/[pcre]/code/trunk/doc/pcreapi.3
ViewVC logotype

Contents of /code/trunk/doc/pcreapi.3

Parent Directory Parent Directory | Revision Log Revision Log


Revision 455 - (hide annotations) (download)
Sat Sep 26 19:12:32 2009 UTC (4 years, 6 months ago) by ph10
File size: 88755 byte(s)
Added lower bound length-finding to pcre_study() and use it when matching; make 
the value available via pcre_fullinfo(); also fixed bugs connected with
pcre_study() in pcre_dfa_exec(). 

1 nigel 79 .TH PCREAPI 3
2 nigel 63 .SH NAME
3     PCRE - Perl-compatible regular expressions
4 nigel 75 .SH "PCRE NATIVE API"
5 nigel 63 .rs
6     .sp
7     .B #include <pcre.h>
8     .PP
9     .SM
10 nigel 75 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
11 nigel 63 .ti +5n
12 nigel 75 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
13 nigel 63 .ti +5n
14 nigel 75 .B const unsigned char *\fItableptr\fP);
15 nigel 63 .PP
16 nigel 77 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
17     .ti +5n
18     .B int *\fIerrorcodeptr\fP,
19     .ti +5n
20     .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
21     .ti +5n
22     .B const unsigned char *\fItableptr\fP);
23     .PP
24 nigel 75 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
25 nigel 63 .ti +5n
26 nigel 75 .B const char **\fIerrptr\fP);
27 nigel 63 .PP
28 nigel 75 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
29 nigel 63 .ti +5n
30 nigel 75 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
31 nigel 63 .ti +5n
32 nigel 75 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
33 nigel 63 .PP
34 nigel 77 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
35     .ti +5n
36     .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
37     .ti +5n
38     .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
39     .ti +5n
40     .B int *\fIworkspace\fP, int \fIwscount\fP);
41     .PP
42 nigel 75 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
43 nigel 63 .ti +5n
44 nigel 75 .B const char *\fIsubject\fP, int *\fIovector\fP,
45 nigel 63 .ti +5n
46 nigel 75 .B int \fIstringcount\fP, const char *\fIstringname\fP,
47 nigel 63 .ti +5n
48 nigel 75 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
49 nigel 63 .PP
50 nigel 75 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
51 nigel 63 .ti +5n
52 nigel 75 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
53 nigel 63 .ti +5n
54 nigel 75 .B int \fIbuffersize\fP);
55 nigel 63 .PP
56 nigel 75 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
57 nigel 63 .ti +5n
58 nigel 75 .B const char *\fIsubject\fP, int *\fIovector\fP,
59 nigel 63 .ti +5n
60 nigel 75 .B int \fIstringcount\fP, const char *\fIstringname\fP,
61 nigel 63 .ti +5n
62 nigel 75 .B const char **\fIstringptr\fP);
63 nigel 63 .PP
64 nigel 75 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
65 nigel 63 .ti +5n
66 nigel 75 .B const char *\fIname\fP);
67 nigel 63 .PP
68 nigel 91 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
69     .ti +5n
70     .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
71     .PP
72 nigel 75 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
73 nigel 63 .ti +5n
74 nigel 75 .B int \fIstringcount\fP, int \fIstringnumber\fP,
75 nigel 63 .ti +5n
76 nigel 75 .B const char **\fIstringptr\fP);
77 nigel 63 .PP
78 nigel 75 .B int pcre_get_substring_list(const char *\fIsubject\fP,
79 nigel 63 .ti +5n
80 nigel 75 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
81 nigel 63 .PP
82 nigel 75 .B void pcre_free_substring(const char *\fIstringptr\fP);
83 nigel 63 .PP
84 nigel 75 .B void pcre_free_substring_list(const char **\fIstringptr\fP);
85 nigel 63 .PP
86     .B const unsigned char *pcre_maketables(void);
87     .PP
88 nigel 75 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
89 nigel 63 .ti +5n
90 nigel 75 .B int \fIwhat\fP, void *\fIwhere\fP);
91 nigel 63 .PP
92 nigel 75 .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
93     .B *\fIfirstcharptr\fP);
94 nigel 63 .PP
95 nigel 77 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
96     .PP
97 nigel 75 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
98 nigel 63 .PP
99     .B char *pcre_version(void);
100     .PP
101     .B void *(*pcre_malloc)(size_t);
102     .PP
103     .B void (*pcre_free)(void *);
104     .PP
105 nigel 73 .B void *(*pcre_stack_malloc)(size_t);
106     .PP
107     .B void (*pcre_stack_free)(void *);
108     .PP
109 nigel 63 .B int (*pcre_callout)(pcre_callout_block *);
110 nigel 75 .
111     .
112     .SH "PCRE API OVERVIEW"
113 nigel 63 .rs
114     .sp
115 nigel 93 PCRE has its own native API, which is described in this document. There are
116     also some wrapper functions that correspond to the POSIX regular expression
117 nigel 77 API. These are described in the
118 nigel 75 .\" HREF
119     \fBpcreposix\fP
120     .\"
121 nigel 77 documentation. Both of these APIs define a set of C function calls. A C++
122     wrapper is distributed with PCRE. It is documented in the
123     .\" HREF
124     \fBpcrecpp\fP
125     .\"
126     page.
127 nigel 75 .P
128 nigel 77 The native API C function prototypes are defined in the header file
129     \fBpcre.h\fP, and on Unix systems the library itself is called \fBlibpcre\fP.
130     It can normally be accessed by adding \fB-lpcre\fP to the command for linking
131     an application that uses PCRE. The header file defines the macros PCRE_MAJOR
132     and PCRE_MINOR to contain the major and minor release numbers for the library.
133 nigel 75 Applications can use these to include support for different releases of PCRE.
134     .P
135 nigel 77 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
136     and \fBpcre_exec()\fP are used for compiling and matching regular expressions
137     in a Perl-compatible manner. A sample program that demonstrates the simplest
138 ph10 429 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
139     source distribution. A listing of this program is given in the
140 nigel 75 .\" HREF
141 ph10 429 \fBpcredemo\fP
142     .\"
143     documentation, and the
144     .\" HREF
145 nigel 75 \fBpcresample\fP
146     .\"
147 ph10 312 documentation describes how to compile and run it.
148 nigel 75 .P
149 nigel 77 A second matching function, \fBpcre_dfa_exec()\fP, which is not
150     Perl-compatible, is also provided. This uses a different algorithm for the
151 nigel 91 matching. The alternative algorithm finds all possible matches (at a given
152 ph10 435 point in the subject), and scans the subject just once (unless there are
153     lookbehind assertions). However, this algorithm does not return captured
154     substrings. A description of the two matching algorithms and their advantages
155     and disadvantages is given in the
156 nigel 77 .\" HREF
157     \fBpcrematching\fP
158     .\"
159     documentation.
160     .P
161 nigel 75 In addition to the main compiling and matching functions, there are convenience
162 nigel 77 functions for extracting captured substrings from a subject string that is
163     matched by \fBpcre_exec()\fP. They are:
164 nigel 75 .sp
165     \fBpcre_copy_substring()\fP
166     \fBpcre_copy_named_substring()\fP
167     \fBpcre_get_substring()\fP
168     \fBpcre_get_named_substring()\fP
169     \fBpcre_get_substring_list()\fP
170     \fBpcre_get_stringnumber()\fP
171 nigel 91 \fBpcre_get_stringtable_entries()\fP
172 nigel 75 .sp
173     \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
174 nigel 63 provided, to free the memory used for extracted strings.
175 nigel 75 .P
176     The function \fBpcre_maketables()\fP is used to build a set of character tables
177 nigel 77 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
178     or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
179     specialist use. Most commonly, no special tables are passed, in which case
180     internal tables that are generated when PCRE is built are used.
181 nigel 75 .P
182     The function \fBpcre_fullinfo()\fP is used to find out information about a
183     compiled pattern; \fBpcre_info()\fP is an obsolete version that returns only
184 nigel 63 some of the available information, but is retained for backwards compatibility.
185 nigel 75 The function \fBpcre_version()\fP returns a pointer to a string containing the
186 nigel 63 version of PCRE and its date of release.
187 nigel 75 .P
188 nigel 77 The function \fBpcre_refcount()\fP maintains a reference count in a data block
189     containing a compiled pattern. This is provided for the benefit of
190     object-oriented applications.
191     .P
192 nigel 75 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
193     the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
194 nigel 63 respectively. PCRE calls the memory management functions via these variables,
195     so a calling program can replace them if it wishes to intercept the calls. This
196     should be done before calling any PCRE functions.
197 nigel 75 .P
198     The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
199 nigel 73 indirections to memory management functions. These special functions are used
200     only when PCRE is compiled to use the heap for remembering data, instead of
201 nigel 91 recursive function calls, when running the \fBpcre_exec()\fP function. See the
202     .\" HREF
203     \fBpcrebuild\fP
204     .\"
205     documentation for details of how to do this. It is a non-standard way of
206     building PCRE, for use in environments that have limited stacks. Because of the
207     greater use of memory management, it runs more slowly. Separate functions are
208     provided so that special-purpose external code can be used for this case. When
209     used, these functions are always called in a stack-like manner (last obtained,
210     first freed), and always for memory blocks of the same size. There is a
211     discussion about PCRE's stack usage in the
212     .\" HREF
213     \fBpcrestack\fP
214     .\"
215     documentation.
216 nigel 75 .P
217     The global variable \fBpcre_callout\fP initially contains NULL. It can be set
218 nigel 63 by the caller to a "callout" function, which PCRE will then call at specified
219 nigel 75 points during a matching operation. Details are given in the
220     .\" HREF
221     \fBpcrecallout\fP
222     .\"
223 nigel 63 documentation.
224 nigel 75 .
225     .
226 ph10 227 .\" HTML <a name="newlines"></a>
227 nigel 91 .SH NEWLINES
228 nigel 93 .rs
229 nigel 91 .sp
230 ph10 149 PCRE supports five different conventions for indicating line breaks in
231 nigel 93 strings: a single CR (carriage return) character, a single LF (linefeed)
232 ph10 149 character, the two-character sequence CRLF, any of the three preceding, or any
233     Unicode newline sequence. The Unicode newline sequences are the three just
234     mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
235     U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
236     (paragraph separator, U+2029).
237 nigel 93 .P
238     Each of the first three conventions is used by at least one operating system as
239     its standard newline sequence. When PCRE is built, a default can be specified.
240     The default default is LF, which is the Unix standard. When PCRE is run, the
241     default can be overridden, either when a pattern is compiled, or when it is
242     matched.
243     .P
244 ph10 227 At compile time, the newline convention can be specified by the \fIoptions\fP
245     argument of \fBpcre_compile()\fP, or it can be specified by special text at the
246     start of the pattern itself; this overrides any other settings. See the
247     .\" HREF
248     \fBpcrepattern\fP
249     .\"
250     page for details of the special character sequences.
251     .P
252 nigel 91 In the PCRE documentation the word "newline" is used to mean "the character or
253 nigel 93 pair of characters that indicate a line break". The choice of newline
254     convention affects the handling of the dot, circumflex, and dollar
255     metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
256     recognized line ending sequence, the match position advancement for a
257 ph10 226 non-anchored pattern. There is more detail about this in the
258     .\" HTML <a href="#execoptions">
259     .\" </a>
260     section on \fBpcre_exec()\fP options
261     .\"
262 ph10 231 below.
263     .P
264     The choice of newline convention does not affect the interpretation of
265     the \en or \er escape sequences, nor does it affect what \eR matches, which is
266     controlled in a similar way, but by separate options.
267 nigel 91 .
268     .
269 nigel 63 .SH MULTITHREADING
270     .rs
271     .sp
272     The PCRE functions can be used in multi-threading applications, with the
273 nigel 75 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
274     \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
275     callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
276     .P
277 nigel 63 The compiled form of a regular expression is not altered during matching, so
278     the same compiled pattern can safely be used by several threads at once.
279 nigel 75 .
280     .
281     .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
282 nigel 63 .rs
283     .sp
284 nigel 75 The compiled form of a regular expression can be saved and re-used at a later
285     time, possibly by a different program, and even on a host other than the one on
286     which it was compiled. Details are given in the
287     .\" HREF
288     \fBpcreprecompile\fP
289     .\"
290 ph10 155 documentation. However, compiling a regular expression with one version of PCRE
291     for use with a different version is not guaranteed to work and may cause
292     crashes.
293 nigel 75 .
294     .
295     .SH "CHECKING BUILD-TIME OPTIONS"
296     .rs
297     .sp
298     .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
299 nigel 63 .PP
300 nigel 75 The function \fBpcre_config()\fP makes it possible for a PCRE client to
301 nigel 63 discover which optional features have been compiled into the PCRE library. The
302     .\" HREF
303 nigel 75 \fBpcrebuild\fP
304 nigel 63 .\"
305     documentation has more details about these optional features.
306 nigel 75 .P
307     The first argument for \fBpcre_config()\fP is an integer, specifying which
308 nigel 63 information is required; the second argument is a pointer to a variable into
309     which the information is placed. The following information is available:
310 nigel 75 .sp
311 nigel 63 PCRE_CONFIG_UTF8
312 nigel 75 .sp
313 nigel 63 The output is an integer that is set to one if UTF-8 support is available;
314     otherwise it is set to zero.
315 nigel 75 .sp
316     PCRE_CONFIG_UNICODE_PROPERTIES
317     .sp
318     The output is an integer that is set to one if support for Unicode character
319     properties is available; otherwise it is set to zero.
320     .sp
321 nigel 63 PCRE_CONFIG_NEWLINE
322 nigel 75 .sp
323 nigel 91 The output is an integer whose value specifies the default character sequence
324 nigel 93 that is recognized as meaning "newline". The four values that are supported
325 ph10 392 are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF, and -1 for ANY.
326 ph10 391 Though they are derived from ASCII, the same values are returned in EBCDIC
327     environments. The default should normally correspond to the standard sequence
328     for your operating system.
329 nigel 75 .sp
330 ph10 231 PCRE_CONFIG_BSR
331     .sp
332     The output is an integer whose value indicates what character sequences the \eR
333     escape sequence matches by default. A value of 0 means that \eR matches any
334     Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
335     or CRLF. The default can be overridden when a pattern is compiled or matched.
336     .sp
337 nigel 63 PCRE_CONFIG_LINK_SIZE
338 nigel 75 .sp
339 nigel 63 The output is an integer that contains the number of bytes used for internal
340     linkage in compiled regular expressions. The value is 2, 3, or 4. Larger values
341     allow larger regular expressions to be compiled, at the expense of slower
342     matching. The default value of 2 is sufficient for all but the most massive
343     patterns, since it allows the compiled pattern to be up to 64K in size.
344 nigel 75 .sp
345 nigel 63 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
346 nigel 75 .sp
347 nigel 63 The output is an integer that contains the threshold above which the POSIX
348 nigel 75 interface uses \fBmalloc()\fP for output vectors. Further details are given in
349     the
350     .\" HREF
351     \fBpcreposix\fP
352     .\"
353     documentation.
354     .sp
355 nigel 63 PCRE_CONFIG_MATCH_LIMIT
356 nigel 75 .sp
357 ph10 376 The output is a long integer that gives the default limit for the number of
358 nigel 75 internal matching function calls in a \fBpcre_exec()\fP execution. Further
359     details are given with \fBpcre_exec()\fP below.
360     .sp
361 nigel 87 PCRE_CONFIG_MATCH_LIMIT_RECURSION
362     .sp
363 ph10 376 The output is a long integer that gives the default limit for the depth of
364 nigel 87 recursion when calling the internal matching function in a \fBpcre_exec()\fP
365     execution. Further details are given with \fBpcre_exec()\fP below.
366     .sp
367 nigel 73 PCRE_CONFIG_STACKRECURSE
368 nigel 75 .sp
369 nigel 77 The output is an integer that is set to one if internal recursion when running
370     \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
371     to remember their state. This is the usual way that PCRE is compiled. The
372     output is zero if PCRE was compiled to use blocks of data on the heap instead
373     of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
374     \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
375     avoiding the use of the stack.
376 nigel 75 .
377     .
378     .SH "COMPILING A PATTERN"
379 nigel 63 .rs
380     .sp
381 nigel 75 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
382 nigel 63 .ti +5n
383 nigel 75 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
384 nigel 63 .ti +5n
385 nigel 75 .B const unsigned char *\fItableptr\fP);
386 nigel 77 .sp
387     .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
388     .ti +5n
389     .B int *\fIerrorcodeptr\fP,
390     .ti +5n
391     .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
392     .ti +5n
393     .B const unsigned char *\fItableptr\fP);
394 nigel 75 .P
395 nigel 77 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
396     called to compile a pattern into an internal form. The only difference between
397     the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
398     \fIerrorcodeptr\fP, via which a numerical error code can be returned.
399 nigel 75 .P
400 nigel 77 The pattern is a C string terminated by a binary zero, and is passed in the
401     \fIpattern\fP argument. A pointer to a single block of memory that is obtained
402     via \fBpcre_malloc\fP is returned. This contains the compiled code and related
403     data. The \fBpcre\fP type is defined for the returned block; this is a typedef
404     for a structure whose contents are not externally defined. It is up to the
405 nigel 91 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
406 nigel 77 .P
407 nigel 63 Although the compiled code of a PCRE regex is relocatable, that is, it does not
408 nigel 75 depend on memory location, the complete \fBpcre\fP data block is not
409     fully relocatable, because it may contain a copy of the \fItableptr\fP
410     argument, which is an address (see below).
411     .P
412 nigel 93 The \fIoptions\fP argument contains various bit settings that affect the
413 nigel 75 compilation. It should be zero if no options are required. The available
414 ph10 412 options are described below. Some of them (in particular, those that are
415     compatible with Perl, but also some others) can also be set and unset from
416     within the pattern (see the detailed description in the
417 nigel 75 .\" HREF
418     \fBpcrepattern\fP
419     .\"
420 ph10 412 documentation). For those options that can be different in different parts of
421     the pattern, the contents of the \fIoptions\fP argument specifies their initial
422     settings at the start of compilation and execution. The PCRE_ANCHORED and
423     PCRE_NEWLINE_\fIxxx\fP options can be set at the time of matching as well as at
424     compile time.
425 nigel 75 .P
426     If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
427     Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
428     NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
429 nigel 87 error message. This is a static string that is part of the library. You must
430 ph10 454 not try to free it. The byte offset from the start of the pattern to the
431     character that was being processes when the error was discovered is placed in
432     the variable pointed to by \fIerroffset\fP, which must not be NULL. If it is,
433     an immediate error is given. Some errors are not detected until checks are
434     carried out when the whole pattern has been scanned; in this case the offset is
435     set to the end of the pattern.
436 nigel 75 .P
437 nigel 77 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
438     \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
439     returned via this argument in the event of an error. This is in addition to the
440     textual error message. Error codes and messages are listed below.
441     .P
442 nigel 75 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
443     character tables that are built when PCRE is compiled, using the default C
444     locale. Otherwise, \fItableptr\fP must be an address that is the result of a
445     call to \fBpcre_maketables()\fP. This value is stored with the compiled
446     pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
447     passed to it. For more discussion, see the section on locale support below.
448     .P
449     This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
450     .sp
451 nigel 63 pcre *re;
452     const char *error;
453     int erroffset;
454     re = pcre_compile(
455     "^A.*Z", /* the pattern */
456     0, /* default options */
457     &error, /* for error message */
458     &erroffset, /* for error offset */
459     NULL); /* use default character tables */
460 nigel 75 .sp
461     The following names for option bits are defined in the \fBpcre.h\fP header
462     file:
463     .sp
464 nigel 63 PCRE_ANCHORED
465 nigel 75 .sp
466 nigel 63 If this bit is set, the pattern is forced to be "anchored", that is, it is
467 nigel 75 constrained to match only at the first matching point in the string that is
468 nigel 63 being searched (the "subject string"). This effect can also be achieved by
469     appropriate constructs in the pattern itself, which is the only way to do it in
470     Perl.
471 nigel 75 .sp
472     PCRE_AUTO_CALLOUT
473     .sp
474     If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
475     all with number 255, before each pattern item. For discussion of the callout
476     facility, see the
477     .\" HREF
478     \fBpcrecallout\fP
479     .\"
480     documentation.
481     .sp
482 ph10 231 PCRE_BSR_ANYCRLF
483     PCRE_BSR_UNICODE
484     .sp
485     These options (which are mutually exclusive) control what the \eR escape
486     sequence matches. The choice is either to match only CR, LF, or CRLF, or to
487     match any Unicode newline sequence. The default is specified when PCRE is
488     built. It can be overridden from within the pattern, or by setting an option
489     when a compiled pattern is matched.
490     .sp
491 nigel 63 PCRE_CASELESS
492 nigel 75 .sp
493 nigel 63 If this bit is set, letters in the pattern match both upper and lower case
494     letters. It is equivalent to Perl's /i option, and it can be changed within a
495 nigel 77 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
496     concept of case for characters whose values are less than 128, so caseless
497     matching is always possible. For characters with higher values, the concept of
498     case is supported if PCRE is compiled with Unicode property support, but not
499     otherwise. If you want to use caseless matching for characters 128 and above,
500     you must ensure that PCRE is compiled with Unicode property support as well as
501     with UTF-8 support.
502 nigel 75 .sp
503 nigel 63 PCRE_DOLLAR_ENDONLY
504 nigel 75 .sp
505 nigel 63 If this bit is set, a dollar metacharacter in the pattern matches only at the
506     end of the subject string. Without this option, a dollar also matches
507 nigel 91 immediately before a newline at the end of the string (but not before any other
508     newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
509     There is no equivalent to this option in Perl, and no way to set it within a
510     pattern.
511 nigel 75 .sp
512 nigel 63 PCRE_DOTALL
513 nigel 75 .sp
514 nigel 63 If this bit is set, a dot metacharater in the pattern matches all characters,
515 nigel 91 including those that indicate newline. Without it, a dot does not match when
516     the current position is at a newline. This option is equivalent to Perl's /s
517     option, and it can be changed within a pattern by a (?s) option setting. A
518 nigel 93 negative class such as [^a] always matches newline characters, independent of
519     the setting of this option.
520 nigel 75 .sp
521 nigel 91 PCRE_DUPNAMES
522     .sp
523     If this bit is set, names used to identify capturing subpatterns need not be
524     unique. This can be helpful for certain types of pattern when it is known that
525     only one instance of the named subpattern can ever be matched. There are more
526     details of named subpatterns below; see also the
527     .\" HREF
528     \fBpcrepattern\fP
529     .\"
530     documentation.
531     .sp
532 nigel 63 PCRE_EXTENDED
533 nigel 75 .sp
534 nigel 63 If this bit is set, whitespace data characters in the pattern are totally
535     ignored except when escaped or inside a character class. Whitespace does not
536     include the VT character (code 11). In addition, characters between an
537 nigel 91 unescaped # outside a character class and the next newline, inclusive, are also
538     ignored. This is equivalent to Perl's /x option, and it can be changed within a
539     pattern by a (?x) option setting.
540 nigel 75 .P
541 nigel 63 This option makes it possible to include comments inside complicated patterns.
542     Note, however, that this applies only to data characters. Whitespace characters
543     may never appear within special character sequences in a pattern, for example
544     within the sequence (?( which introduces a conditional subpattern.
545 nigel 75 .sp
546 nigel 63 PCRE_EXTRA
547 nigel 75 .sp
548 nigel 63 This option was invented in order to turn on additional functionality of PCRE
549     that is incompatible with Perl, but it is currently of very little use. When
550     set, any backslash in a pattern that is followed by a letter that has no
551     special meaning causes an error, thus reserving these combinations for future
552     expansion. By default, as in Perl, a backslash followed by a letter with no
553 nigel 91 special meaning is treated as a literal. (Perl can, however, be persuaded to
554     give a warning for this.) There are at present no other features controlled by
555     this option. It can also be set by a (?X) option setting within a pattern.
556 nigel 75 .sp
557 nigel 77 PCRE_FIRSTLINE
558     .sp
559     If this option is set, an unanchored pattern is required to match before or at
560 nigel 91 the first newline in the subject string, though the matched text may continue
561     over the newline.
562 nigel 77 .sp
563 ph10 336 PCRE_JAVASCRIPT_COMPAT
564     .sp
565 ph10 345 If this option is set, PCRE's behaviour is changed in some ways so that it is
566 ph10 336 compatible with JavaScript rather than Perl. The changes are as follows:
567     .P
568     (1) A lone closing square bracket in a pattern causes a compile-time error,
569     because this is illegal in JavaScript (by default it is treated as a data
570     character). Thus, the pattern AB]CD becomes illegal when this option is set.
571     .P
572     (2) At run time, a back reference to an unset subpattern group matches an empty
573 ph10 345 string (by default this causes the current matching alternative to fail). A
574     pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
575     an "a" in the subject), whereas it fails by default, for Perl compatibility.
576 ph10 336 .sp
577 nigel 63 PCRE_MULTILINE
578 nigel 75 .sp
579     By default, PCRE treats the subject string as consisting of a single line of
580     characters (even if it actually contains newlines). The "start of line"
581 nigel 63 metacharacter (^) matches only at the start of the string, while the "end of
582     line" metacharacter ($) matches only at the end of the string, or before a
583     terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
584     Perl.
585 nigel 75 .P
586 nigel 63 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
587 nigel 91 match immediately following or immediately before internal newlines in the
588     subject string, respectively, as well as at the very start and end. This is
589     equivalent to Perl's /m option, and it can be changed within a pattern by a
590     (?m) option setting. If there are no newlines in a subject string, or no
591 nigel 63 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
592 nigel 75 .sp
593 nigel 91 PCRE_NEWLINE_CR
594     PCRE_NEWLINE_LF
595     PCRE_NEWLINE_CRLF
596 ph10 150 PCRE_NEWLINE_ANYCRLF
597 nigel 93 PCRE_NEWLINE_ANY
598 nigel 91 .sp
599     These options override the default newline definition that was chosen when PCRE
600     was built. Setting the first or the second specifies that a newline is
601 nigel 93 indicated by a single character (CR or LF, respectively). Setting
602     PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
603 ph10 149 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
604     preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
605     that any Unicode newline sequence should be recognized. The Unicode newline
606     sequences are the three just mentioned, plus the single characters VT (vertical
607     tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
608     separator, U+2028), and PS (paragraph separator, U+2029). The last two are
609     recognized only in UTF-8 mode.
610 nigel 91 .P
611 nigel 93 The newline setting in the options word uses three bits that are treated
612 ph10 149 as a number, giving eight possibilities. Currently only six are used (default
613     plus the five values above). This means that if you set more than one newline
614 nigel 93 option, the combination may or may not be sensible. For example,
615     PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
616 ph10 149 other combinations may yield unused numbers and cause an error.
617 nigel 93 .P
618     The only time that a line break is specially recognized when compiling a
619     pattern is if PCRE_EXTENDED is set, and an unescaped # outside a character
620     class is encountered. This indicates a comment that lasts until after the next
621     line break sequence. In other circumstances, line break sequences are treated
622     as literal data, except that in PCRE_EXTENDED mode, both CR and LF are treated
623     as whitespace characters and are therefore ignored.
624     .P
625     The newline option that is set at compile time becomes the default that is used
626     for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
627 nigel 91 .sp
628 nigel 63 PCRE_NO_AUTO_CAPTURE
629 nigel 75 .sp
630 nigel 63 If this option is set, it disables the use of numbered capturing parentheses in
631     the pattern. Any opening parenthesis that is not followed by ? behaves as if it
632     were followed by ?: but named parentheses can still be used for capturing (and
633     they acquire numbers in the usual way). There is no equivalent of this option
634     in Perl.
635 nigel 75 .sp
636 nigel 63 PCRE_UNGREEDY
637 nigel 75 .sp
638 nigel 63 This option inverts the "greediness" of the quantifiers so that they are not
639     greedy by default, but become greedy if followed by "?". It is not compatible
640     with Perl. It can also be set by a (?U) option setting within the pattern.
641 nigel 75 .sp
642 nigel 63 PCRE_UTF8
643 nigel 75 .sp
644 nigel 63 This option causes PCRE to regard both the pattern and the subject as strings
645     of UTF-8 characters instead of single-byte character strings. However, it is
646 nigel 75 available only when PCRE is built to include UTF-8 support. If not, the use
647 nigel 63 of this option provokes an error. Details of how this option changes the
648     behaviour of PCRE are given in the
649     .\" HTML <a href="pcre.html#utf8support">
650     .\" </a>
651     section on UTF-8 support
652     .\"
653     in the main
654     .\" HREF
655 nigel 75 \fBpcre\fP
656 nigel 63 .\"
657     page.
658 nigel 75 .sp
659 nigel 71 PCRE_NO_UTF8_CHECK
660 nigel 75 .sp
661 nigel 71 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
662 ph10 211 automatically checked. There is a discussion about the
663     .\" HTML <a href="pcre.html#utf8strings">
664     .\" </a>
665     validity of UTF-8 strings
666     .\"
667     in the main
668     .\" HREF
669     \fBpcre\fP
670     .\"
671     page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_compile()\fP
672     returns an error. If you already know that your pattern is valid, and you want
673     to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
674     option. When it is set, the effect of passing an invalid UTF-8 string as a
675     pattern is undefined. It may cause your program to crash. Note that this option
676     can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
677     the UTF-8 validity checking of subject strings.
678 nigel 75 .
679     .
680 nigel 77 .SH "COMPILATION ERROR CODES"
681     .rs
682     .sp
683     The following table lists the error codes than may be returned by
684     \fBpcre_compile2()\fP, along with the error messages that may be returned by
685 nigel 93 both compiling functions. As PCRE has developed, some error codes have fallen
686     out of use. To avoid confusion, they have not been re-used.
687 nigel 77 .sp
688     0 no error
689     1 \e at end of pattern
690     2 \ec at end of pattern
691     3 unrecognized character follows \e
692     4 numbers out of order in {} quantifier
693     5 number too big in {} quantifier
694     6 missing terminating ] for character class
695     7 invalid escape sequence in character class
696     8 range out of order in character class
697     9 nothing to repeat
698 nigel 93 10 [this code is not in use]
699 nigel 77 11 internal error: unexpected repeat
700 ph10 290 12 unrecognized character after (? or (?-
701 nigel 77 13 POSIX named classes are supported only within a class
702     14 missing )
703     15 reference to non-existent subpattern
704     16 erroffset passed as NULL
705     17 unknown option bit(s) set
706     18 missing ) after comment
707 nigel 93 19 [this code is not in use]
708 ph10 290 20 regular expression is too large
709 nigel 77 21 failed to get memory
710     22 unmatched parentheses
711     23 internal error: code overflow
712     24 unrecognized character after (?<
713     25 lookbehind assertion is not fixed length
714 nigel 91 26 malformed number or name after (?(
715 nigel 77 27 conditional group contains more than two branches
716     28 assertion expected after (?(
717 ph10 181 29 (?R or (?[+-]digits must be followed by )
718 nigel 77 30 unknown POSIX class name
719     31 POSIX collating elements are not supported
720     32 this version of PCRE is not compiled with PCRE_UTF8 support
721 nigel 93 33 [this code is not in use]
722 nigel 77 34 character value in \ex{...} sequence is too large
723     35 invalid condition (?(0)
724     36 \eC not allowed in lookbehind assertion
725     37 PCRE does not support \eL, \el, \eN, \eU, or \eu
726     38 number after (?C is > 255
727     39 closing ) for (?C expected
728     40 recursive call could loop indefinitely
729     41 unrecognized character after (?P
730 nigel 93 42 syntax error in subpattern name (missing terminator)
731 nigel 91 43 two named subpatterns have the same name
732 nigel 77 44 invalid UTF-8 string
733     45 support for \eP, \ep, and \eX has not been compiled
734     46 malformed \eP or \ep sequence
735     47 unknown property name after \eP or \ep
736 nigel 91 48 subpattern name is too long (maximum 32 characters)
737 ph10 290 49 too many named subpatterns (maximum 10000)
738 ph10 202 50 [this code is not in use]
739 nigel 91 51 octal value is greater than \e377 (not in UTF-8 mode)
740 nigel 93 52 internal error: overran compiling workspace
741     53 internal error: previously-checked referenced subpattern not found
742     54 DEFINE group contains more than one branch
743     55 repeating a DEFINE group is not allowed
744 ph10 231 56 inconsistent NEWLINE options
745 ph10 345 57 \eg is not followed by a braced, angle-bracketed, or quoted
746     name/number or by a plain number
747 ph10 336 58 a numbered reference must not be zero
748 ph10 290 59 (*VERB) with an argument is not supported
749     60 (*VERB) not recognized
750     61 number is too big
751     62 subpattern name expected
752 ph10 292 63 digit expected after (?+
753 ph10 336 64 ] is an invalid data character in JavaScript compatibility mode
754 ph10 290 .sp
755 ph10 292 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
756 ph10 290 be used if the limits were changed when PCRE was built.
757 nigel 77 .
758     .
759 nigel 75 .SH "STUDYING A PATTERN"
760 nigel 63 .rs
761     .sp
762 nigel 77 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
763 nigel 63 .ti +5n
764 nigel 75 .B const char **\fIerrptr\fP);
765 nigel 63 .PP
766 nigel 75 If a compiled pattern is going to be used several times, it is worth spending
767     more time analyzing it in order to speed up the time taken for matching. The
768     function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
769     argument. If studying the pattern produces additional information that will
770     help speed up matching, \fBpcre_study()\fP returns a pointer to a
771     \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
772     results of the study.
773     .P
774     The returned value from \fBpcre_study()\fP can be passed directly to
775 ph10 455 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
776     also contains other fields that can be set by the caller before the block is
777     passed; these are described
778 nigel 75 .\" HTML <a href="#extradata">
779     .\" </a>
780     below
781     .\"
782     in the section on matching a pattern.
783     .P
784 ph10 455 If studying the pattern does not produce any useful information,
785 nigel 75 \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program
786 ph10 455 wants to pass any of the other fields to \fBpcre_exec()\fP or
787     \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block.
788 nigel 75 .P
789     The second argument of \fBpcre_study()\fP contains option bits. At present, no
790     options are defined, and this argument should always be zero.
791     .P
792     The third argument for \fBpcre_study()\fP is a pointer for an error message. If
793 nigel 63 studying succeeds (even if no data is returned), the variable it points to is
794 nigel 87 set to NULL. Otherwise it is set to point to a textual error message. This is a
795     static string that is part of the library. You must not try to free it. You
796     should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
797     sure that it has run successfully.
798 nigel 75 .P
799     This is a typical call to \fBpcre_study\fP():
800     .sp
801 nigel 63 pcre_extra *pe;
802     pe = pcre_study(
803     re, /* result of pcre_compile() */
804     0, /* no options exist */
805     &error); /* set to NULL or points to a message */
806 nigel 75 .sp
807 ph10 455 Studying a pattern does two things: first, a lower bound for the length of
808     subject string that is needed to match the pattern is computed. This does not
809     mean that there are any strings of that length that match, but it does
810     guarantee that no shorter strings match. The value is used by
811     \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP to avoid wasting time by trying to
812     match strings that are shorter than the lower bound. You can find out the value
813     in a calling program via the \fBpcre_fullinfo()\fP function.
814     .P
815     Studying a pattern is also useful for non-anchored patterns that do not have a
816     single fixed starting character. A bitmap of possible starting bytes is
817     created. This speeds up finding a position in the subject at which to start
818     matching.
819 nigel 75 .
820     .
821 nigel 63 .\" HTML <a name="localesupport"></a>
822 nigel 75 .SH "LOCALE SUPPORT"
823 nigel 63 .rs
824     .sp
825 ph10 139 PCRE handles caseless matching, and determines whether characters are letters,
826 nigel 75 digits, or whatever, by reference to a set of tables, indexed by character
827 nigel 77 value. When running in UTF-8 mode, this applies only to characters with codes
828 nigel 75 less than 128. Higher-valued codes never match escapes such as \ew or \ed, but
829     can be tested with \ep if PCRE is built with Unicode character property
830 ph10 142 support. The use of locales with Unicode is discouraged. If you are handling
831     characters with codes greater than 128, you should either use UTF-8 and
832 ph10 139 Unicode, or use locales, but not try to mix the two.
833 nigel 75 .P
834 ph10 139 PCRE contains an internal set of tables that are used when the final argument
835     of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
836 ph10 142 Normally, the internal tables recognize only ASCII characters. However, when
837 ph10 139 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
838     default "C" locale of the local system, which may cause them to be different.
839 nigel 75 .P
840 ph10 139 The internal tables can always be overridden by tables supplied by the
841     application that calls PCRE. These may be created in a different locale from
842     the default. As more and more applications change to using Unicode, the need
843     for this locale support is expected to die away.
844     .P
845 nigel 75 External tables are built by calling the \fBpcre_maketables()\fP function,
846     which has no arguments, in the relevant locale. The result can then be passed
847     to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
848     example, to build and use tables that are appropriate for the French locale
849     (where accented characters with values greater than 128 are treated as letters),
850     the following code could be used:
851     .sp
852     setlocale(LC_CTYPE, "fr_FR");
853 nigel 63 tables = pcre_maketables();
854     re = pcre_compile(..., tables);
855 nigel 75 .sp
856 ph10 142 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
857 ph10 139 are using Windows, the name for the French locale is "french".
858     .P
859 nigel 75 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
860     obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
861     that the memory containing the tables remains available for as long as it is
862     needed.
863     .P
864     The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
865     pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
866     and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
867     pattern, compilation, studying and matching all happen in the same locale, but
868     different patterns can be compiled in different locales.
869     .P
870     It is possible to pass a table pointer or NULL (indicating the use of the
871     internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
872     this facility could be used to match a pattern in a different locale from the
873     one in which it was compiled. Passing table pointers at run time is discussed
874     below in the section on matching a pattern.
875     .
876     .
877     .SH "INFORMATION ABOUT A PATTERN"
878 nigel 63 .rs
879     .sp
880 nigel 75 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
881 nigel 63 .ti +5n
882 nigel 75 .B int \fIwhat\fP, void *\fIwhere\fP);
883 nigel 63 .PP
884 nigel 75 The \fBpcre_fullinfo()\fP function returns information about a compiled
885     pattern. It replaces the obsolete \fBpcre_info()\fP function, which is
886 nigel 63 nevertheless retained for backwards compability (and is documented below).
887 nigel 75 .P
888     The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
889     pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
890 nigel 63 the pattern was not studied. The third argument specifies which piece of
891     information is required, and the fourth argument is a pointer to a variable
892     to receive the data. The yield of the function is zero for success, or one of
893     the following negative numbers:
894 nigel 75 .sp
895     PCRE_ERROR_NULL the argument \fIcode\fP was NULL
896     the argument \fIwhere\fP was NULL
897 nigel 63 PCRE_ERROR_BADMAGIC the "magic number" was not found
898 nigel 75 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
899     .sp
900     The "magic number" is placed at the start of each compiled pattern as an simple
901     check against passing an arbitrary memory pointer. Here is a typical call of
902     \fBpcre_fullinfo()\fP, to obtain the length of the compiled pattern:
903     .sp
904 nigel 63 int rc;
905 nigel 91 size_t length;
906 nigel 63 rc = pcre_fullinfo(
907     re, /* result of pcre_compile() */
908     pe, /* result of pcre_study(), or NULL */
909     PCRE_INFO_SIZE, /* what is required */
910     &length); /* where to put the data */
911 nigel 75 .sp
912     The possible values for the third argument are defined in \fBpcre.h\fP, and are
913 nigel 63 as follows:
914 nigel 75 .sp
915 nigel 63 PCRE_INFO_BACKREFMAX
916 nigel 75 .sp
917 nigel 63 Return the number of the highest back reference in the pattern. The fourth
918 nigel 75 argument should point to an \fBint\fP variable. Zero is returned if there are
919 nigel 63 no back references.
920 nigel 75 .sp
921 nigel 63 PCRE_INFO_CAPTURECOUNT
922 nigel 75 .sp
923 nigel 63 Return the number of capturing subpatterns in the pattern. The fourth argument
924 nigel 75 should point to an \fBint\fP variable.
925     .sp
926 nigel 77 PCRE_INFO_DEFAULT_TABLES
927 nigel 75 .sp
928     Return a pointer to the internal default character tables within PCRE. The
929     fourth argument should point to an \fBunsigned char *\fP variable. This
930     information call is provided for internal use by the \fBpcre_study()\fP
931     function. External callers can cause PCRE to use its internal tables by passing
932     a NULL table pointer.
933     .sp
934 nigel 63 PCRE_INFO_FIRSTBYTE
935 nigel 75 .sp
936 nigel 63 Return information about the first byte of any matched string, for a
937 nigel 91 non-anchored pattern. The fourth argument should point to an \fBint\fP
938     variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
939     still recognized for backwards compatibility.)
940 nigel 75 .P
941     If there is a fixed first byte, for example, from a pattern such as
942 nigel 93 (cat|cow|coyote), its value is returned. Otherwise, if either
943 nigel 75 .sp
944 nigel 63 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
945     starts with "^", or
946 nigel 75 .sp
947 nigel 63 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
948     (if it were set, the pattern would be anchored),
949 nigel 75 .sp
950 nigel 63 -1 is returned, indicating that the pattern matches only at the start of a
951     subject string or after any newline within the string. Otherwise -2 is
952     returned. For anchored patterns, -2 is returned.
953 nigel 75 .sp
954 nigel 63 PCRE_INFO_FIRSTTABLE
955 nigel 75 .sp
956 nigel 63 If the pattern was studied, and this resulted in the construction of a 256-bit
957     table indicating a fixed set of bytes for the first byte in any matching
958     string, a pointer to the table is returned. Otherwise NULL is returned. The
959 nigel 75 fourth argument should point to an \fBunsigned char *\fP variable.
960     .sp
961 ph10 226 PCRE_INFO_HASCRORLF
962     .sp
963 ph10 227 Return 1 if the pattern contains any explicit matches for CR or LF characters,
964 ph10 243 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
965 ph10 231 explicit match is either a literal CR or LF character, or \er or \en.
966 ph10 226 .sp
967 ph10 169 PCRE_INFO_JCHANGED
968     .sp
969 ph10 278 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
970     0. The fourth argument should point to an \fBint\fP variable. (?J) and
971     (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
972 ph10 169 .sp
973 nigel 63 PCRE_INFO_LASTLITERAL
974 nigel 75 .sp
975 nigel 65 Return the value of the rightmost literal byte that must exist in any matched
976     string, other than at its start, if such a byte has been recorded. The fourth
977 nigel 75 argument should point to an \fBint\fP variable. If there is no such byte, -1 is
978 nigel 65 returned. For anchored patterns, a last literal byte is recorded only if it
979     follows something of variable length. For example, for the pattern
980 nigel 75 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
981 nigel 65 is -1.
982 nigel 75 .sp
983 ph10 455 PCRE_INFO_MINLENGTH
984     .sp
985     If the pattern was studied and a minimum length for matching subject strings
986     was computed, its value is returned. Otherwise the returned value is -1. The
987     value is a number of characters, not bytes (there may be a difference in UTF-8
988     mode). The fourth argument should point to an \fBint\fP variable. A
989     non-negative value is a lower bound to the length of any matching string. There
990     may not be any strings of that length that do actually match, but every string
991     that does match is at least that long.
992     .sp
993 nigel 63 PCRE_INFO_NAMECOUNT
994     PCRE_INFO_NAMEENTRYSIZE
995     PCRE_INFO_NAMETABLE
996 nigel 75 .sp
997 nigel 63 PCRE supports the use of named as well as numbered capturing parentheses. The
998     names are just an additional way of identifying the parentheses, which still
999 nigel 91 acquire numbers. Several convenience functions such as
1000     \fBpcre_get_named_substring()\fP are provided for extracting captured
1001     substrings by name. It is also possible to extract the data directly, by first
1002     converting the name to a number in order to access the correct pointers in the
1003     output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1004     you need to use the name-to-number map, which is described by these three
1005     values.
1006 nigel 75 .P
1007 nigel 63 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1008     the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1009 nigel 75 entry; both of these return an \fBint\fP value. The entry size depends on the
1010 nigel 63 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1011 nigel 75 entry of the table (a pointer to \fBchar\fP). The first two bytes of each entry
1012 nigel 63 are the number of the capturing parenthesis, most significant byte first. The
1013     rest of the entry is the corresponding name, zero terminated. The names are in
1014 nigel 91 alphabetical order. When PCRE_DUPNAMES is set, duplicate names are in order of
1015     their parentheses numbers. For example, consider the following pattern (assume
1016 nigel 63 PCRE_EXTENDED is set, so white space - including newlines - is ignored):
1017 nigel 75 .sp
1018     .\" JOIN
1019 nigel 93 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1020     (?<month>\ed\ed) - (?<day>\ed\ed) )
1021 nigel 75 .sp
1022 nigel 63 There are four named subpatterns, so the table has four entries, and each entry
1023     in the table is eight bytes long. The table is as follows, with non-printing
1024 nigel 75 bytes shows in hexadecimal, and undefined bytes shown as ??:
1025     .sp
1026 nigel 63 00 01 d a t e 00 ??
1027     00 05 d a y 00 ?? ??
1028     00 04 m o n t h 00
1029     00 02 y e a r 00 ??
1030 nigel 75 .sp
1031     When writing code to extract data from named subpatterns using the
1032 nigel 91 name-to-number map, remember that the length of the entries is likely to be
1033 nigel 75 different for each compiled pattern.
1034     .sp
1035 ph10 169 PCRE_INFO_OKPARTIAL
1036     .sp
1037 ph10 435 Return 1 if the pattern can be used for partial matching with
1038     \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1039     \fBint\fP variable. From release 8.00, this always returns 1, because the
1040     restrictions that previously applied to partial matching have been lifted. The
1041 ph10 169 .\" HREF
1042     \fBpcrepartial\fP
1043     .\"
1044 ph10 426 documentation gives details of partial matching.
1045 ph10 169 .sp
1046 nigel 63 PCRE_INFO_OPTIONS
1047 nigel 75 .sp
1048 nigel 63 Return a copy of the options with which the pattern was compiled. The fourth
1049 nigel 75 argument should point to an \fBunsigned long int\fP variable. These option bits
1050     are those specified in the call to \fBpcre_compile()\fP, modified by any
1051 ph10 196 top-level option settings at the start of the pattern itself. In other words,
1052     they are the options that will be in force when matching starts. For example,
1053     if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1054     result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1055 nigel 75 .P
1056 nigel 63 A pattern is automatically anchored by PCRE if all of its top-level
1057     alternatives begin with one of the following:
1058 nigel 75 .sp
1059 nigel 63 ^ unless PCRE_MULTILINE is set
1060 nigel 75 \eA always
1061     \eG always
1062     .\" JOIN
1063 nigel 63 .* if PCRE_DOTALL is set and there are no back
1064     references to the subpattern in which .* appears
1065 nigel 75 .sp
1066 nigel 63 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1067 nigel 75 \fBpcre_fullinfo()\fP.
1068     .sp
1069 nigel 63 PCRE_INFO_SIZE
1070 nigel 75 .sp
1071 nigel 63 Return the size of the compiled pattern, that is, the value that was passed as
1072 nigel 75 the argument to \fBpcre_malloc()\fP when PCRE was getting memory in which to
1073     place the compiled data. The fourth argument should point to a \fBsize_t\fP
1074 nigel 63 variable.
1075 nigel 75 .sp
1076 nigel 63 PCRE_INFO_STUDYSIZE
1077 nigel 75 .sp
1078     Return the size of the data block pointed to by the \fIstudy_data\fP field in
1079     a \fBpcre_extra\fP block. That is, it is the value that was passed to
1080     \fBpcre_malloc()\fP when PCRE was getting memory into which to place the data
1081 ph10 455 created by \fBpcre_study()\fP. If \fBpcre_extra\fP is NULL, or there is no
1082     study data, zero is returned. The fourth argument should point to a
1083 nigel 75 \fBsize_t\fP variable.
1084     .
1085     .
1086     .SH "OBSOLETE INFO FUNCTION"
1087 nigel 63 .rs
1088     .sp
1089 nigel 75 .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
1090     .B *\fIfirstcharptr\fP);
1091 nigel 63 .PP
1092 nigel 75 The \fBpcre_info()\fP function is now obsolete because its interface is too
1093 nigel 63 restrictive to return all the available data about a compiled pattern. New
1094 nigel 75 programs should use \fBpcre_fullinfo()\fP instead. The yield of
1095     \fBpcre_info()\fP is the number of capturing subpatterns, or one of the
1096 nigel 63 following negative numbers:
1097 nigel 75 .sp
1098     PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1099 nigel 63 PCRE_ERROR_BADMAGIC the "magic number" was not found
1100 nigel 75 .sp
1101     If the \fIoptptr\fP argument is not NULL, a copy of the options with which the
1102 nigel 63 pattern was compiled is placed in the integer it points to (see
1103     PCRE_INFO_OPTIONS above).
1104 nigel 75 .P
1105     If the pattern is not anchored and the \fIfirstcharptr\fP argument is not NULL,
1106 nigel 63 it is used to pass back information about the first character of any matched
1107     string (see PCRE_INFO_FIRSTBYTE above).
1108 nigel 75 .
1109     .
1110 nigel 77 .SH "REFERENCE COUNTS"
1111 nigel 63 .rs
1112     .sp
1113 nigel 77 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1114     .PP
1115     The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1116     data block that contains a compiled pattern. It is provided for the benefit of
1117     applications that operate in an object-oriented manner, where different parts
1118     of the application may be using the same compiled pattern, but you want to free
1119     the block when they are all done.
1120     .P
1121     When a pattern is compiled, the reference count field is initialized to zero.
1122     It is changed only by calling this function, whose action is to add the
1123     \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1124     function is the new value. However, the value of the count is constrained to
1125     lie between 0 and 65535, inclusive. If the new value is outside these limits,
1126     it is forced to the appropriate limit value.
1127     .P
1128     Except when it is zero, the reference count is not correctly preserved if a
1129     pattern is compiled on one host and then transferred to a host whose byte-order
1130     is different. (This seems a highly unlikely scenario.)
1131     .
1132     .
1133     .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1134     .rs
1135     .sp
1136 nigel 75 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1137 nigel 63 .ti +5n
1138 nigel 75 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1139 nigel 63 .ti +5n
1140 nigel 75 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1141     .P
1142     The function \fBpcre_exec()\fP is called to match a subject string against a
1143     compiled pattern, which is passed in the \fIcode\fP argument. If the
1144 ph10 455 pattern was studied, the result of the study should be passed in the
1145 nigel 77 \fIextra\fP argument. This function is the main matching facility of the
1146     library, and it operates in a Perl-like manner. For specialist use there is
1147     also an alternative matching function, which is described
1148     .\" HTML <a href="#dfamatch">
1149     .\" </a>
1150     below
1151     .\"
1152     in the section about the \fBpcre_dfa_exec()\fP function.
1153 nigel 75 .P
1154     In most applications, the pattern will have been compiled (and optionally
1155     studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1156     possible to save compiled patterns and study data, and then use them later
1157     in different processes, possibly even on different hosts. For a discussion
1158     about this, see the
1159     .\" HREF
1160     \fBpcreprecompile\fP
1161     .\"
1162     documentation.
1163     .P
1164     Here is an example of a simple call to \fBpcre_exec()\fP:
1165     .sp
1166 nigel 63 int rc;
1167     int ovector[30];
1168     rc = pcre_exec(
1169     re, /* result of pcre_compile() */
1170     NULL, /* we didn't study the pattern */
1171     "some string", /* the subject string */
1172     11, /* the length of the subject string */
1173     0, /* start at offset 0 in the subject */
1174     0, /* default options */
1175 nigel 75 ovector, /* vector of integers for substring information */
1176 nigel 77 30); /* number of elements (NOT size in bytes) */
1177 nigel 75 .
1178     .\" HTML <a name="extradata"></a>
1179     .SS "Extra data for \fBpcre_exec()\fR"
1180     .rs
1181     .sp
1182     If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1183     data block. The \fBpcre_study()\fP function returns such a block (when it
1184 nigel 63 doesn't return NULL), but you can also create one for yourself, and pass
1185 nigel 87 additional information in it. The \fBpcre_extra\fP block contains the following
1186     fields (not necessarily in this order):
1187 nigel 75 .sp
1188     unsigned long int \fIflags\fP;
1189     void *\fIstudy_data\fP;
1190     unsigned long int \fImatch_limit\fP;
1191 nigel 87 unsigned long int \fImatch_limit_recursion\fP;
1192 nigel 75 void *\fIcallout_data\fP;
1193     const unsigned char *\fItables\fP;
1194     .sp
1195     The \fIflags\fP field is a bitmap that specifies which of the other fields
1196 nigel 63 are set. The flag bits are:
1197 nigel 75 .sp
1198 nigel 63 PCRE_EXTRA_STUDY_DATA
1199     PCRE_EXTRA_MATCH_LIMIT
1200 nigel 87 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1201 nigel 63 PCRE_EXTRA_CALLOUT_DATA
1202 nigel 75 PCRE_EXTRA_TABLES
1203     .sp
1204     Other flag bits should be set to zero. The \fIstudy_data\fP field is set in the
1205     \fBpcre_extra\fP block that is returned by \fBpcre_study()\fP, together with
1206     the appropriate flag bit. You should not set this yourself, but you may add to
1207     the block by setting the other fields and their corresponding flag bits.
1208     .P
1209     The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1210 nigel 63 vast amount of resources when running patterns that are not going to match,
1211     but which have a very large number of possibilities in their search trees. The
1212 nigel 75 classic example is the use of nested unlimited repeats.
1213     .P
1214     Internally, PCRE uses a function called \fBmatch()\fP which it calls repeatedly
1215 nigel 87 (sometimes recursively). The limit set by \fImatch_limit\fP is imposed on the
1216     number of times this function is called during a match, which has the effect of
1217     limiting the amount of backtracking that can take place. For patterns that are
1218     not anchored, the count restarts from zero for each position in the subject
1219     string.
1220 nigel 75 .P
1221 nigel 87 The default value for the limit can be set when PCRE is built; the default
1222 nigel 63 default is 10 million, which handles all but the most extreme cases. You can
1223 nigel 87 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1224     block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1225     the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1226     PCRE_ERROR_MATCHLIMIT.
1227 nigel 75 .P
1228 nigel 87 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1229     instead of limiting the total number of times that \fBmatch()\fP is called, it
1230     limits the depth of recursion. The recursion depth is a smaller number than the
1231     total number of calls, because not all calls to \fBmatch()\fP are recursive.
1232     This limit is of use only if it is set smaller than \fImatch_limit\fP.
1233     .P
1234     Limiting the recursion depth limits the amount of stack that can be used, or,
1235     when PCRE has been compiled to use memory on the heap instead of the stack, the
1236     amount of heap memory that can be used.
1237     .P
1238     The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1239     built; the default default is the same value as the default for
1240     \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1241     with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1242     PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1243     is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1244     .P
1245 ph10 440 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1246     and is described in the
1247 nigel 75 .\" HREF
1248     \fBpcrecallout\fP
1249     .\"
1250     documentation.
1251     .P
1252     The \fItables\fP field is used to pass a character tables pointer to
1253     \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1254     pattern. A non-NULL value is stored with the compiled pattern only if custom
1255     tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1256     If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1257     internal tables to be used. This facility is helpful when re-using patterns
1258     that have been saved after compiling with an external set of tables, because
1259     the external tables might be at a different address when \fBpcre_exec()\fP is
1260     called. See the
1261     .\" HREF
1262     \fBpcreprecompile\fP
1263     .\"
1264     documentation for a discussion of saving compiled patterns for later use.
1265     .
1266 ph10 226 .\" HTML <a name="execoptions"></a>
1267 nigel 75 .SS "Option bits for \fBpcre_exec()\fP"
1268     .rs
1269     .sp
1270     The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1271 nigel 91 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1272 ph10 442 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1273     PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_SOFT, and
1274     PCRE_PARTIAL_HARD.
1275 nigel 75 .sp
1276     PCRE_ANCHORED
1277     .sp
1278     The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1279     matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1280     to be anchored by virtue of its contents, it cannot be made unachored at
1281     matching time.
1282     .sp
1283 ph10 231 PCRE_BSR_ANYCRLF
1284     PCRE_BSR_UNICODE
1285     .sp
1286     These options (which are mutually exclusive) control what the \eR escape
1287     sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1288     match any Unicode newline sequence. These options override the choice that was
1289     made or defaulted when the pattern was compiled.
1290     .sp
1291 nigel 91 PCRE_NEWLINE_CR
1292     PCRE_NEWLINE_LF
1293     PCRE_NEWLINE_CRLF
1294 ph10 150 PCRE_NEWLINE_ANYCRLF
1295 nigel 93 PCRE_NEWLINE_ANY
1296 nigel 91 .sp
1297     These options override the newline definition that was chosen or defaulted when
1298 nigel 93 the pattern was compiled. For details, see the description of
1299     \fBpcre_compile()\fP above. During matching, the newline choice affects the
1300     behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1301     the way the match position is advanced after a match failure for an unanchored
1302 ph10 227 pattern.
1303 ph10 225 .P
1304 ph10 226 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1305     match attempt for an unanchored pattern fails when the current position is at a
1306 ph10 230 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1307 ph10 226 characters, the match position is advanced by two characters instead of one, in
1308     other words, to after the CRLF.
1309     .P
1310 ph10 227 The above rule is a compromise that makes the most common cases work as
1311     expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1312     set), it does not match the string "\er\enA" because, after failing at the
1313     start, it skips both the CR and the LF before retrying. However, the pattern
1314     [\er\en]A does match that string, because it contains an explicit CR or LF
1315 ph10 226 reference, and so advances only by one character after the first failure.
1316     .P
1317 ph10 231 An explicit match for CR of LF is either a literal appearance of one of those
1318     characters, or one of the \er or \en escape sequences. Implicit matches such as
1319 ph10 230 [^X] do not count, nor does \es (which includes CR and LF in the characters
1320     that it matches).
1321     .P
1322 ph10 226 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1323     valid newline sequence and explicit \er or \en escapes appear in the pattern.
1324 nigel 91 .sp
1325 nigel 63 PCRE_NOTBOL
1326 nigel 75 .sp
1327     This option specifies that first character of the subject string is not the
1328     beginning of a line, so the circumflex metacharacter should not match before
1329     it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1330     never to match. This option affects only the behaviour of the circumflex
1331     metacharacter. It does not affect \eA.
1332     .sp
1333 nigel 63 PCRE_NOTEOL
1334 nigel 75 .sp
1335     This option specifies that the end of the subject string is not the end of a
1336     line, so the dollar metacharacter should not match it nor (except in multiline
1337     mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1338     compile time) causes dollar never to match. This option affects only the
1339     behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1340     .sp
1341 nigel 63 PCRE_NOTEMPTY
1342 nigel 75 .sp
1343 nigel 63 An empty string is not considered to be a valid match if this option is set. If
1344     there are alternatives in the pattern, they are tried. If all the alternatives
1345     match the empty string, the entire match fails. For example, if the pattern
1346 nigel 75 .sp
1347 nigel 63 a?b?
1348 nigel 75 .sp
1349 ph10 442 is applied to a string not beginning with "a" or "b", it matches an empty
1350 nigel 63 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1351     valid, so PCRE searches further into the string for occurrences of "a" or "b".
1352 ph10 442 .sp
1353     PCRE_NOTEMPTY_ATSTART
1354     .sp
1355     This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1356     the start of the subject is permitted. If the pattern is anchored, such a match
1357     can occur only if the pattern contains \eK.
1358 nigel 75 .P
1359 ph10 442 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1360     does make a special case of a pattern match of the empty string within its
1361     \fBsplit()\fP function, and when using the /g modifier. It is possible to
1362     emulate Perl's behaviour after matching a null string by first trying the match
1363     again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1364     if that fails, by advancing the starting offset (see below) and trying an
1365     ordinary match again. There is some code that demonstrates how to do this in
1366     the
1367 ph10 429 .\" HREF
1368     \fBpcredemo\fP
1369     .\"
1370     sample program.
1371 nigel 75 .sp
1372 ph10 389 PCRE_NO_START_OPTIMIZE
1373     .sp
1374 ph10 392 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1375 ph10 389 a match, in order to speed up the process. For example, if it is known that a
1376     match must start with a specific character, it searches the subject for that
1377     character, and fails immediately if it cannot find it, without actually running
1378     the main matching function. When callouts are in use, these optimizations can
1379     cause them to be skipped. This option disables the "start-up" optimizations,
1380     causing performance to suffer, but ensuring that the callouts do occur.
1381     .sp
1382 nigel 75 PCRE_NO_UTF8_CHECK
1383     .sp
1384     When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1385     string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1386 ph10 211 The value of \fIstartoffset\fP is also checked to ensure that it points to the
1387     start of a UTF-8 character. There is a discussion about the validity of UTF-8
1388     strings in the
1389     .\" HTML <a href="pcre.html#utf8strings">
1390     .\" </a>
1391     section on UTF-8 support
1392     .\"
1393     in the main
1394     .\" HREF
1395     \fBpcre\fP
1396     .\"
1397     page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_exec()\fP returns
1398     the error PCRE_ERROR_BADUTF8. If \fIstartoffset\fP contains an invalid value,
1399 ph10 209 PCRE_ERROR_BADUTF8_OFFSET is returned.
1400 nigel 75 .P
1401     If you already know that your subject is valid, and you want to skip these
1402     checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1403     calling \fBpcre_exec()\fP. You might want to do this for the second and
1404     subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1405     all the matches in a single subject string. However, you should be sure that
1406     the value of \fIstartoffset\fP points to the start of a UTF-8 character. When
1407     PCRE_NO_UTF8_CHECK is set, the effect of passing an invalid UTF-8 string as a
1408     subject, or a value of \fIstartoffset\fP that does not point to the start of a
1409     UTF-8 character, is undefined. Your program may crash.
1410     .sp
1411 ph10 428 PCRE_PARTIAL_HARD
1412     PCRE_PARTIAL_SOFT
1413 nigel 75 .sp
1414 ph10 428 These options turn on the partial matching feature. For backwards
1415     compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1416     occurs if the end of the subject string is reached successfully, but there are
1417     not enough subject characters to complete the match. If this happens when
1418     PCRE_PARTIAL_HARD is set, \fBpcre_exec()\fP immediately returns
1419     PCRE_ERROR_PARTIAL. Otherwise, if PCRE_PARTIAL_SOFT is set, matching continues
1420     by testing any other alternatives. Only if they all fail is PCRE_ERROR_PARTIAL
1421     returned (instead of PCRE_ERROR_NOMATCH). The portion of the string that
1422 ph10 435 was inspected when the partial match was found is set as the first matching
1423     string. There is a more detailed discussion in the
1424 nigel 75 .\" HREF
1425     \fBpcrepartial\fP
1426     .\"
1427     documentation.
1428     .
1429     .SS "The string to be matched by \fBpcre_exec()\fP"
1430     .rs
1431     .sp
1432     The subject string is passed to \fBpcre_exec()\fP as a pointer in
1433 ph10 368 \fIsubject\fP, a length (in bytes) in \fIlength\fP, and a starting byte offset
1434     in \fIstartoffset\fP. In UTF-8 mode, the byte offset must point to the start of
1435     a UTF-8 character. Unlike the pattern string, the subject may contain binary
1436     zero bytes. When the starting offset is zero, the search for a match starts at
1437     the beginning of the subject, and this is by far the most common case.
1438 nigel 75 .P
1439 nigel 63 A non-zero starting offset is useful when searching for another match in the
1440 nigel 75 same subject by calling \fBpcre_exec()\fP again after a previous success.
1441     Setting \fIstartoffset\fP differs from just passing over a shortened string and
1442 nigel 63 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1443     lookbehind. For example, consider the pattern
1444 nigel 75 .sp
1445     \eBiss\eB
1446     .sp
1447     which finds occurrences of "iss" in the middle of words. (\eB matches only if
1448 nigel 63 the current position in the subject is not a word boundary.) When applied to
1449 nigel 75 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1450     occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1451     subject, namely "issipi", it does not match, because \eB is always false at the
1452 nigel 63 start of the subject, which is deemed to be a word boundary. However, if
1453 nigel 75 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1454 nigel 63 set to 4, it finds the second occurrence of "iss" because it is able to look
1455     behind the starting point to discover that it is preceded by a letter.
1456 nigel 75 .P
1457 nigel 63 If a non-zero starting offset is passed when the pattern is anchored, one
1458 nigel 75 attempt to match at the given offset is made. This can only succeed if the
1459 nigel 63 pattern does not require the match to be at the start of the subject.
1460 nigel 75 .
1461     .SS "How \fBpcre_exec()\fP returns captured substrings"
1462     .rs
1463     .sp
1464 nigel 63 In general, a pattern matches a certain portion of the subject, and in
1465     addition, further substrings from the subject may be picked out by parts of the
1466     pattern. Following the usage in Jeffrey Friedl's book, this is called
1467     "capturing" in what follows, and the phrase "capturing subpattern" is used for
1468     a fragment of a pattern that picks out a substring. PCRE supports several other
1469     kinds of parenthesized subpattern that do not cause substrings to be captured.
1470 nigel 75 .P
1471 ph10 368 Captured substrings are returned to the caller via a vector of integers whose
1472     address is passed in \fIovector\fP. The number of elements in the vector is
1473     passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
1474     argument is NOT the size of \fIovector\fP in bytes.
1475 nigel 75 .P
1476     The first two-thirds of the vector is used to pass back captured substrings,
1477     each substring using a pair of integers. The remaining third of the vector is
1478     used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1479 ph10 368 and is not available for passing back information. The number passed in
1480 nigel 75 \fIovecsize\fP should always be a multiple of three. If it is not, it is
1481     rounded down.
1482     .P
1483     When a match is successful, information about captured substrings is returned
1484     in pairs of integers, starting at the beginning of \fIovector\fP, and
1485 ph10 371 continuing up to two-thirds of its length at the most. The first element of
1486 ph10 368 each pair is set to the byte offset of the first character in a substring, and
1487     the second is set to the byte offset of the first character after the end of a
1488     substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
1489     mode. They are not character counts.
1490 nigel 75 .P
1491 ph10 368 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
1492     portion of the subject string matched by the entire pattern. The next pair is
1493     used for the first capturing subpattern, and so on. The value returned by
1494     \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
1495     For example, if two substrings have been captured, the returned value is 3. If
1496     there are no capturing subpatterns, the return value from a successful match is
1497     1, indicating that just the first pair of offsets has been set.
1498     .P
1499 nigel 63 If a capturing subpattern is matched repeatedly, it is the last portion of the
1500 nigel 75 string that it matched that is returned.
1501     .P
1502     If the vector is too small to hold all the captured substring offsets, it is
1503     used as far as possible (up to two-thirds of its length), and the function
1504 ph10 368 returns a value of zero. If the substring offsets are not of interest,
1505     \fBpcre_exec()\fP may be called with \fIovector\fP passed as NULL and
1506 nigel 75 \fIovecsize\fP as zero. However, if the pattern contains back references and
1507     the \fIovector\fP is not big enough to remember the related substrings, PCRE
1508     has to get additional memory for use during matching. Thus it is usually
1509     advisable to supply an \fIovector\fP.
1510     .P
1511 nigel 91 The \fBpcre_info()\fP function can be used to find out how many capturing
1512 nigel 63 subpatterns there are in a compiled pattern. The smallest size for
1513 nigel 75 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
1514     the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
1515 nigel 91 .P
1516     It is possible for capturing subpattern number \fIn+1\fP to match some part of
1517     the subject when subpattern \fIn\fP has not been used at all. For example, if
1518     the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1519     function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1520     happens, both values in the offset pairs corresponding to unused subpatterns
1521     are set to -1.
1522     .P
1523     Offset values that correspond to unused subpatterns at the end of the
1524     expression are also set to -1. For example, if the string "abc" is matched
1525     against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1526     return from the function is 2, because the highest used capturing subpattern
1527     number is 1. However, you can refer to the offsets for the second and third
1528     capturing subpatterns if you wish (assuming the vector is large enough, of
1529     course).
1530     .P
1531     Some convenience functions are provided for extracting the captured substrings
1532     as separate strings. These are described below.
1533 nigel 75 .
1534 nigel 77 .\" HTML <a name="errorlist"></a>
1535 nigel 91 .SS "Error return values from \fBpcre_exec()\fP"
1536 nigel 75 .rs
1537     .sp
1538     If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1539 nigel 63 defined in the header file:
1540 nigel 75 .sp
1541 nigel 63 PCRE_ERROR_NOMATCH (-1)
1542 nigel 75 .sp
1543 nigel 63 The subject string did not match the pattern.
1544 nigel 75 .sp
1545 nigel 63 PCRE_ERROR_NULL (-2)
1546 nigel 75 .sp
1547     Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
1548     NULL and \fIovecsize\fP was not zero.
1549     .sp
1550 nigel 63 PCRE_ERROR_BADOPTION (-3)
1551 nigel 75 .sp
1552     An unrecognized bit was set in the \fIoptions\fP argument.
1553     .sp
1554 nigel 63 PCRE_ERROR_BADMAGIC (-4)
1555 nigel 75 .sp
1556 nigel 63 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1557 nigel 75 the case when it is passed a junk pointer and to detect when a pattern that was
1558     compiled in an environment of one endianness is run in an environment with the
1559     other endianness. This is the error that PCRE gives when the magic number is
1560     not present.
1561     .sp
1562 nigel 93 PCRE_ERROR_UNKNOWN_OPCODE (-5)
1563 nigel 75 .sp
1564 nigel 63 While running the pattern match, an unknown item was encountered in the
1565     compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1566     of the compiled pattern.
1567 nigel 75 .sp
1568 nigel 63 PCRE_ERROR_NOMEMORY (-6)
1569 nigel 75 .sp
1570     If a pattern contains back references, but the \fIovector\fP that is passed to
1571     \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
1572 nigel 63 gets a block of memory at the start of matching to use for this purpose. If the
1573 nigel 75 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
1574     automatically freed at the end of matching.
1575     .sp
1576 nigel 63 PCRE_ERROR_NOSUBSTRING (-7)
1577 nigel 75 .sp
1578     This error is used by the \fBpcre_copy_substring()\fP,
1579     \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
1580     below). It is never returned by \fBpcre_exec()\fP.
1581     .sp
1582 nigel 63 PCRE_ERROR_MATCHLIMIT (-8)
1583 nigel 75 .sp
1584 nigel 87 The backtracking limit, as specified by the \fImatch_limit\fP field in a
1585     \fBpcre_extra\fP structure (or defaulted) was reached. See the description
1586     above.
1587     .sp
1588 nigel 63 PCRE_ERROR_CALLOUT (-9)
1589 nigel 75 .sp
1590     This error is never generated by \fBpcre_exec()\fP itself. It is provided for
1591 nigel 63 use by callout functions that want to yield a distinctive error code. See the
1592 nigel 75 .\" HREF
1593     \fBpcrecallout\fP
1594     .\"
1595     documentation for details.
1596     .sp
1597 nigel 73 PCRE_ERROR_BADUTF8 (-10)
1598 nigel 75 .sp
1599 nigel 71 A string that contains an invalid UTF-8 byte sequence was passed as a subject.
1600 nigel 75 .sp
1601 nigel 73 PCRE_ERROR_BADUTF8_OFFSET (-11)
1602 nigel 75 .sp
1603 nigel 73 The UTF-8 byte sequence that was passed as a subject was valid, but the value
1604 nigel 75 of \fIstartoffset\fP did not point to the beginning of a UTF-8 character.
1605     .sp
1606 nigel 77 PCRE_ERROR_PARTIAL (-12)
1607 nigel 75 .sp
1608     The subject string did not match, but it did match partially. See the
1609     .\" HREF
1610     \fBpcrepartial\fP
1611     .\"
1612     documentation for details of partial matching.
1613     .sp
1614 nigel 77 PCRE_ERROR_BADPARTIAL (-13)
1615 nigel 75 .sp
1616 ph10 426 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1617     option was used with a compiled pattern containing items that were not
1618     supported for partial matching. From release 8.00 onwards, there are no
1619     restrictions on partial matching.
1620 nigel 75 .sp
1621 nigel 77 PCRE_ERROR_INTERNAL (-14)
1622 nigel 75 .sp
1623     An unexpected internal error has occurred. This error could be caused by a bug
1624     in PCRE or by overwriting of the compiled pattern.
1625     .sp
1626 nigel 77 PCRE_ERROR_BADCOUNT (-15)
1627 nigel 75 .sp
1628     This error is given if the value of the \fIovecsize\fP argument is negative.
1629 nigel 93 .sp
1630     PCRE_ERROR_RECURSIONLIMIT (-21)
1631     .sp
1632     The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
1633     field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
1634     description above.
1635     .sp
1636     PCRE_ERROR_BADNEWLINE (-23)
1637     .sp
1638     An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
1639     .P
1640 ph10 197 Error numbers -16 to -20 and -22 are not used by \fBpcre_exec()\fP.
1641 nigel 75 .
1642     .
1643     .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
1644 nigel 63 .rs
1645     .sp
1646 nigel 75 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1647 nigel 63 .ti +5n
1648 nigel 75 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
1649 nigel 63 .ti +5n
1650 nigel 75 .B int \fIbuffersize\fP);
1651 nigel 63 .PP
1652 nigel 75 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1653 nigel 63 .ti +5n
1654 nigel 75 .B int \fIstringcount\fP, int \fIstringnumber\fP,
1655 nigel 63 .ti +5n
1656 nigel 75 .B const char **\fIstringptr\fP);
1657 nigel 63 .PP
1658 nigel 75 .B int pcre_get_substring_list(const char *\fIsubject\fP,
1659 nigel 63 .ti +5n
1660 nigel 75 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
1661 nigel 63 .PP
1662     Captured substrings can be accessed directly by using the offsets returned by
1663 nigel 75 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
1664     \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
1665     \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
1666 nigel 63 as new, separate, zero-terminated strings. These functions identify substrings
1667     by number. The next section describes functions for extracting named
1668 nigel 91 substrings.
1669 nigel 75 .P
1670 nigel 91 A substring that contains a binary zero is correctly extracted and has a
1671     further zero added on the end, but the result is not, of course, a C string.
1672     However, you can process such a string by referring to the length that is
1673     returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
1674     Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
1675     for handling strings containing binary zeros, because the end of the final
1676     string is not independently indicated.
1677     .P
1678 nigel 63 The first three arguments are the same for all three of these functions:
1679 nigel 75 \fIsubject\fP is the subject string that has just been successfully matched,
1680     \fIovector\fP is a pointer to the vector of integer offsets that was passed to
1681     \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
1682 nigel 63 captured by the match, including the substring that matched the entire regular
1683 nigel 75 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
1684     than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
1685     space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
1686     number of elements in the vector divided by three.
1687     .P
1688     The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
1689     extract a single substring, whose number is given as \fIstringnumber\fP. A
1690     value of zero extracts the substring that matched the entire pattern, whereas
1691     higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
1692     the string is placed in \fIbuffer\fP, whose length is given by
1693     \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
1694     obtained via \fBpcre_malloc\fP, and its address is returned via
1695     \fIstringptr\fP. The yield of the function is the length of the string, not
1696 nigel 93 including the terminating zero, or one of these error codes:
1697 nigel 75 .sp
1698 nigel 63 PCRE_ERROR_NOMEMORY (-6)
1699 nigel 75 .sp
1700     The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
1701     memory failed for \fBpcre_get_substring()\fP.
1702     .sp
1703 nigel 63 PCRE_ERROR_NOSUBSTRING (-7)
1704 nigel 75 .sp
1705     There is no substring whose number is \fIstringnumber\fP.
1706     .P
1707     The \fBpcre_get_substring_list()\fP function extracts all available substrings
1708 nigel 63 and builds a list of pointers to them. All this is done in a single block of
1709 nigel 75 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
1710     is returned via \fIlistptr\fP, which is also the start of the list of string
1711 nigel 63 pointers. The end of the list is marked by a NULL pointer. The yield of the
1712 nigel 93 function is zero if all went well, or the error code
1713 nigel 75 .sp
1714 nigel 63 PCRE_ERROR_NOMEMORY (-6)
1715 nigel 75 .sp
1716 nigel 63 if the attempt to get the memory block failed.
1717 nigel 75 .P
1718 nigel 63 When any of these functions encounter a substring that is unset, which can
1719 nigel 75 happen when capturing subpattern number \fIn+1\fP matches some part of the
1720     subject, but subpattern \fIn\fP has not been used at all, they return an empty
1721 nigel 63 string. This can be distinguished from a genuine zero-length substring by
1722 nigel 75 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
1723 nigel 63 substrings.
1724 nigel 75 .P
1725     The two convenience functions \fBpcre_free_substring()\fP and
1726     \fBpcre_free_substring_list()\fP can be used to free the memory returned by
1727     a previous call of \fBpcre_get_substring()\fP or
1728     \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
1729     the function pointed to by \fBpcre_free\fP, which of course could be called
1730 nigel 63 directly from a C program. However, PCRE is used in some situations where it is
1731 nigel 91 linked via a special interface to another programming language that cannot use
1732 nigel 75 \fBpcre_free\fP directly; it is for these cases that the functions are
1733 nigel 63 provided.
1734 nigel 75 .
1735     .
1736     .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
1737 nigel 63 .rs
1738     .sp
1739 nigel 75 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
1740 nigel 63 .ti +5n
1741 nigel 75 .B const char *\fIname\fP);
1742 nigel 63 .PP
1743 nigel 75 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
1744 nigel 63 .ti +5n
1745 nigel 75 .B const char *\fIsubject\fP, int *\fIovector\fP,
1746     .ti +5n
1747     .B int \fIstringcount\fP, const char *\fIstringname\fP,
1748     .ti +5n
1749     .B char *\fIbuffer\fP, int \fIbuffersize\fP);
1750 nigel 63 .PP
1751 nigel 75 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
1752 nigel 63 .ti +5n
1753 nigel 75 .B const char *\fIsubject\fP, int *\fIovector\fP,
1754 nigel 63 .ti +5n
1755 nigel 75 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1756 nigel 63 .ti +5n
1757 nigel 75 .B const char **\fIstringptr\fP);
1758 nigel 63 .PP
1759 nigel 75 To extract a substring by name, you first have to find associated number.
1760     For example, for this pattern
1761     .sp
1762 nigel 93 (a+)b(?<xxx>\ed+)...
1763 nigel 75 .sp
1764 nigel 91 the number of the subpattern called "xxx" is 2. If the name is known to be
1765     unique (PCRE_DUPNAMES was not set), you can find the number from the name by
1766     calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
1767     pattern, and the second is the name. The yield of the function is the
1768 nigel 75 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
1769     that name.
1770     .P
1771     Given the number, you can extract the substring directly, or use one of the
1772     functions described in the previous section. For convenience, there are also
1773     two functions that do the whole job.
1774     .P
1775 nigel 91 Most of the arguments of \fBpcre_copy_named_substring()\fP and
1776     \fBpcre_get_named_substring()\fP are the same as those for the similarly named
1777 nigel 75 functions that extract by number. As these are described in the previous
1778     section, they are not re-described here. There are just two differences:
1779     .P
1780 nigel 63 First, instead of a substring number, a substring name is given. Second, there
1781     is an extra argument, given at the start, which is a pointer to the compiled
1782     pattern. This is needed in order to gain access to the name-to-number
1783     translation table.
1784 nigel 75 .P
1785     These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
1786 ph10 127 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
1787     appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
1788 ph10 128 the behaviour may not be what you want (see the next section).
1789 ph10 385 .P
1790     \fBWarning:\fP If the pattern uses the "(?|" feature to set up multiple
1791 ph10 392 subpatterns with the same number, you cannot use names to distinguish them,
1792     because names are not included in the compiled code. The matching process uses
1793 ph10 385 only numbers.
1794 nigel 77 .
1795 nigel 91 .SH "DUPLICATE SUBPATTERN NAMES"
1796     .rs
1797     .sp
1798     .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
1799     .ti +5n
1800     .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
1801     .PP
1802     When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
1803     are not required to be unique. Normally, patterns with duplicate names are such
1804     that in any one match, only one of the named subpatterns participates. An
1805     example is shown in the
1806     .\" HREF
1807     \fBpcrepattern\fP
1808     .\"
1809 ph10 208 documentation.
1810 ph10 203 .P
1811     When duplicates are present, \fBpcre_copy_named_substring()\fP and
1812     \fBpcre_get_named_substring()\fP return the first substring corresponding to
1813     the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
1814     returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
1815     returns one of the numbers that are associated with the name, but it is not
1816     defined which it is.
1817     .P
1818 nigel 91 If you want to get full details of all captured substrings for a given name,
1819     you must use the \fBpcre_get_stringtable_entries()\fP function. The first
1820     argument is the compiled pattern, and the second is the name. The third and
1821     fourth are pointers to variables which are updated by the function. After it
1822     has run, they point to the first and last entries in the name-to-number table
1823     for the given name. The function itself returns the length of each entry, or
1824 nigel 93 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
1825     described above in the section entitled \fIInformation about a pattern\fP.
1826     Given all the relevant entries for the name, you can extract each of their
1827     numbers, and hence the captured data, if any.
1828 nigel 91 .
1829     .
1830 nigel 77 .SH "FINDING ALL POSSIBLE MATCHES"
1831     .rs
1832     .sp
1833     The traditional matching function uses a similar algorithm to Perl, which stops
1834     when it finds the first match, starting at a given point in the subject. If you
1835     want to find all possible matches, or the longest possible match, consider
1836     using the alternative matching function (see below) instead. If you cannot use
1837     the alternative function, but still need to find all possible matches, you
1838     can kludge it up by making use of the callout facility, which is described in
1839     the
1840     .\" HREF
1841     \fBpcrecallout\fP
1842     .\"
1843     documentation.
1844 nigel 75 .P
1845 nigel 77 What you have to do is to insert a callout right at the end of the pattern.
1846     When your callout function is called, extract and save the current matched
1847     substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
1848     other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
1849     will yield PCRE_ERROR_NOMATCH.
1850     .
1851     .
1852     .\" HTML <a name="dfamatch"></a>
1853     .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
1854     .rs
1855     .sp
1856     .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1857     .ti +5n
1858     .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1859     .ti +5n
1860     .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
1861     .ti +5n
1862     .B int *\fIworkspace\fP, int \fIwscount\fP);
1863     .P
1864     The function \fBpcre_dfa_exec()\fP is called to match a subject string against
1865 nigel 93 a compiled pattern, using a matching algorithm that scans the subject string
1866     just once, and does not backtrack. This has different characteristics to the
1867     normal algorithm, and is not compatible with Perl. Some of the features of PCRE
1868     patterns are not supported. Nevertheless, there are times when this kind of
1869 ph10 435 matching can be useful. For a discussion of the two matching algorithms, and a
1870     list of features that \fBpcre_dfa_exec()\fP does not support, see the
1871 nigel 77 .\" HREF
1872     \fBpcrematching\fP
1873     .\"
1874     documentation.
1875     .P
1876     The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
1877     \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
1878     different way, and this is described below. The other common arguments are used
1879     in the same way as for \fBpcre_exec()\fP, so their description is not repeated
1880     here.
1881     .P
1882     The two additional arguments provide workspace for the function. The workspace
1883     vector should contain at least 20 elements. It is used for keeping track of
1884     multiple paths through the pattern tree. More workspace will be needed for
1885 nigel 91 patterns and subjects where there are a lot of potential matches.
1886 nigel 77 .P
1887 nigel 87 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
1888 nigel 77 .sp
1889     int rc;
1890     int ovector[10];
1891     int wspace[20];
1892 nigel 87 rc = pcre_dfa_exec(
1893 nigel 77 re, /* result of pcre_compile() */
1894     NULL, /* we didn't study the pattern */
1895     "some string", /* the subject string */
1896     11, /* the length of the subject string */
1897     0, /* start at offset 0 in the subject */
1898     0, /* default options */
1899     ovector, /* vector of integers for substring information */
1900     10, /* number of elements (NOT size in bytes) */
1901     wspace, /* working space vector */
1902     20); /* number of elements (NOT size in bytes) */
1903     .
1904     .SS "Option bits for \fBpcre_dfa_exec()\fP"
1905     .rs
1906     .sp
1907     The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
1908 nigel 91 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1909 ph10 442 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1910     PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST,
1911     and PCRE_DFA_RESTART. All but the last four of these are exactly the same as
1912     for \fBpcre_exec()\fP, so their description is not repeated here.
1913 nigel 77 .sp
1914 ph10 428 PCRE_PARTIAL_HARD
1915     PCRE_PARTIAL_SOFT
1916 nigel 77 .sp
1917 ph10 428 These have the same general effect as they do for \fBpcre_exec()\fP, but the
1918     details are slightly different. When PCRE_PARTIAL_HARD is set for
1919     \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
1920     is reached and there is still at least one matching possibility that requires
1921     additional characters. This happens even if some complete matches have also
1922     been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
1923     is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
1924     there have been no complete matches, but there is still at least one matching
1925 ph10 435 possibility. The portion of the string that was inspected when the longest
1926     partial match was found is set as the first matching string in both cases.
1927 nigel 77 .sp
1928     PCRE_DFA_SHORTEST
1929     .sp
1930     Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
1931 nigel 93 soon as it has found one match. Because of the way the alternative algorithm
1932     works, this is necessarily the shortest possible match at the first possible
1933     matching point in the subject string.
1934 nigel 77 .sp
1935     PCRE_DFA_RESTART
1936     .sp
1937 ph10 428 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
1938     again, with additional subject characters, and have it continue with the same
1939     match. The PCRE_DFA_RESTART option requests this action; when it is set, the
1940     \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
1941     before because data about the match so far is left in them after a partial
1942     match. There is more discussion of this facility in the
1943 nigel 77 .\" HREF
1944     \fBpcrepartial\fP
1945     .\"
1946     documentation.
1947     .
1948     .SS "Successful returns from \fBpcre_dfa_exec()\fP"
1949     .rs
1950     .sp
1951     When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
1952     substring in the subject. Note, however, that all the matches from one run of
1953     the function start at the same point in the subject. The shorter matches are
1954     all initial substrings of the longer matches. For example, if the pattern
1955     .sp
1956     <.*>
1957     .sp
1958     is matched against the string
1959     .sp
1960     This is <something> <something else> <something further> no more
1961     .sp
1962     the three matched strings are
1963     .sp
1964     <something>
1965     <something> <something else>
1966     <something> <something else> <something further>
1967     .sp
1968     On success, the yield of the function is a number greater than zero, which is
1969     the number of matched substrings. The substrings themselves are returned in
1970     \fIovector\fP. Each string uses two elements; the first is the offset to the
1971 nigel 93 start, and the second is the offset to the end. In fact, all the strings have
1972     the same start offset. (Space could have been saved by giving this only once,
1973     but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
1974     returns data, even though the meaning of the strings is different.)
1975 nigel 77 .P
1976     The strings are returned in reverse order of length; that is, the longest
1977     matching string is given first. If there were too many matches to fit into
1978     \fIovector\fP, the yield of the function is zero, and the vector is filled with
1979     the longest matches.
1980     .
1981     .SS "Error returns from \fBpcre_dfa_exec()\fP"
1982     .rs
1983     .sp
1984     The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
1985     Many of the errors are the same as for \fBpcre_exec()\fP, and these are
1986     described
1987     .\" HTML <a href="#errorlist">
1988     .\" </a>
1989     above.
1990     .\"
1991     There are in addition the following errors that are specific to
1992     \fBpcre_dfa_exec()\fP:
1993     .sp
1994     PCRE_ERROR_DFA_UITEM (-16)
1995     .sp
1996     This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
1997     that it does not support, for instance, the use of \eC or a back reference.
1998     .sp
1999     PCRE_ERROR_DFA_UCOND (-17)
2000     .sp
2001 nigel 93 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2002     uses a back reference for the condition, or a test for recursion in a specific
2003     group. These are not supported.
2004 nigel 77 .sp
2005     PCRE_ERROR_DFA_UMLIMIT (-18)
2006     .sp
2007     This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2008     block that contains a setting of the \fImatch_limit\fP field. This is not
2009     supported (it is meaningless).
2010     .sp
2011     PCRE_ERROR_DFA_WSSIZE (-19)
2012     .sp
2013     This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2014     \fIworkspace\fP vector.
2015     .sp
2016     PCRE_ERROR_DFA_RECURSE (-20)
2017     .sp
2018     When a recursive subpattern is processed, the matching function calls itself
2019     recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2020     error is given if the output vector is not large enough. This should be
2021     extremely rare, as a vector of size 1000 is used.
2022 nigel 93 .
2023     .
2024     .SH "SEE ALSO"
2025     .rs
2026     .sp
2027     \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
2028     \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
2029     \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
2030 ph10 99 .
2031     .
2032     .SH AUTHOR
2033     .rs
2034     .sp
2035     .nf
2036     Philip Hazel
2037     University Computing Service
2038     Cambridge CB2 3QH, England.
2039     .fi
2040     .
2041     .
2042     .SH REVISION
2043     .rs
2044     .sp
2045     .nf
2046 ph10 455 Last updated: 26 September 2009
2047 ph10 376 Copyright (c) 1997-2009 University of Cambridge.
2048 ph10 99 .fi

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