substring. <b>Note</b>: these values are always byte offsets, even in UTF-8
1648     mode. They are not character counts.
1649 nigel 63 </P>
1650     <P>
1651 ph10 371 The first pair of integers, <i>ovector[0]</i> and <i>ovector[1]</i>, identify the
1652     portion of the subject string matched by the entire pattern. The next pair is
1653     used for the first capturing subpattern, and so on. The value returned by
1654     <b>pcre_exec()</b> is one more than the highest numbered pair that has been set.
1655     For example, if two substrings have been captured, the returned value is 3. If
1656     there are no capturing subpatterns, the return value from a successful match is
1657     1, indicating that just the first pair of offsets has been set.
1658     </P>
1659     <P>
1660 nigel 63 If a capturing subpattern is matched repeatedly, it is the last portion of the
1661 nigel 75 string that it matched that is returned.
1662 nigel 63 </P>
1663     <P>
1664 nigel 75 If the vector is too small to hold all the captured substring offsets, it is
1665     used as far as possible (up to two-thirds of its length), and the function
1666 ph10 371 returns a value of zero. If the substring offsets are not of interest,
1667     <b>pcre_exec()</b> may be called with <i>ovector</i> passed as NULL and
1668 nigel 63 <i>ovecsize</i> as zero. However, if the pattern contains back references and
1669 nigel 75 the <i>ovector</i> is not big enough to remember the related substrings, PCRE
1670     has to get additional memory for use during matching. Thus it is usually
1671     advisable to supply an <i>ovector</i>.
1672 nigel 63 </P>
1673     <P>
1674 ph10 461 The <b>pcre_fullinfo()</b> function can be used to find out how many capturing
1675 nigel 63 subpatterns there are in a compiled pattern. The smallest size for
1676     <i>ovector</i> that will allow for <i>n</i> captured substrings, in addition to
1677     the offsets of the substring matched by the whole pattern, is (<i>n</i>+1)*3.
1678 nigel 91 </P>
1679     <P>
1680     It is possible for capturing subpattern number <i>n+1</i> to match some part of
1681     the subject when subpattern <i>n</i> has not been used at all. For example, if
1682     the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1683     function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1684     happens, both values in the offset pairs corresponding to unused subpatterns
1685     are set to -1.
1686     </P>
1687     <P>
1688     Offset values that correspond to unused subpatterns at the end of the
1689     expression are also set to -1. For example, if the string "abc" is matched
1690     against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1691     return from the function is 2, because the highest used capturing subpattern
1692 ph10 572 number is 1, and the offsets for for the second and third capturing subpatterns
1693     (assuming the vector is large enough, of course) are set to -1.
1694 nigel 91 </P>
1695     <P>
1696 ph10 572 <b>Note</b>: Elements of <i>ovector</i> that do not correspond to capturing
1697     parentheses in the pattern are never changed. That is, if a pattern contains
1698     <i>n</i> capturing parentheses, no more than <i>ovector[0]</i> to
1699     <i>ovector[2n+1]</i> are set by <b>pcre_exec()</b>. The other elements retain
1700     whatever values they previously had.
1701     </P>
1702     <P>
1703 nigel 91 Some convenience functions are provided for extracting the captured substrings
1704     as separate strings. These are described below.
1705 nigel 77 <a name="errorlist"></a></P>
1706 nigel 75 <br><b>
1707 nigel 91 Error return values from <b>pcre_exec()</b>
1708 nigel 75 </b><br>
1709 nigel 63 <P>
1710     If <b>pcre_exec()</b> fails, it returns a negative number. The following are
1711     defined in the header file:
1712     <pre>
1713     PCRE_ERROR_NOMATCH (-1)
1714 nigel 75 </pre>
1715 nigel 63 The subject string did not match the pattern.
1716     <pre>
1717     PCRE_ERROR_NULL (-2)
1718 nigel 75 </pre>
1719 nigel 63 Either <i>code</i> or <i>subject</i> was passed as NULL, or <i>ovector</i> was
1720     NULL and <i>ovecsize</i> was not zero.
1721     <pre>
1723 nigel 75 </pre>
1724 nigel 63 An unrecognized bit was set in the <i>options</i> argument.
1725     <pre>
1727 nigel 75 </pre>
1728 nigel 63 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1729 nigel 75 the case when it is passed a junk pointer and to detect when a pattern that was
1730     compiled in an environment of one endianness is run in an environment with the
1731     other endianness. This is the error that PCRE gives when the magic number is
1732     not present.
1733 nigel 63 <pre>
1734 nigel 93 PCRE_ERROR_UNKNOWN_OPCODE (-5)
1735 nigel 75 </pre>
1736 nigel 63 While running the pattern match, an unknown item was encountered in the
1737     compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1738     of the compiled pattern.
1739     <pre>
1741 nigel 75 </pre>
1742 nigel 63 If a pattern contains back references, but the <i>ovector</i> that is passed to
1743     <b>pcre_exec()</b> is not big enough to remember the referenced substrings, PCRE
1744     gets a block of memory at the start of matching to use for this purpose. If the
1745 nigel 75 call via <b>pcre_malloc()</b> fails, this error is given. The memory is
1746     automatically freed at the end of matching.
1747 ph10 535 </P>
1748     <P>
1749     This error is also given if <b>pcre_stack_malloc()</b> fails in
1750     <b>pcre_exec()</b>. This can happen only when PCRE has been compiled with
1751     <b>--disable-stack-for-recursion</b>.
1752 nigel 63 <pre>
1754 nigel 75 </pre>
1755 nigel 63 This error is used by the <b>pcre_copy_substring()</b>,
1756     <b>pcre_get_substring()</b>, and <b>pcre_get_substring_list()</b> functions (see
1757     below). It is never returned by <b>pcre_exec()</b>.
1758     <pre>
1760 nigel 75 </pre>
1761 nigel 87 The backtracking limit, as specified by the <i>match_limit</i> field in a
1762     <b>pcre_extra</b> structure (or defaulted) was reached. See the description
1763     above.
1764     <pre>
1765 nigel 63 PCRE_ERROR_CALLOUT (-9)
1766 nigel 75 </pre>
1767 nigel 63 This error is never generated by <b>pcre_exec()</b> itself. It is provided for
1768     use by callout functions that want to yield a distinctive error code. See the
1769 nigel 75 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1770     documentation for details.
1771 nigel 71 <pre>
1772 nigel 73 PCRE_ERROR_BADUTF8 (-10)
1773 nigel 75 </pre>
1774 nigel 71 A string that contains an invalid UTF-8 byte sequence was passed as a subject.
1775 ph10 572 However, if PCRE_PARTIAL_HARD is set and the problem is a truncated UTF-8
1776     character at the end of the subject, PCRE_ERROR_SHORTUTF8 is used instead.
1777 nigel 73 <pre>
1779 nigel 75 </pre>
1780 nigel 73 The UTF-8 byte sequence that was passed as a subject was valid, but the value
1781 ph10 572 of <i>startoffset</i> did not point to the beginning of a UTF-8 character or the
1782     end of the subject.
1783 nigel 75 <pre>
1784 nigel 77 PCRE_ERROR_PARTIAL (-12)
1785 nigel 75 </pre>
1786     The subject string did not match, but it did match partially. See the
1787     <a href="pcrepartial.html"><b>pcrepartial</b></a>
1788     documentation for details of partial matching.
1789     <pre>
1790 nigel 77 PCRE_ERROR_BADPARTIAL (-13)
1791 nigel 75 </pre>
1792 ph10 429 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1793     option was used with a compiled pattern containing items that were not
1794 ph10 461 supported for partial matching. From release 8.00 onwards, there are no
1795 ph10 429 restrictions on partial matching.
1796 nigel 75 <pre>
1797 nigel 77 PCRE_ERROR_INTERNAL (-14)
1798 nigel 75 </pre>
1799     An unexpected internal error has occurred. This error could be caused by a bug
1800     in PCRE or by overwriting of the compiled pattern.
1801     <pre>
1802 nigel 77 PCRE_ERROR_BADCOUNT (-15)
1803 nigel 75 </pre>
1804     This error is given if the value of the <i>ovecsize</i> argument is negative.
1805 nigel 93 <pre>
1807     </pre>
1808     The internal recursion limit, as specified by the <i>match_limit_recursion</i>
1809     field in a <b>pcre_extra</b> structure (or defaulted) was reached. See the
1810     description above.
1811     <pre>
1813     </pre>
1814     An invalid combination of PCRE_NEWLINE_<i>xxx</i> options was given.
1815 ph10 567 <pre>
1817     </pre>
1818 ph10 572 The value of <i>startoffset</i> was negative or greater than the length of the
1819 ph10 567 subject, that is, the value in <i>length</i>.
1820 ph10 572 <pre>
1821     PCRE_ERROR_SHORTUTF8 (-25)
1822     </pre>
1823     The subject string ended with an incomplete (truncated) UTF-8 character, and
1824     the PCRE_PARTIAL_HARD option was set. Without this option, PCRE_ERROR_BADUTF8
1825     is returned in this situation.
1826 nigel 73 </P>
1827 nigel 93 <P>
1828 ph10 197 Error numbers -16 to -20 and -22 are not used by <b>pcre_exec()</b>.
1829 nigel 93 </P>
1830 nigel 91 <br><a name="SEC15" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NUMBER</a><br>
1831 nigel 63 <P>
1832     <b>int pcre_copy_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
1833     <b>int <i>stringcount</i>, int <i>stringnumber</i>, char *<i>buffer</i>,</b>
1834     <b>int <i>buffersize</i>);</b>
1835     </P>
1836     <P>
1837     <b>int pcre_get_substring(const char *<i>subject</i>, int *<i>ovector</i>,</b>
1838     <b>int <i>stringcount</i>, int <i>stringnumber</i>,</b>
1839     <b>const char **<i>stringptr</i>);</b>
1840     </P>
1841     <P>
1842     <b>int pcre_get_substring_list(const char *<i>subject</i>,</b>
1843     <b>int *<i>ovector</i>, int <i>stringcount</i>, const char ***<i>listptr</i>);</b>
1844     </P>
1845     <P>
1846     Captured substrings can be accessed directly by using the offsets returned by
1847     <b>pcre_exec()</b> in <i>ovector</i>. For convenience, the functions
1848     <b>pcre_copy_substring()</b>, <b>pcre_get_substring()</b>, and
1849     <b>pcre_get_substring_list()</b> are provided for extracting captured substrings
1850     as new, separate, zero-terminated strings. These functions identify substrings
1851     by number. The next section describes functions for extracting named
1852 nigel 91 substrings.
1853 nigel 63 </P>
1854     <P>
1855 nigel 91 A substring that contains a binary zero is correctly extracted and has a
1856     further zero added on the end, but the result is not, of course, a C string.
1857     However, you can process such a string by referring to the length that is
1858     returned by <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>.
1859     Unfortunately, the interface to <b>pcre_get_substring_list()</b> is not adequate
1860     for handling strings containing binary zeros, because the end of the final
1861     string is not independently indicated.
1862     </P>
1863     <P>
1864 nigel 63 The first three arguments are the same for all three of these functions:
1865 nigel 75 <i>subject</i> is the subject string that has just been successfully matched,
1866 nigel 63 <i>ovector</i> is a pointer to the vector of integer offsets that was passed to
1867     <b>pcre_exec()</b>, and <i>stringcount</i> is the number of substrings that were
1868     captured by the match, including the substring that matched the entire regular
1869 nigel 75 expression. This is the value returned by <b>pcre_exec()</b> if it is greater
1870     than zero. If <b>pcre_exec()</b> returned zero, indicating that it ran out of
1871     space in <i>ovector</i>, the value passed as <i>stringcount</i> should be the
1872     number of elements in the vector divided by three.
1873 nigel 63 </P>
1874     <P>
1875     The functions <b>pcre_copy_substring()</b> and <b>pcre_get_substring()</b>
1876     extract a single substring, whose number is given as <i>stringnumber</i>. A
1877 nigel 75 value of zero extracts the substring that matched the entire pattern, whereas
1878 nigel 63 higher values extract the captured substrings. For <b>pcre_copy_substring()</b>,
1879     the string is placed in <i>buffer</i>, whose length is given by
1880     <i>buffersize</i>, while for <b>pcre_get_substring()</b> a new block of memory is
1881     obtained via <b>pcre_malloc</b>, and its address is returned via
1882     <i>stringptr</i>. The yield of the function is the length of the string, not
1883 nigel 93 including the terminating zero, or one of these error codes:
1884 nigel 63 <pre>
1886 nigel 75 </pre>
1887 nigel 63 The buffer was too small for <b>pcre_copy_substring()</b>, or the attempt to get
1888     memory failed for <b>pcre_get_substring()</b>.
1889     <pre>
1891 nigel 75 </pre>
1892 nigel 63 There is no substring whose number is <i>stringnumber</i>.
1893     </P>
1894     <P>
1895     The <b>pcre_get_substring_list()</b> function extracts all available substrings
1896     and builds a list of pointers to them. All this is done in a single block of
1897 nigel 75 memory that is obtained via <b>pcre_malloc</b>. The address of the memory block
1898 nigel 63 is returned via <i>listptr</i>, which is also the start of the list of string
1899     pointers. The end of the list is marked by a NULL pointer. The yield of the
1900 nigel 93 function is zero if all went well, or the error code
1901 nigel 63 <pre>
1903 nigel 75 </pre>
1904 nigel 63 if the attempt to get the memory block failed.
1905     </P>
1906     <P>
1907     When any of these functions encounter a substring that is unset, which can
1908     happen when capturing subpattern number <i>n+1</i> matches some part of the
1909     subject, but subpattern <i>n</i> has not been used at all, they return an empty
1910     string. This can be distinguished from a genuine zero-length substring by
1911     inspecting the appropriate offset in <i>ovector</i>, which is negative for unset
1912     substrings.
1913     </P>
1914     <P>
1915     The two convenience functions <b>pcre_free_substring()</b> and
1916     <b>pcre_free_substring_list()</b> can be used to free the memory returned by
1917     a previous call of <b>pcre_get_substring()</b> or
1918     <b>pcre_get_substring_list()</b>, respectively. They do nothing more than call
1919     the function pointed to by <b>pcre_free</b>, which of course could be called
1920     directly from a C program. However, PCRE is used in some situations where it is
1921 nigel 91 linked via a special interface to another programming language that cannot use
1922 nigel 63 <b>pcre_free</b> directly; it is for these cases that the functions are
1923     provided.
1924     </P>
1925 nigel 91 <br><a name="SEC16" href="#TOC1">EXTRACTING CAPTURED SUBSTRINGS BY NAME</a><br>
1926 nigel 63 <P>
1927 nigel 75 <b>int pcre_get_stringnumber(const pcre *<i>code</i>,</b>
1928     <b>const char *<i>name</i>);</b>
1929     </P>
1930     <P>
1931 nigel 63 <b>int pcre_copy_named_substring(const pcre *<i>code</i>,</b>
1932     <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
1933     <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
1934     <b>char *<i>buffer</i>, int <i>buffersize</i>);</b>
1935     </P>
1936     <P>
1937     <b>int pcre_get_named_substring(const pcre *<i>code</i>,</b>
1938     <b>const char *<i>subject</i>, int *<i>ovector</i>,</b>
1939     <b>int <i>stringcount</i>, const char *<i>stringname</i>,</b>
1940     <b>const char **<i>stringptr</i>);</b>
1941     </P>
1942     <P>
1943 nigel 75 To extract a substring by name, you first have to find associated number.
1944     For example, for this pattern
1945 nigel 63 <pre>
1946 nigel 93 (a+)b(?&#60;xxx&#62;\d+)...
1947 nigel 75 </pre>
1948 nigel 91 the number of the subpattern called "xxx" is 2. If the name is known to be
1949     unique (PCRE_DUPNAMES was not set), you can find the number from the name by
1950     calling <b>pcre_get_stringnumber()</b>. The first argument is the compiled
1951     pattern, and the second is the name. The yield of the function is the
1952 nigel 75 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
1953     that name.
1954 nigel 63 </P>
1955     <P>
1956 nigel 75 Given the number, you can extract the substring directly, or use one of the
1957     functions described in the previous section. For convenience, there are also
1958     two functions that do the whole job.
1959 nigel 63 </P>
1960     <P>
1961 nigel 91 Most of the arguments of <b>pcre_copy_named_substring()</b> and
1962     <b>pcre_get_named_substring()</b> are the same as those for the similarly named
1963 nigel 75 functions that extract by number. As these are described in the previous
1964     section, they are not re-described here. There are just two differences:
1965 nigel 63 </P>
1966     <P>
1967     First, instead of a substring number, a substring name is given. Second, there
1968     is an extra argument, given at the start, which is a pointer to the compiled
1969     pattern. This is needed in order to gain access to the name-to-number
1970     translation table.
1971     </P>
1972     <P>
1973     These functions call <b>pcre_get_stringnumber()</b>, and if it succeeds, they
1974 ph10 128 then call <b>pcre_copy_substring()</b> or <b>pcre_get_substring()</b>, as
1975     appropriate. <b>NOTE:</b> If PCRE_DUPNAMES is set and there are duplicate names,
1976     the behaviour may not be what you want (see the next section).
1977 nigel 63 </P>
1978 ph10 392 <P>
1979 ph10 461 <b>Warning:</b> If the pattern uses the (?| feature to set up multiple
1980     subpatterns with the same number, as described in the
1981     <a href="pcrepattern.html#dupsubpatternnumber">section on duplicate subpattern numbers</a>
1982     in the
1983     <a href="pcrepattern.html"><b>pcrepattern</b></a>
1984     page, you cannot use names to distinguish the different subpatterns, because
1985     names are not included in the compiled code. The matching process uses only
1986     numbers. For this reason, the use of different names for subpatterns of the
1987     same number causes an error at compile time.
1988 ph10 392 </P>
1989 nigel 91 <br><a name="SEC17" href="#TOC1">DUPLICATE SUBPATTERN NAMES</a><br>
1990 nigel 63 <P>
1991 nigel 91 <b>int pcre_get_stringtable_entries(const pcre *<i>code</i>,</b>
1992     <b>const char *<i>name</i>, char **<i>first</i>, char **<i>last</i>);</b>
1993     </P>
1994     <P>
1995     When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
1996 ph10 461 are not required to be unique. (Duplicate names are always allowed for
1997     subpatterns with the same number, created by using the (?| feature. Indeed, if
1998     such subpatterns are named, they are required to use the same names.)
1999     </P>
2000     <P>
2001     Normally, patterns with duplicate names are such that in any one match, only
2002     one of the named subpatterns participates. An example is shown in the
2003 nigel 91 <a href="pcrepattern.html"><b>pcrepattern</b></a>
2004 ph10 208 documentation.
2005     </P>
2006     <P>
2007     When duplicates are present, <b>pcre_copy_named_substring()</b> and
2008     <b>pcre_get_named_substring()</b> return the first substring corresponding to
2009     the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2010     returned; no data is returned. The <b>pcre_get_stringnumber()</b> function
2011     returns one of the numbers that are associated with the name, but it is not
2012     defined which it is.
2013     </P>
2014     <P>
2015 nigel 91 If you want to get full details of all captured substrings for a given name,
2016     you must use the <b>pcre_get_stringtable_entries()</b> function. The first
2017     argument is the compiled pattern, and the second is the name. The third and
2018     fourth are pointers to variables which are updated by the function. After it
2019     has run, they point to the first and last entries in the name-to-number table
2020     for the given name. The function itself returns the length of each entry, or
2021 nigel 93 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2022     described above in the section entitled <i>Information about a pattern</i>.
2023     Given all the relevant entries for the name, you can extract each of their
2024     numbers, and hence the captured data, if any.
2025 nigel 91 </P>
2026     <br><a name="SEC18" href="#TOC1">FINDING ALL POSSIBLE MATCHES</a><br>
2027     <P>
2028 nigel 77 The traditional matching function uses a similar algorithm to Perl, which stops
2029     when it finds the first match, starting at a given point in the subject. If you
2030     want to find all possible matches, or the longest possible match, consider
2031     using the alternative matching function (see below) instead. If you cannot use
2032     the alternative function, but still need to find all possible matches, you
2033     can kludge it up by making use of the callout facility, which is described in
2034     the
2035     <a href="pcrecallout.html"><b>pcrecallout</b></a>
2036     documentation.
2037     </P>
2038     <P>
2039     What you have to do is to insert a callout right at the end of the pattern.
2040     When your callout function is called, extract and save the current matched
2041     substring. Then return 1, which forces <b>pcre_exec()</b> to backtrack and try
2042     other alternatives. Ultimately, when it runs out of matches, <b>pcre_exec()</b>
2043     will yield PCRE_ERROR_NOMATCH.
2044     <a name="dfamatch"></a></P>
2045 nigel 91 <br><a name="SEC19" href="#TOC1">MATCHING A PATTERN: THE ALTERNATIVE FUNCTION</a><br>
2046 nigel 77 <P>
2047     <b>int pcre_dfa_exec(const pcre *<i>code</i>, const pcre_extra *<i>extra</i>,</b>
2048     <b>const char *<i>subject</i>, int <i>length</i>, int <i>startoffset</i>,</b>
2049     <b>int <i>options</i>, int *<i>ovector</i>, int <i>ovecsize</i>,</b>
2050     <b>int *<i>workspace</i>, int <i>wscount</i>);</b>
2051     </P>
2052     <P>
2053     The function <b>pcre_dfa_exec()</b> is called to match a subject string against
2054 nigel 93 a compiled pattern, using a matching algorithm that scans the subject string
2055     just once, and does not backtrack. This has different characteristics to the
2056     normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2057     patterns are not supported. Nevertheless, there are times when this kind of
2058 ph10 461 matching can be useful. For a discussion of the two matching algorithms, and a
2059 ph10 453 list of features that <b>pcre_dfa_exec()</b> does not support, see the
2060 nigel 77 <a href="pcrematching.html"><b>pcrematching</b></a>
2061     documentation.
2062     </P>
2063     <P>
2064     The arguments for the <b>pcre_dfa_exec()</b> function are the same as for
2065     <b>pcre_exec()</b>, plus two extras. The <i>ovector</i> argument is used in a
2066     different way, and this is described below. The other common arguments are used
2067     in the same way as for <b>pcre_exec()</b>, so their description is not repeated
2068     here.
2069     </P>
2070     <P>
2071     The two additional arguments provide workspace for the function. The workspace
2072     vector should contain at least 20 elements. It is used for keeping track of
2073     multiple paths through the pattern tree. More workspace will be needed for
2074 nigel 91 patterns and subjects where there are a lot of potential matches.
2075 nigel 77 </P>
2076     <P>
2077 nigel 87 Here is an example of a simple call to <b>pcre_dfa_exec()</b>:
2078 nigel 77 <pre>
2079     int rc;
2080     int ovector[10];
2081     int wspace[20];
2082 nigel 87 rc = pcre_dfa_exec(
2083 nigel 77 re, /* result of pcre_compile() */
2084     NULL, /* we didn't study the pattern */
2085     "some string", /* the subject string */
2086     11, /* the length of the subject string */
2087     0, /* start at offset 0 in the subject */
2088     0, /* default options */
2089     ovector, /* vector of integers for substring information */
2090     10, /* number of elements (NOT size in bytes) */
2091     wspace, /* working space vector */
2092     20); /* number of elements (NOT size in bytes) */
2093     </PRE>
2094     </P>
2095     <br><b>
2096     Option bits for <b>pcre_dfa_exec()</b>
2097     </b><br>
2098     <P>
2099     The unused bits of the <i>options</i> argument for <b>pcre_dfa_exec()</b> must be
2100 nigel 91 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_<i>xxx</i>,
2104     All but the last four of these are exactly the same as for <b>pcre_exec()</b>,
2105     so their description is not repeated here.
2106 nigel 77 <pre>
2107 ph10 429 PCRE_PARTIAL_HARD
2108 ph10 461 PCRE_PARTIAL_SOFT
2109 nigel 77 </pre>
2110 ph10 429 These have the same general effect as they do for <b>pcre_exec()</b>, but the
2111     details are slightly different. When PCRE_PARTIAL_HARD is set for
2112     <b>pcre_dfa_exec()</b>, it returns PCRE_ERROR_PARTIAL if the end of the subject
2113     is reached and there is still at least one matching possibility that requires
2114     additional characters. This happens even if some complete matches have also
2115     been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2116     is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2117     there have been no complete matches, but there is still at least one matching
2118 ph10 453 possibility. The portion of the string that was inspected when the longest
2119     partial match was found is set as the first matching string in both cases.
2120 ph10 567 There is a more detailed discussion of partial and multi-segment matching, with
2121     examples, in the
2122     <a href="pcrepartial.html"><b>pcrepartial</b></a>
2123     documentation.
2124 nigel 77 <pre>
2126     </pre>
2127     Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2128 nigel 93 soon as it has found one match. Because of the way the alternative algorithm
2129     works, this is necessarily the shortest possible match at the first possible
2130     matching point in the subject string.
2131 nigel 77 <pre>
2133     </pre>
2134 ph10 429 When <b>pcre_dfa_exec()</b> returns a partial match, it is possible to call it
2135     again, with additional subject characters, and have it continue with the same
2136     match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2137     <i>workspace</i> and <i>wscount</i> options must reference the same vector as
2138     before because data about the match so far is left in them after a partial
2139     match. There is more discussion of this facility in the
2140 nigel 77 <a href="pcrepartial.html"><b>pcrepartial</b></a>
2141     documentation.
2142     </P>
2143     <br><b>
2144     Successful returns from <b>pcre_dfa_exec()</b>
2145     </b><br>
2146     <P>
2147     When <b>pcre_dfa_exec()</b> succeeds, it may have matched more than one
2148     substring in the subject. Note, however, that all the matches from one run of
2149     the function start at the same point in the subject. The shorter matches are
2150     all initial substrings of the longer matches. For example, if the pattern
2151     <pre>
2152     &#60;.*&#62;
2153     </pre>
2154     is matched against the string
2155     <pre>
2156     This is &#60;something&#62; &#60;something else&#62; &#60;something further&#62; no more
2157     </pre>
2158     the three matched strings are
2159     <pre>
2160     &#60;something&#62;
2161     &#60;something&#62; &#60;something else&#62;
2162     &#60;something&#62; &#60;something else&#62; &#60;something further&#62;
2163     </pre>
2164     On success, the yield of the function is a number greater than zero, which is
2165     the number of matched substrings. The substrings themselves are returned in
2166     <i>ovector</i>. Each string uses two elements; the first is the offset to the
2167 nigel 93 start, and the second is the offset to the end. In fact, all the strings have
2168     the same start offset. (Space could have been saved by giving this only once,
2169     but it was decided to retain some compatibility with the way <b>pcre_exec()</b>
2170     returns data, even though the meaning of the strings is different.)
2171 nigel 77 </P>
2172     <P>
2173     The strings are returned in reverse order of length; that is, the longest
2174     matching string is given first. If there were too many matches to fit into
2175     <i>ovector</i>, the yield of the function is zero, and the vector is filled with
2176     the longest matches.
2177     </P>
2178     <br><b>
2179     Error returns from <b>pcre_dfa_exec()</b>
2180     </b><br>
2181     <P>
2182     The <b>pcre_dfa_exec()</b> function returns a negative number when it fails.
2183     Many of the errors are the same as for <b>pcre_exec()</b>, and these are
2184     described
2185     <a href="#errorlist">above.</a>
2186     There are in addition the following errors that are specific to
2187     <b>pcre_dfa_exec()</b>:
2188     <pre>
2189     PCRE_ERROR_DFA_UITEM (-16)
2190     </pre>
2191     This return is given if <b>pcre_dfa_exec()</b> encounters an item in the pattern
2192     that it does not support, for instance, the use of \C or a back reference.
2193     <pre>
2194     PCRE_ERROR_DFA_UCOND (-17)
2195     </pre>
2196 nigel 93 This return is given if <b>pcre_dfa_exec()</b> encounters a condition item that
2197     uses a back reference for the condition, or a test for recursion in a specific
2198     group. These are not supported.
2199 nigel 77 <pre>
2201     </pre>
2202     This return is given if <b>pcre_dfa_exec()</b> is called with an <i>extra</i>
2203     block that contains a setting of the <i>match_limit</i> field. This is not
2204     supported (it is meaningless).
2205     <pre>
2206     PCRE_ERROR_DFA_WSSIZE (-19)
2207     </pre>
2208     This return is given if <b>pcre_dfa_exec()</b> runs out of space in the
2209     <i>workspace</i> vector.
2210     <pre>
2212     </pre>
2213     When a recursive subpattern is processed, the matching function calls itself
2214     recursively, using private vectors for <i>ovector</i> and <i>workspace</i>. This
2215     error is given if the output vector is not large enough. This should be
2216     extremely rare, as a vector of size 1000 is used.
2217     </P>
2218 nigel 93 <br><a name="SEC20" href="#TOC1">SEE ALSO</a><br>
2219 nigel 77 <P>
2220 nigel 93 <b>pcrebuild</b>(3), <b>pcrecallout</b>(3), <b>pcrecpp(3)</b>(3),
2221     <b>pcrematching</b>(3), <b>pcrepartial</b>(3), <b>pcreposix</b>(3),
2222     <b>pcreprecompile</b>(3), <b>pcresample</b>(3), <b>pcrestack</b>(3).
2223     </P>
2224 ph10 99 <br><a name="SEC21" href="#TOC1">AUTHOR</a><br>
2225 nigel 93 <P>
2226 ph10 99 Philip Hazel
2227 nigel 63 <br>
2228 ph10 99 University Computing Service
2229     <br>
2230     Cambridge CB2 3QH, England.
2231     <br>
2232     </P>
2233     <br><a name="SEC22" href="#TOC1">REVISION</a><br>
2234     <P>
2235 ph10 572 Last updated: 13 November 2010
2236 ph10 99 <br>
2237 ph10 512 Copyright &copy; 1997-2010 University of Cambridge.
2238 ph10 99 <br>
2239 nigel 75 <p>
2240     Return to the <a href="index.html">PCRE index page</a>.
2241     </p>


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

ViewVC Help
Powered by ViewVC 1.1.12