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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 412 - (hide annotations) (download)
Sat Apr 11 10:34:37 2009 UTC (5 years, 5 months ago) by ph10
File size: 86077 byte(s)
Add support for (*UTF8).

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