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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 678 - (hide annotations) (download) (as text)
Sun Aug 28 15:23:03 2011 UTC (2 years, 11 months ago) by ph10
File MIME type: text/html
File size: 108112 byte(s)
Documentation for JIT support.

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

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12