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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1459 - (hide annotations) (download)
Tue Mar 4 10:45:15 2014 UTC (9 months, 2 weeks ago) by ph10
File size: 129199 byte(s)
Preparations for next release.

1 ph10 1454 .TH PCREAPI 3 "09 February 2014" "PCRE 8.35"
2 nigel 63 .SH NAME
3     PCRE - Perl-compatible regular expressions
4 ph10 856 .sp
5     .B #include <pcre.h>
6     .
7     .
8 ph10 678 .SH "PCRE NATIVE API BASIC FUNCTIONS"
9 nigel 63 .rs
10     .sp
11 ph10 1339 .nf
12 nigel 75 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
13 ph10 1339 .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
14     .B " const unsigned char *\fItableptr\fP);"
15     .sp
16 nigel 77 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
17 ph10 1339 .B " int *\fIerrorcodeptr\fP,"
18     .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
19     .B " const unsigned char *\fItableptr\fP);"
20     .sp
21 nigel 75 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
22 ph10 1339 .B " const char **\fIerrptr\fP);"
23     .sp
24 ph10 678 .B void pcre_free_study(pcre_extra *\fIextra\fP);
25 ph10 1339 .sp
26 nigel 75 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
27 ph10 1339 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
28     .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);"
29     .sp
30 nigel 77 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
31 ph10 1339 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
32     .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
33     .B " int *\fIworkspace\fP, int \fIwscount\fP);"
34     .fi
35 ph10 856 .
36     .
37     .SH "PCRE NATIVE API STRING EXTRACTION FUNCTIONS"
38     .rs
39     .sp
40 ph10 1339 .nf
41 nigel 75 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
42 ph10 1339 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
43     .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
44     .B " char *\fIbuffer\fP, int \fIbuffersize\fP);"
45     .sp
46 nigel 75 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
47 ph10 1339 .B " int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,"
48     .B " int \fIbuffersize\fP);"
49     .sp
50 nigel 75 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
51 ph10 1339 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
52     .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
53     .B " const char **\fIstringptr\fP);"
54     .sp
55 nigel 75 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
56 ph10 1339 .B " const char *\fIname\fP);"
57     .sp
58 nigel 91 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
59 ph10 1339 .B " const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);"
60     .sp
61 nigel 75 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
62 ph10 1339 .B " int \fIstringcount\fP, int \fIstringnumber\fP,"
63     .B " const char **\fIstringptr\fP);"
64     .sp
65 nigel 75 .B int pcre_get_substring_list(const char *\fIsubject\fP,
66 ph10 1339 .B " int *\fIovector\fP, int \fIstringcount\fP, const char ***\fIlistptr\fP);"
67     .sp
68 nigel 75 .B void pcre_free_substring(const char *\fIstringptr\fP);
69 ph10 1339 .sp
70 nigel 75 .B void pcre_free_substring_list(const char **\fIstringptr\fP);
71 ph10 1339 .fi
72 ph10 856 .
73     .
74     .SH "PCRE NATIVE API AUXILIARY FUNCTIONS"
75     .rs
76     .sp
77 ph10 1339 .nf
78 ph10 1194 .B int pcre_jit_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
79 ph10 1339 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
80     .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
81     .B " pcre_jit_stack *\fIjstack\fP);"
82     .sp
83 ph10 856 .B pcre_jit_stack *pcre_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
84 ph10 1339 .sp
85 ph10 856 .B void pcre_jit_stack_free(pcre_jit_stack *\fIstack\fP);
86 ph10 1339 .sp
87 ph10 856 .B void pcre_assign_jit_stack(pcre_extra *\fIextra\fP,
88 ph10 1339 .B " pcre_jit_callback \fIcallback\fP, void *\fIdata\fP);"
89     .sp
90 nigel 63 .B const unsigned char *pcre_maketables(void);
91 ph10 1339 .sp
92 nigel 75 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
93 ph10 1339 .B " int \fIwhat\fP, void *\fIwhere\fP);"
94     .sp
95 nigel 77 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
96 ph10 1339 .sp
97 nigel 75 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
98 ph10 1339 .sp
99 ph10 856 .B const char *pcre_version(void);
100 ph10 1339 .sp
101 ph10 856 .B int pcre_pattern_to_host_byte_order(pcre *\fIcode\fP,
102 ph10 1339 .B " pcre_extra *\fIextra\fP, const unsigned char *\fItables\fP);"
103     .fi
104 ph10 686 .
105     .
106     .SH "PCRE NATIVE API INDIRECTED FUNCTIONS"
107     .rs
108     .sp
109 ph10 1339 .nf
110 nigel 63 .B void *(*pcre_malloc)(size_t);
111 ph10 1339 .sp
112 nigel 63 .B void (*pcre_free)(void *);
113 ph10 1339 .sp
114 nigel 73 .B void *(*pcre_stack_malloc)(size_t);
115 ph10 1339 .sp
116 nigel 73 .B void (*pcre_stack_free)(void *);
117 ph10 1339 .sp
118 nigel 63 .B int (*pcre_callout)(pcre_callout_block *);
119 ph10 1454 .sp
120     .B int (*pcre_stack_guard)(void);
121 ph10 1339 .fi
122 nigel 75 .
123     .
124 ph10 1214 .SH "PCRE 8-BIT, 16-BIT, AND 32-BIT LIBRARIES"
125 ph10 856 .rs
126     .sp
127 ph10 1214 As well as support for 8-bit character strings, PCRE also supports 16-bit
128     strings (from release 8.30) and 32-bit strings (from release 8.32), by means of
129     two additional libraries. They can be built as well as, or instead of, the
130     8-bit library. To avoid too much complication, this document describes the
131     8-bit versions of the functions, with only occasional references to the 16-bit
132     and 32-bit libraries.
133 ph10 856 .P
134 ph10 1191 The 16-bit and 32-bit functions operate in the same way as their 8-bit
135     counterparts; they just use different data types for their arguments and
136     results, and their names start with \fBpcre16_\fP or \fBpcre32_\fP instead of
137     \fBpcre_\fP. For every option that has UTF8 in its name (for example,
138     PCRE_UTF8), there are corresponding 16-bit and 32-bit names with UTF8 replaced
139     by UTF16 or UTF32, respectively. This facility is in fact just cosmetic; the
140     16-bit and 32-bit option names define the same bit values.
141 ph10 856 .P
142     References to bytes and UTF-8 in this document should be read as references to
143 ph10 1328 16-bit data units and UTF-16 when using the 16-bit library, or 32-bit data
144     units and UTF-32 when using the 32-bit library, unless specified otherwise.
145     More details of the specific differences for the 16-bit and 32-bit libraries
146     are given in the
147 ph10 856 .\" HREF
148     \fBpcre16\fP
149     .\"
150 ph10 1191 and
151 chpe 1055 .\" HREF
152     \fBpcre32\fP
153     .\"
154 ph10 1191 pages.
155 ph10 856 .
156 chpe 1055 .
157 nigel 75 .SH "PCRE API OVERVIEW"
158 nigel 63 .rs
159     .sp
160 nigel 93 PCRE has its own native API, which is described in this document. There are
161 ph10 856 also some wrapper functions (for the 8-bit library only) that correspond to the
162     POSIX regular expression API, but they do not give access to all the
163     functionality. They are described in the
164 nigel 75 .\" HREF
165     \fBpcreposix\fP
166     .\"
167 nigel 77 documentation. Both of these APIs define a set of C function calls. A C++
168 ph10 856 wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
169     documented in the
170 nigel 77 .\" HREF
171     \fBpcrecpp\fP
172     .\"
173     page.
174 nigel 75 .P
175 nigel 77 The native API C function prototypes are defined in the header file
176 ph10 856 \fBpcre.h\fP, and on Unix-like systems the (8-bit) library itself is called
177     \fBlibpcre\fP. It can normally be accessed by adding \fB-lpcre\fP to the
178     command for linking an application that uses PCRE. The header file defines the
179     macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
180     for the library. Applications can use these to include support for different
181     releases of PCRE.
182 nigel 75 .P
183 ph10 525 In a Windows environment, if you want to statically link an application program
184     against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
185 ph10 526 including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
186     \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
187 ph10 525 \fB__declspec(dllimport)\fP, with unwanted results.
188     .P
189 nigel 77 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
190     and \fBpcre_exec()\fP are used for compiling and matching regular expressions
191     in a Perl-compatible manner. A sample program that demonstrates the simplest
192 ph10 429 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
193     source distribution. A listing of this program is given in the
194 nigel 75 .\" HREF
195 ph10 429 \fBpcredemo\fP
196     .\"
197     documentation, and the
198     .\" HREF
199 nigel 75 \fBpcresample\fP
200     .\"
201 ph10 312 documentation describes how to compile and run it.
202 nigel 75 .P
203 ph10 678 Just-in-time compiler support is an optional feature of PCRE that can be built
204 ph10 686 in appropriate hardware environments. It greatly speeds up the matching
205     performance of many patterns. Simple programs can easily request that it be
206     used if available, by setting an option that is ignored when it is not
207     relevant. More complicated programs might need to make use of the functions
208     \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
209     \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
210 ph10 1191 .P
211     From release 8.32 there is also a direct interface for JIT execution, which
212     gives improved performance. The JIT-specific functions are discussed in the
213 ph10 678 .\" HREF
214     \fBpcrejit\fP
215     .\"
216     documentation.
217     .P
218 nigel 77 A second matching function, \fBpcre_dfa_exec()\fP, which is not
219     Perl-compatible, is also provided. This uses a different algorithm for the
220 nigel 91 matching. The alternative algorithm finds all possible matches (at a given
221 ph10 435 point in the subject), and scans the subject just once (unless there are
222     lookbehind assertions). However, this algorithm does not return captured
223     substrings. A description of the two matching algorithms and their advantages
224     and disadvantages is given in the
225 nigel 77 .\" HREF
226     \fBpcrematching\fP
227     .\"
228     documentation.
229     .P
230 nigel 75 In addition to the main compiling and matching functions, there are convenience
231 nigel 77 functions for extracting captured substrings from a subject string that is
232     matched by \fBpcre_exec()\fP. They are:
233 nigel 75 .sp
234     \fBpcre_copy_substring()\fP
235     \fBpcre_copy_named_substring()\fP
236     \fBpcre_get_substring()\fP
237     \fBpcre_get_named_substring()\fP
238     \fBpcre_get_substring_list()\fP
239     \fBpcre_get_stringnumber()\fP
240 nigel 91 \fBpcre_get_stringtable_entries()\fP
241 nigel 75 .sp
242     \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
243 nigel 63 provided, to free the memory used for extracted strings.
244 nigel 75 .P
245     The function \fBpcre_maketables()\fP is used to build a set of character tables
246 nigel 77 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
247     or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
248     specialist use. Most commonly, no special tables are passed, in which case
249     internal tables that are generated when PCRE is built are used.
250 nigel 75 .P
251     The function \fBpcre_fullinfo()\fP is used to find out information about a
252 ph10 856 compiled pattern. The function \fBpcre_version()\fP returns a pointer to a
253     string containing the version of PCRE and its date of release.
254 nigel 75 .P
255 nigel 77 The function \fBpcre_refcount()\fP maintains a reference count in a data block
256     containing a compiled pattern. This is provided for the benefit of
257     object-oriented applications.
258     .P
259 nigel 75 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
260     the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
261 nigel 63 respectively. PCRE calls the memory management functions via these variables,
262     so a calling program can replace them if it wishes to intercept the calls. This
263     should be done before calling any PCRE functions.
264 nigel 75 .P
265     The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
266 nigel 73 indirections to memory management functions. These special functions are used
267     only when PCRE is compiled to use the heap for remembering data, instead of
268 nigel 91 recursive function calls, when running the \fBpcre_exec()\fP function. See the
269     .\" HREF
270     \fBpcrebuild\fP
271     .\"
272     documentation for details of how to do this. It is a non-standard way of
273     building PCRE, for use in environments that have limited stacks. Because of the
274     greater use of memory management, it runs more slowly. Separate functions are
275     provided so that special-purpose external code can be used for this case. When
276     used, these functions are always called in a stack-like manner (last obtained,
277     first freed), and always for memory blocks of the same size. There is a
278     discussion about PCRE's stack usage in the
279     .\" HREF
280     \fBpcrestack\fP
281     .\"
282     documentation.
283 nigel 75 .P
284     The global variable \fBpcre_callout\fP initially contains NULL. It can be set
285 nigel 63 by the caller to a "callout" function, which PCRE will then call at specified
286 nigel 75 points during a matching operation. Details are given in the
287     .\" HREF
288     \fBpcrecallout\fP
289     .\"
290 nigel 63 documentation.
291 ph10 1454 .P
292 ph10 1459 The global variable \fBpcre_stack_guard\fP initially contains NULL. It can be
293     set by the caller to a function that is called by PCRE whenever it starts
294     to compile a parenthesized part of a pattern. When parentheses are nested, PCRE
295     uses recursive function calls, which use up the system stack. This function is
296     provided so that applications with restricted stacks can force a compilation
297     error if the stack runs out. The function should return zero if all is well, or
298 ph10 1454 non-zero to force an error.
299 nigel 75 .
300     .
301 ph10 227 .\" HTML <a name="newlines"></a>
302 nigel 91 .SH NEWLINES
303 nigel 93 .rs
304 nigel 91 .sp
305 ph10 149 PCRE supports five different conventions for indicating line breaks in
306 nigel 93 strings: a single CR (carriage return) character, a single LF (linefeed)
307 ph10 149 character, the two-character sequence CRLF, any of the three preceding, or any
308     Unicode newline sequence. The Unicode newline sequences are the three just
309 ph10 968 mentioned, plus the single characters VT (vertical tab, U+000B), FF (form feed,
310 ph10 149 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
311     (paragraph separator, U+2029).
312 nigel 93 .P
313     Each of the first three conventions is used by at least one operating system as
314     its standard newline sequence. When PCRE is built, a default can be specified.
315     The default default is LF, which is the Unix standard. When PCRE is run, the
316     default can be overridden, either when a pattern is compiled, or when it is
317     matched.
318     .P
319 ph10 227 At compile time, the newline convention can be specified by the \fIoptions\fP
320     argument of \fBpcre_compile()\fP, or it can be specified by special text at the
321     start of the pattern itself; this overrides any other settings. See the
322     .\" HREF
323     \fBpcrepattern\fP
324     .\"
325     page for details of the special character sequences.
326     .P
327 nigel 91 In the PCRE documentation the word "newline" is used to mean "the character or
328 nigel 93 pair of characters that indicate a line break". The choice of newline
329     convention affects the handling of the dot, circumflex, and dollar
330     metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
331     recognized line ending sequence, the match position advancement for a
332 ph10 226 non-anchored pattern. There is more detail about this in the
333     .\" HTML <a href="#execoptions">
334     .\" </a>
335     section on \fBpcre_exec()\fP options
336     .\"
337 ph10 231 below.
338     .P
339     The choice of newline convention does not affect the interpretation of
340     the \en or \er escape sequences, nor does it affect what \eR matches, which is
341     controlled in a similar way, but by separate options.
342 nigel 91 .
343     .
344 nigel 63 .SH MULTITHREADING
345     .rs
346     .sp
347     The PCRE functions can be used in multi-threading applications, with the
348 nigel 75 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
349     \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
350 ph10 1459 callout and stack-checking functions pointed to by \fBpcre_callout\fP and
351 ph10 1454 \fBpcre_stack_guard\fP, are shared by all threads.
352 nigel 75 .P
353 nigel 63 The compiled form of a regular expression is not altered during matching, so
354     the same compiled pattern can safely be used by several threads at once.
355 ph10 678 .P
356 ph10 691 If the just-in-time optimization feature is being used, it needs separate
357 ph10 678 memory stack areas for each thread. See the
358     .\" HREF
359     \fBpcrejit\fP
360     .\"
361     documentation for more details.
362 nigel 75 .
363     .
364     .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
365 nigel 63 .rs
366     .sp
367 nigel 75 The compiled form of a regular expression can be saved and re-used at a later
368     time, possibly by a different program, and even on a host other than the one on
369     which it was compiled. Details are given in the
370     .\" HREF
371     \fBpcreprecompile\fP
372     .\"
373 ph10 856 documentation, which includes a description of the
374     \fBpcre_pattern_to_host_byte_order()\fP function. However, compiling a regular
375     expression with one version of PCRE for use with a different version is not
376     guaranteed to work and may cause crashes.
377 nigel 75 .
378     .
379     .SH "CHECKING BUILD-TIME OPTIONS"
380     .rs
381     .sp
382     .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
383 nigel 63 .PP
384 nigel 75 The function \fBpcre_config()\fP makes it possible for a PCRE client to
385 nigel 63 discover which optional features have been compiled into the PCRE library. The
386     .\" HREF
387 nigel 75 \fBpcrebuild\fP
388 nigel 63 .\"
389     documentation has more details about these optional features.
390 nigel 75 .P
391     The first argument for \fBpcre_config()\fP is an integer, specifying which
392 nigel 63 information is required; the second argument is a pointer to a variable into
393 ph10 856 which the information is placed. The returned value is zero on success, or the
394     negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
395     not recognized. The following information is available:
396 nigel 75 .sp
397 nigel 63 PCRE_CONFIG_UTF8
398 nigel 75 .sp
399 nigel 63 The output is an integer that is set to one if UTF-8 support is available;
400 chpe 1055 otherwise it is set to zero. This value should normally be given to the 8-bit
401     version of this function, \fBpcre_config()\fP. If it is given to the 16-bit
402     or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
403 nigel 75 .sp
404 ph10 856 PCRE_CONFIG_UTF16
405     .sp
406     The output is an integer that is set to one if UTF-16 support is available;
407     otherwise it is set to zero. This value should normally be given to the 16-bit
408     version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
409 chpe 1055 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
410 ph10 856 .sp
411 chpe 1055 PCRE_CONFIG_UTF32
412     .sp
413     The output is an integer that is set to one if UTF-32 support is available;
414     otherwise it is set to zero. This value should normally be given to the 32-bit
415     version of this function, \fBpcre32_config()\fP. If it is given to the 8-bit
416     or 16-bit version of this function, the result is PCRE_ERROR_BADOPTION.
417     .sp
418 nigel 75 PCRE_CONFIG_UNICODE_PROPERTIES
419     .sp
420     The output is an integer that is set to one if support for Unicode character
421     properties is available; otherwise it is set to zero.
422     .sp
423 ph10 678 PCRE_CONFIG_JIT
424     .sp
425 ph10 691 The output is an integer that is set to one if support for just-in-time
426     compiling is available; otherwise it is set to zero.
427 ph10 678 .sp
428 ph10 887 PCRE_CONFIG_JITTARGET
429     .sp
430     The output is a pointer to a zero-terminated "const char *" string. If JIT
431 ph10 903 support is available, the string contains the name of the architecture for
432     which the JIT compiler is configured, for example "x86 32bit (little endian +
433 ph10 887 unaligned)". If JIT support is not available, the result is NULL.
434     .sp
435 nigel 63 PCRE_CONFIG_NEWLINE
436 nigel 75 .sp
437 nigel 91 The output is an integer whose value specifies the default character sequence
438 ph10 1221 that is recognized as meaning "newline". The values that are supported in
439 ph10 1031 ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for
440     ANYCRLF, and -1 for ANY. In EBCDIC environments, CR, ANYCRLF, and ANY yield the
441     same values. However, the value for LF is normally 21, though some EBCDIC
442     environments use 37. The corresponding values for CRLF are 3349 and 3365. The
443     default should normally correspond to the standard sequence for your operating
444     system.
445 nigel 75 .sp
446 ph10 231 PCRE_CONFIG_BSR
447     .sp
448     The output is an integer whose value indicates what character sequences the \eR
449     escape sequence matches by default. A value of 0 means that \eR matches any
450     Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
451     or CRLF. The default can be overridden when a pattern is compiled or matched.
452     .sp
453 nigel 63 PCRE_CONFIG_LINK_SIZE
454 nigel 75 .sp
455 nigel 63 The output is an integer that contains the number of bytes used for internal
456 ph10 856 linkage in compiled regular expressions. For the 8-bit library, the value can
457     be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
458 chpe 1055 a number of bytes. For the 32-bit library, the value is either 2 or 4 and is
459     still a number of bytes. The default value of 2 is sufficient for all but the
460     most massive patterns, since it allows the compiled pattern to be up to 64K in
461     size. Larger values allow larger regular expressions to be compiled, at the
462     expense of slower matching.
463 nigel 75 .sp
464 nigel 63 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
465 nigel 75 .sp
466 nigel 63 The output is an integer that contains the threshold above which the POSIX
467 nigel 75 interface uses \fBmalloc()\fP for output vectors. Further details are given in
468     the
469     .\" HREF
470     \fBpcreposix\fP
471     .\"
472     documentation.
473     .sp
474 ph10 1389 PCRE_CONFIG_PARENS_LIMIT
475     .sp
476 ph10 1404 The output is a long integer that gives the maximum depth of nesting of
477     parentheses (of any kind) in a pattern. This limit is imposed to cap the amount
478 ph10 1389 of system stack used when a pattern is compiled. It is specified when PCRE is
479 ph10 1459 built; the default is 250. This limit does not take into account the stack that
480     may already be used by the calling application. For finer control over
481 ph10 1454 compilation stack usage, you can set a pointer to an external checking function
482     in \fBpcre_stack_guard\fP.
483 ph10 1389 .sp
484 nigel 63 PCRE_CONFIG_MATCH_LIMIT
485 nigel 75 .sp
486 ph10 376 The output is a long integer that gives the default limit for the number of
487 nigel 75 internal matching function calls in a \fBpcre_exec()\fP execution. Further
488     details are given with \fBpcre_exec()\fP below.
489     .sp
490 nigel 87 PCRE_CONFIG_MATCH_LIMIT_RECURSION
491     .sp
492 ph10 376 The output is a long integer that gives the default limit for the depth of
493 nigel 87 recursion when calling the internal matching function in a \fBpcre_exec()\fP
494     execution. Further details are given with \fBpcre_exec()\fP below.
495     .sp
496 nigel 73 PCRE_CONFIG_STACKRECURSE
497 nigel 75 .sp
498 nigel 77 The output is an integer that is set to one if internal recursion when running
499     \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
500     to remember their state. This is the usual way that PCRE is compiled. The
501     output is zero if PCRE was compiled to use blocks of data on the heap instead
502     of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
503     \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
504     avoiding the use of the stack.
505 nigel 75 .
506     .
507     .SH "COMPILING A PATTERN"
508 nigel 63 .rs
509     .sp
510 ph10 1339 .nf
511 nigel 75 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
512 ph10 1339 .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
513     .B " const unsigned char *\fItableptr\fP);"
514 nigel 77 .sp
515     .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
516 ph10 1339 .B " int *\fIerrorcodeptr\fP,"
517     .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
518     .B " const unsigned char *\fItableptr\fP);"
519     .fi
520 nigel 75 .P
521 nigel 77 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
522     called to compile a pattern into an internal form. The only difference between
523     the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
524 ph10 461 \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
525     too much repetition, we refer just to \fBpcre_compile()\fP below, but the
526 ph10 456 information applies equally to \fBpcre_compile2()\fP.
527 nigel 75 .P
528 nigel 77 The pattern is a C string terminated by a binary zero, and is passed in the
529     \fIpattern\fP argument. A pointer to a single block of memory that is obtained
530     via \fBpcre_malloc\fP is returned. This contains the compiled code and related
531     data. The \fBpcre\fP type is defined for the returned block; this is a typedef
532     for a structure whose contents are not externally defined. It is up to the
533 nigel 91 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
534 nigel 77 .P
535 nigel 63 Although the compiled code of a PCRE regex is relocatable, that is, it does not
536 nigel 75 depend on memory location, the complete \fBpcre\fP data block is not
537     fully relocatable, because it may contain a copy of the \fItableptr\fP
538     argument, which is an address (see below).
539     .P
540 nigel 93 The \fIoptions\fP argument contains various bit settings that affect the
541 nigel 75 compilation. It should be zero if no options are required. The available
542 ph10 412 options are described below. Some of them (in particular, those that are
543 ph10 456 compatible with Perl, but some others as well) can also be set and unset from
544 ph10 412 within the pattern (see the detailed description in the
545 nigel 75 .\" HREF
546     \fBpcrepattern\fP
547     .\"
548 ph10 412 documentation). For those options that can be different in different parts of
549 ph10 456 the pattern, the contents of the \fIoptions\fP argument specifies their
550 ph10 461 settings at the start of compilation and execution. The PCRE_ANCHORED,
551 ph10 576 PCRE_BSR_\fIxxx\fP, PCRE_NEWLINE_\fIxxx\fP, PCRE_NO_UTF8_CHECK, and
552 ph10 930 PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
553 ph10 576 compile time.
554 nigel 75 .P
555     If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
556     Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
557     NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
558 nigel 87 error message. This is a static string that is part of the library. You must
559 ph10 598 not try to free it. Normally, the offset from the start of the pattern to the
560 ph10 1328 data unit that was being processed when the error was discovered is placed in
561     the variable pointed to by \fIerroffset\fP, which must not be NULL (if it is,
562     an immediate error is given). However, for an invalid UTF-8 or UTF-16 string,
563     the offset is that of the first data unit of the failing character.
564 nigel 75 .P
565 ph10 856 Some errors are not detected until the whole pattern has been scanned; in these
566     cases, the offset passed back is the length of the pattern. Note that the
567 ph10 1328 offset is in data units, not characters, even in a UTF mode. It may sometimes
568     point into the middle of a UTF-8 or UTF-16 character.
569 ph10 569 .P
570 nigel 77 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
571     \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
572     returned via this argument in the event of an error. This is in addition to the
573     textual error message. Error codes and messages are listed below.
574     .P
575 nigel 75 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
576     character tables that are built when PCRE is compiled, using the default C
577     locale. Otherwise, \fItableptr\fP must be an address that is the result of a
578     call to \fBpcre_maketables()\fP. This value is stored with the compiled
579 ph10 1400 pattern, and used again by \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP when the
580     pattern is matched. For more discussion, see the section on locale support
581     below.
582 nigel 75 .P
583     This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
584     .sp
585 nigel 63 pcre *re;
586     const char *error;
587     int erroffset;
588     re = pcre_compile(
589     "^A.*Z", /* the pattern */
590     0, /* default options */
591     &error, /* for error message */
592     &erroffset, /* for error offset */
593     NULL); /* use default character tables */
594 nigel 75 .sp
595     The following names for option bits are defined in the \fBpcre.h\fP header
596     file:
597     .sp
598 nigel 63 PCRE_ANCHORED
599 nigel 75 .sp
600 nigel 63 If this bit is set, the pattern is forced to be "anchored", that is, it is
601 nigel 75 constrained to match only at the first matching point in the string that is
602 nigel 63 being searched (the "subject string"). This effect can also be achieved by
603     appropriate constructs in the pattern itself, which is the only way to do it in
604     Perl.
605 nigel 75 .sp
606     PCRE_AUTO_CALLOUT
607     .sp
608     If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
609     all with number 255, before each pattern item. For discussion of the callout
610     facility, see the
611     .\" HREF
612     \fBpcrecallout\fP
613     .\"
614     documentation.
615     .sp
616 ph10 231 PCRE_BSR_ANYCRLF
617     PCRE_BSR_UNICODE
618     .sp
619     These options (which are mutually exclusive) control what the \eR escape
620     sequence matches. The choice is either to match only CR, LF, or CRLF, or to
621     match any Unicode newline sequence. The default is specified when PCRE is
622     built. It can be overridden from within the pattern, or by setting an option
623     when a compiled pattern is matched.
624     .sp
625 nigel 63 PCRE_CASELESS
626 nigel 75 .sp
627 nigel 63 If this bit is set, letters in the pattern match both upper and lower case
628     letters. It is equivalent to Perl's /i option, and it can be changed within a
629 nigel 77 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
630     concept of case for characters whose values are less than 128, so caseless
631     matching is always possible. For characters with higher values, the concept of
632     case is supported if PCRE is compiled with Unicode property support, but not
633     otherwise. If you want to use caseless matching for characters 128 and above,
634     you must ensure that PCRE is compiled with Unicode property support as well as
635     with UTF-8 support.
636 nigel 75 .sp
637 nigel 63 PCRE_DOLLAR_ENDONLY
638 nigel 75 .sp
639 nigel 63 If this bit is set, a dollar metacharacter in the pattern matches only at the
640     end of the subject string. Without this option, a dollar also matches
641 nigel 91 immediately before a newline at the end of the string (but not before any other
642     newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
643     There is no equivalent to this option in Perl, and no way to set it within a
644     pattern.
645 nigel 75 .sp
646 nigel 63 PCRE_DOTALL
647 nigel 75 .sp
648 ph10 572 If this bit is set, a dot metacharacter in the pattern matches a character of
649     any value, including one that indicates a newline. However, it only ever
650     matches one character, even if newlines are coded as CRLF. Without this option,
651     a dot does not match when the current position is at a newline. This option is
652     equivalent to Perl's /s option, and it can be changed within a pattern by a
653     (?s) option setting. A negative class such as [^a] always matches newline
654     characters, independent of the setting of this option.
655 nigel 75 .sp
656 nigel 91 PCRE_DUPNAMES
657     .sp
658     If this bit is set, names used to identify capturing subpatterns need not be
659     unique. This can be helpful for certain types of pattern when it is known that
660     only one instance of the named subpattern can ever be matched. There are more
661     details of named subpatterns below; see also the
662     .\" HREF
663     \fBpcrepattern\fP
664     .\"
665     documentation.
666     .sp
667 nigel 63 PCRE_EXTENDED
668 nigel 75 .sp
669 ph10 1396 If this bit is set, most white space characters in the pattern are totally
670     ignored except when escaped or inside a character class. However, white space
671     is not allowed within sequences such as (?> that introduce various
672     parenthesized subpatterns, nor within a numerical quantifier such as {1,3}.
673     However, ignorable white space is permitted between an item and a following
674 ph10 1404 quantifier and between a quantifier and a following + that indicates
675 ph10 1396 possessiveness.
676 nigel 75 .P
677 ph10 1396 White space did not used to include the VT character (code 11), because Perl
678     did not treat this character as white space. However, Perl changed at release
679     5.18, so PCRE followed at release 8.34, and VT is now treated as white space.
680     .P
681     PCRE_EXTENDED also causes characters between an unescaped # outside a character
682     class and the next newline, inclusive, to be ignored. PCRE_EXTENDED is
683     equivalent to Perl's /x option, and it can be changed within a pattern by a
684     (?x) option setting.
685     .P
686 ph10 598 Which characters are interpreted as newlines is controlled by the options
687     passed to \fBpcre_compile()\fP or by a special sequence at the start of the
688     pattern, as described in the section entitled
689 ph10 572 .\" HTML <a href="pcrepattern.html#newlines">
690     .\" </a>
691     "Newline conventions"
692     .\"
693     in the \fBpcrepattern\fP documentation. Note that the end of this type of
694     comment is a literal newline sequence in the pattern; escape sequences that
695     happen to represent a newline do not count.
696     .P
697 nigel 63 This option makes it possible to include comments inside complicated patterns.
698 ph10 968 Note, however, that this applies only to data characters. White space characters
699 nigel 63 may never appear within special character sequences in a pattern, for example
700 ph10 572 within the sequence (?( that introduces a conditional subpattern.
701 nigel 75 .sp
702 nigel 63 PCRE_EXTRA
703 nigel 75 .sp
704 nigel 63 This option was invented in order to turn on additional functionality of PCRE
705     that is incompatible with Perl, but it is currently of very little use. When
706     set, any backslash in a pattern that is followed by a letter that has no
707     special meaning causes an error, thus reserving these combinations for future
708     expansion. By default, as in Perl, a backslash followed by a letter with no
709 nigel 91 special meaning is treated as a literal. (Perl can, however, be persuaded to
710 ph10 513 give an error for this, by running it with the -w option.) There are at present
711     no other features controlled by this option. It can also be set by a (?X)
712     option setting within a pattern.
713 nigel 75 .sp
714 nigel 77 PCRE_FIRSTLINE
715     .sp
716     If this option is set, an unanchored pattern is required to match before or at
717 nigel 91 the first newline in the subject string, though the matched text may continue
718     over the newline.
719 nigel 77 .sp
720 ph10 336 PCRE_JAVASCRIPT_COMPAT
721     .sp
722 ph10 345 If this option is set, PCRE's behaviour is changed in some ways so that it is
723 ph10 336 compatible with JavaScript rather than Perl. The changes are as follows:
724     .P
725     (1) A lone closing square bracket in a pattern causes a compile-time error,
726     because this is illegal in JavaScript (by default it is treated as a data
727     character). Thus, the pattern AB]CD becomes illegal when this option is set.
728     .P
729     (2) At run time, a back reference to an unset subpattern group matches an empty
730 ph10 345 string (by default this causes the current matching alternative to fail). A
731     pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
732     an "a" in the subject), whereas it fails by default, for Perl compatibility.
733 ph10 745 .P
734 ph10 836 (3) \eU matches an upper case "U" character; by default \eU causes a compile
735 ph10 745 time error (Perl uses \eU to upper case subsequent characters).
736     .P
737 ph10 836 (4) \eu matches a lower case "u" character unless it is followed by four
738     hexadecimal digits, in which case the hexadecimal number defines the code point
739     to match. By default, \eu causes a compile time error (Perl uses it to upper
740 ph10 745 case the following character).
741     .P
742 ph10 836 (5) \ex matches a lower case "x" character unless it is followed by two
743     hexadecimal digits, in which case the hexadecimal number defines the code point
744     to match. By default, as in Perl, a hexadecimal number is always expected after
745     \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
746 ph10 745 binary zero character followed by z).
747 ph10 336 .sp
748 nigel 63 PCRE_MULTILINE
749 nigel 75 .sp
750 ph10 1326 By default, for the purposes of matching "start of line" and "end of line",
751     PCRE treats the subject string as consisting of a single line of characters,
752     even if it actually contains newlines. The "start of line" metacharacter (^)
753     matches only at the start of the string, and the "end of line" metacharacter
754     ($) matches only at the end of the string, or before a terminating newline
755     (except when PCRE_DOLLAR_ENDONLY is set). Note, however, that unless
756     PCRE_DOTALL is set, the "any character" metacharacter (.) does not match at a
757     newline. This behaviour (for ^, $, and dot) is the same as Perl.
758 nigel 75 .P
759 nigel 63 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
760 nigel 91 match immediately following or immediately before internal newlines in the
761     subject string, respectively, as well as at the very start and end. This is
762     equivalent to Perl's /m option, and it can be changed within a pattern by a
763     (?m) option setting. If there are no newlines in a subject string, or no
764 nigel 63 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
765 nigel 75 .sp
766 ph10 1309 PCRE_NEVER_UTF
767     .sp
768     This option locks out interpretation of the pattern as UTF-8 (or UTF-16 or
769     UTF-32 in the 16-bit and 32-bit libraries). In particular, it prevents the
770     creator of the pattern from switching to UTF interpretation by starting the
771     pattern with (*UTF). This may be useful in applications that process patterns
772 ph10 1314 from external sources. The combination of PCRE_UTF8 and PCRE_NEVER_UTF also
773 ph10 1309 causes an error.
774     .sp
775 nigel 91 PCRE_NEWLINE_CR
776     PCRE_NEWLINE_LF
777     PCRE_NEWLINE_CRLF
778 ph10 150 PCRE_NEWLINE_ANYCRLF
779 nigel 93 PCRE_NEWLINE_ANY
780 nigel 91 .sp
781     These options override the default newline definition that was chosen when PCRE
782     was built. Setting the first or the second specifies that a newline is
783 nigel 93 indicated by a single character (CR or LF, respectively). Setting
784     PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
785 ph10 149 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
786     preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
787 ph10 1221 that any Unicode newline sequence should be recognized.
788 nigel 91 .P
789 ph10 1031 In an ASCII/Unicode environment, the Unicode newline sequences are the three
790     just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
791     feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
792     (paragraph separator, U+2029). For the 8-bit library, the last two are
793     recognized only in UTF-8 mode.
794     .P
795     When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
796     CR is 0x0d, the same as ASCII. However, the character code for LF is normally
797 ph10 1221 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
798     not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
799 ph10 1031 less than 256. For more details, see the
800     .\" HREF
801     \fBpcrebuild\fP
802     .\"
803     documentation.
804     .P
805 nigel 93 The newline setting in the options word uses three bits that are treated
806 ph10 149 as a number, giving eight possibilities. Currently only six are used (default
807     plus the five values above). This means that if you set more than one newline
808 nigel 93 option, the combination may or may not be sensible. For example,
809     PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
810 ph10 149 other combinations may yield unused numbers and cause an error.
811 nigel 93 .P
812 ph10 572 The only time that a line break in a pattern is specially recognized when
813 ph10 968 compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
814 ph10 572 and so are ignored in this mode. Also, an unescaped # outside a character class
815     indicates a comment that lasts until after the next line break sequence. In
816     other circumstances, line break sequences in patterns are treated as literal
817     data.
818 nigel 93 .P
819     The newline option that is set at compile time becomes the default that is used
820     for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
821 nigel 91 .sp
822 nigel 63 PCRE_NO_AUTO_CAPTURE
823 nigel 75 .sp
824 nigel 63 If this option is set, it disables the use of numbered capturing parentheses in
825     the pattern. Any opening parenthesis that is not followed by ? behaves as if it
826     were followed by ?: but named parentheses can still be used for capturing (and
827     they acquire numbers in the usual way). There is no equivalent of this option
828     in Perl.
829 nigel 75 .sp
830 ph10 1398 PCRE_NO_AUTO_POSSESS
831 ph10 1363 .sp
832 ph10 1404 If this option is set, it disables "auto-possessification". This is an
833 ph10 1363 optimization that, for example, turns a+b into a++b in order to avoid
834 ph10 1395 backtracks into a+ that can never be successful. However, if callouts are in
835     use, auto-possessification means that some of them are never taken. You can set
836 ph10 1404 this option if you want the matching functions to do a full unoptimized search
837 ph10 1395 and run all the callouts, but it is mainly provided for testing purposes.
838 ph10 1363 .sp
839 ph10 1301 PCRE_NO_START_OPTIMIZE
840 ph10 576 .sp
841     This is an option that acts at matching time; that is, it is really an option
842     for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
843 ph10 1301 it is remembered with the compiled pattern and assumed at matching time. This
844     is necessary if you want to use JIT execution, because the JIT compiler needs
845     to know whether or not this option is set. For details see the discussion of
846     PCRE_NO_START_OPTIMIZE
847 ph10 576 .\" HTML <a href="#execoptions">
848     .\" </a>
849     below.
850     .\"
851     .sp
852 ph10 518 PCRE_UCP
853     .sp
854 ph10 572 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
855     \ew, and some of the POSIX character classes. By default, only ASCII characters
856     are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
857     classify characters. More details are given in the section on
858 ph10 518 .\" HTML <a href="pcre.html#genericchartypes">
859     .\" </a>
860     generic character types
861     .\"
862     in the
863     .\" HREF
864     \fBpcrepattern\fP
865     .\"
866     page. If you set PCRE_UCP, matching one of the items it affects takes much
867     longer. The option is available only if PCRE has been compiled with Unicode
868     property support.
869     .sp
870 nigel 63 PCRE_UNGREEDY
871 nigel 75 .sp
872 nigel 63 This option inverts the "greediness" of the quantifiers so that they are not
873     greedy by default, but become greedy if followed by "?". It is not compatible
874     with Perl. It can also be set by a (?U) option setting within the pattern.
875 nigel 75 .sp
876 nigel 63 PCRE_UTF8
877 nigel 75 .sp
878 nigel 63 This option causes PCRE to regard both the pattern and the subject as strings
879 ph10 856 of UTF-8 characters instead of single-byte strings. However, it is available
880     only when PCRE is built to include UTF support. If not, the use of this option
881     provokes an error. Details of how this option changes the behaviour of PCRE are
882     given in the
883 nigel 63 .\" HREF
884 ph10 678 \fBpcreunicode\fP
885 nigel 63 .\"
886     page.
887 nigel 75 .sp
888 nigel 71 PCRE_NO_UTF8_CHECK
889 nigel 75 .sp
890 ph10 1191 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
891     automatically checked. There is a discussion about the
892 ph10 856 .\" HTML <a href="pcreunicode.html#utf8strings">
893 ph10 211 .\" </a>
894 ph10 903 validity of UTF-8 strings
895 ph10 211 .\"
896 ph10 856 in the
897 ph10 211 .\" HREF
898 ph10 856 \fBpcreunicode\fP
899 ph10 211 .\"
900 ph10 856 page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
901     error. If you already know that your pattern is valid, and you want to skip
902     this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
903     When it is set, the effect of passing an invalid UTF-8 string as a pattern is
904 ph10 1363 undefined. It may cause your program to crash or loop. Note that this option
905     can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
906     the validity checking of subject strings only. If the same string is being
907     matched many times, the option can be safely set for the second and subsequent
908 ph10 1191 matchings to improve performance.
909 nigel 75 .
910     .
911 nigel 77 .SH "COMPILATION ERROR CODES"
912     .rs
913     .sp
914     The following table lists the error codes than may be returned by
915     \fBpcre_compile2()\fP, along with the error messages that may be returned by
916 ph10 903 both compiling functions. Note that error messages are always 8-bit ASCII
917 chpe 1055 strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
918     have fallen out of use. To avoid confusion, they have not been re-used.
919 nigel 77 .sp
920     0 no error
921     1 \e at end of pattern
922     2 \ec at end of pattern
923     3 unrecognized character follows \e
924     4 numbers out of order in {} quantifier
925     5 number too big in {} quantifier
926     6 missing terminating ] for character class
927     7 invalid escape sequence in character class
928     8 range out of order in character class
929     9 nothing to repeat
930 nigel 93 10 [this code is not in use]
931 nigel 77 11 internal error: unexpected repeat
932 ph10 290 12 unrecognized character after (? or (?-
933 nigel 77 13 POSIX named classes are supported only within a class
934     14 missing )
935     15 reference to non-existent subpattern
936     16 erroffset passed as NULL
937     17 unknown option bit(s) set
938     18 missing ) after comment
939 nigel 93 19 [this code is not in use]
940 ph10 290 20 regular expression is too large
941 nigel 77 21 failed to get memory
942     22 unmatched parentheses
943     23 internal error: code overflow
944     24 unrecognized character after (?<
945     25 lookbehind assertion is not fixed length
946 nigel 91 26 malformed number or name after (?(
947 nigel 77 27 conditional group contains more than two branches
948     28 assertion expected after (?(
949 ph10 181 29 (?R or (?[+-]digits must be followed by )
950 nigel 77 30 unknown POSIX class name
951     31 POSIX collating elements are not supported
952 ph10 856 32 this version of PCRE is compiled without UTF support
953 nigel 93 33 [this code is not in use]
954 ph10 1370 34 character value in \ex{} or \eo{} is too large
955 nigel 77 35 invalid condition (?(0)
956     36 \eC not allowed in lookbehind assertion
957 ph10 656 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
958 nigel 77 38 number after (?C is > 255
959     39 closing ) for (?C expected
960     40 recursive call could loop indefinitely
961     41 unrecognized character after (?P
962 nigel 93 42 syntax error in subpattern name (missing terminator)
963 nigel 91 43 two named subpatterns have the same name
964 ph10 856 44 invalid UTF-8 string (specifically UTF-8)
965 nigel 77 45 support for \eP, \ep, and \eX has not been compiled
966     46 malformed \eP or \ep sequence
967     47 unknown property name after \eP or \ep
968 nigel 91 48 subpattern name is too long (maximum 32 characters)
969 ph10 290 49 too many named subpatterns (maximum 10000)
970 ph10 202 50 [this code is not in use]
971 ph10 856 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
972 nigel 93 52 internal error: overran compiling workspace
973 ph10 548 53 internal error: previously-checked referenced subpattern
974 ph10 546 not found
975 nigel 93 54 DEFINE group contains more than one branch
976     55 repeating a DEFINE group is not allowed
977 ph10 231 56 inconsistent NEWLINE options
978 ph10 345 57 \eg is not followed by a braced, angle-bracketed, or quoted
979     name/number or by a plain number
980 ph10 336 58 a numbered reference must not be zero
981 ph10 510 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
982 ph10 1313 60 (*VERB) not recognized or malformed
983 ph10 290 61 number is too big
984     62 subpattern name expected
985 ph10 292 63 digit expected after (?+
986 ph10 336 64 ] is an invalid data character in JavaScript compatibility mode
987 ph10 548 65 different names for subpatterns of the same number are
988 ph10 546 not allowed
989 ph10 510 66 (*MARK) must have an argument
990 ph10 903 67 this version of PCRE is not compiled with Unicode property
991 ph10 856 support
992 ph10 656 68 \ec must be followed by an ASCII character
993 ph10 659 69 \ek is not followed by a braced, angle-bracketed, or quoted name
994 ph10 856 70 internal error: unknown opcode in find_fixedlength()
995     71 \eN is not supported in a class
996     72 too many forward references
997 ph10 903 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
998 ph10 856 74 invalid UTF-16 string (specifically UTF-16)
999 ph10 975 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
1000 ph10 978 76 character value in \eu.... sequence is too large
1001 chpe 1055 77 invalid UTF-32 string (specifically UTF-32)
1002 ph10 1370 78 setting UTF is disabled by the application
1003     79 non-hex character in \ex{} (closing brace missing?)
1004     80 non-octal character in \eo{} (closing brace missing?)
1005 ph10 1404 81 missing opening brace after \eo
1006 ph10 1392 82 parentheses are too deeply nested
1007     83 invalid range in character class
1008 ph10 1454 84 group name must start with a non-digit
1009 ph10 1459 85 parentheses are too deeply nested (stack check)
1010 ph10 290 .sp
1011 ph10 292 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
1012 ph10 290 be used if the limits were changed when PCRE was built.
1013 nigel 77 .
1014     .
1015 ph10 656 .\" HTML <a name="studyingapattern"></a>
1016 nigel 75 .SH "STUDYING A PATTERN"
1017 nigel 63 .rs
1018     .sp
1019 ph10 1339 .nf
1020     .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
1021     .B " const char **\fIerrptr\fP);"
1022     .fi
1023 nigel 63 .PP
1024 nigel 75 If a compiled pattern is going to be used several times, it is worth spending
1025     more time analyzing it in order to speed up the time taken for matching. The
1026     function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
1027     argument. If studying the pattern produces additional information that will
1028     help speed up matching, \fBpcre_study()\fP returns a pointer to a
1029     \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
1030     results of the study.
1031     .P
1032     The returned value from \fBpcre_study()\fP can be passed directly to
1033 ph10 455 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
1034     also contains other fields that can be set by the caller before the block is
1035     passed; these are described
1036 nigel 75 .\" HTML <a href="#extradata">
1037     .\" </a>
1038     below
1039     .\"
1040     in the section on matching a pattern.
1041     .P
1042 ph10 455 If studying the pattern does not produce any useful information,
1043 ph10 1022 \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
1044     calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
1045 ph10 1221 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
1046 ph10 1022 if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
1047     returns a \fBpcre_extra\fP block even if studying did not find any additional
1048     information. It may still return NULL, however, if an error occurs in
1049     \fBpcre_study()\fP.
1050 nigel 75 .P
1051 ph10 921 The second argument of \fBpcre_study()\fP contains option bits. There are three
1052 ph10 1022 further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1053 ph10 921 .sp
1054     PCRE_STUDY_JIT_COMPILE
1055     PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1056     PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1057 ph10 975 .sp
1058 ph10 921 If any of these are set, and the just-in-time compiler is available, the
1059     pattern is further compiled into machine code that executes much faster than
1060     the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
1061 ph10 1022 compiler is not available, these options are ignored. All undefined bits in the
1062 ph10 921 \fIoptions\fP argument must be zero.
1063 nigel 75 .P
1064 ph10 691 JIT compilation is a heavyweight optimization. It can take some time for
1065 ph10 678 patterns to be analyzed, and for one-off matches and simple patterns the
1066     benefit of faster execution might be offset by a much slower study time.
1067 ph10 691 Not all patterns can be optimized by the JIT compiler. For those that cannot be
1068 ph10 678 handled, matching automatically falls back to the \fBpcre_exec()\fP
1069     interpreter. For more details, see the
1070     .\" HREF
1071     \fBpcrejit\fP
1072     .\"
1073     documentation.
1074     .P
1075 nigel 75 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
1076 nigel 63 studying succeeds (even if no data is returned), the variable it points to is
1077 nigel 87 set to NULL. Otherwise it is set to point to a textual error message. This is a
1078     static string that is part of the library. You must not try to free it. You
1079     should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
1080     sure that it has run successfully.
1081 nigel 75 .P
1082 ph10 678 When you are finished with a pattern, you can free the memory used for the
1083     study data by calling \fBpcre_free_study()\fP. This function was added to the
1084     API for release 8.20. For earlier versions, the memory could be freed with
1085     \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1086 ph10 921 where JIT optimization is not used, but it is advisable to change to the new
1087     function when convenient.
1088 ph10 678 .P
1089 ph10 691 This is a typical way in which \fBpcre_study\fP() is used (except that in a
1090 ph10 678 real application there should be tests for errors):
1091 nigel 75 .sp
1092 ph10 678 int rc;
1093     pcre *re;
1094     pcre_extra *sd;
1095     re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1096     sd = pcre_study(
1097 nigel 63 re, /* result of pcre_compile() */
1098 ph10 678 0, /* no options */
1099 nigel 63 &error); /* set to NULL or points to a message */
1100 ph10 678 rc = pcre_exec( /* see below for details of pcre_exec() options */
1101 ph10 691 re, sd, "subject", 7, 0, 0, ovector, 30);
1102 ph10 678 ...
1103     pcre_free_study(sd);
1104 ph10 691 pcre_free(re);
1105 nigel 75 .sp
1106 ph10 455 Studying a pattern does two things: first, a lower bound for the length of
1107 ph10 461 subject string that is needed to match the pattern is computed. This does not
1108     mean that there are any strings of that length that match, but it does
1109 ph10 1022 guarantee that no shorter strings match. The value is used to avoid wasting
1110     time by trying to match strings that are shorter than the lower bound. You can
1111     find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
1112 ph10 455 .P
1113     Studying a pattern is also useful for non-anchored patterns that do not have a
1114     single fixed starting character. A bitmap of possible starting bytes is
1115 ph10 461 created. This speeds up finding a position in the subject at which to start
1116 chpe 1055 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1117     In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1118 ph10 547 .P
1119 ph10 691 These two optimizations apply to both \fBpcre_exec()\fP and
1120 ph10 921 \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1121 ph10 1314 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option.
1122 ph10 1301 You might want to do this if your pattern contains callouts or (*MARK) and you
1123     want to make use of these facilities in cases where matching fails.
1124     .P
1125     PCRE_NO_START_OPTIMIZE can be specified at either compile time or execution
1126 ph10 1314 time. However, if PCRE_NO_START_OPTIMIZE is passed to \fBpcre_exec()\fP, (that
1127     is, after any JIT compilation has happened) JIT execution is disabled. For JIT
1128 ph10 1301 execution to work with PCRE_NO_START_OPTIMIZE, the option must be set at
1129     compile time.
1130     .P
1131     There is a longer discussion of PCRE_NO_START_OPTIMIZE
1132 ph10 547 .\" HTML <a href="#execoptions">
1133     .\" </a>
1134     below.
1135     .\"
1136 nigel 75 .
1137     .
1138 nigel 63 .\" HTML <a name="localesupport"></a>
1139 nigel 75 .SH "LOCALE SUPPORT"
1140 nigel 63 .rs
1141     .sp
1142 ph10 139 PCRE handles caseless matching, and determines whether characters are letters,
1143 nigel 75 digits, or whatever, by reference to a set of tables, indexed by character
1144 ph10 1400 code point. When running in UTF-8 mode, or in the 16- or 32-bit libraries, this
1145     applies only to characters with code points less than 256. By default,
1146     higher-valued code points never match escapes such as \ew or \ed. However, if
1147     PCRE is built with Unicode property support, all characters can be tested with
1148     \ep and \eP, or, alternatively, the PCRE_UCP option can be set when a pattern
1149     is compiled; this causes \ew and friends to use Unicode property support
1150     instead of the built-in tables.
1151 nigel 75 .P
1152 ph10 1400 The use of locales with Unicode is discouraged. If you are handling characters
1153     with code points greater than 128, you should either use Unicode support, or
1154     use locales, but not try to mix the two.
1155     .P
1156 ph10 139 PCRE contains an internal set of tables that are used when the final argument
1157     of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1158 ph10 142 Normally, the internal tables recognize only ASCII characters. However, when
1159 ph10 139 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1160     default "C" locale of the local system, which may cause them to be different.
1161 nigel 75 .P
1162 ph10 139 The internal tables can always be overridden by tables supplied by the
1163     application that calls PCRE. These may be created in a different locale from
1164     the default. As more and more applications change to using Unicode, the need
1165     for this locale support is expected to die away.
1166     .P
1167 nigel 75 External tables are built by calling the \fBpcre_maketables()\fP function,
1168     which has no arguments, in the relevant locale. The result can then be passed
1169 ph10 1400 to \fBpcre_compile()\fP as often as necessary. For example, to build and use
1170     tables that are appropriate for the French locale (where accented characters
1171     with values greater than 128 are treated as letters), the following code could
1172     be used:
1173 nigel 75 .sp
1174     setlocale(LC_CTYPE, "fr_FR");
1175 nigel 63 tables = pcre_maketables();
1176     re = pcre_compile(..., tables);
1177 nigel 75 .sp
1178 ph10 142 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1179 ph10 139 are using Windows, the name for the French locale is "french".
1180     .P
1181 nigel 75 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1182     obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1183     that the memory containing the tables remains available for as long as it is
1184     needed.
1185     .P
1186     The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1187     pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1188 ph10 1400 and also by \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP. Thus, for any single
1189 nigel 75 pattern, compilation, studying and matching all happen in the same locale, but
1190 ph10 1400 different patterns can be processed in different locales.
1191 nigel 75 .P
1192     It is possible to pass a table pointer or NULL (indicating the use of the
1193 ph10 1400 internal tables) to \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP (see the
1194     discussion below in the section on matching a pattern). This facility is
1195     provided for use with pre-compiled patterns that have been saved and reloaded.
1196     Character tables are not saved with patterns, so if a non-standard table was
1197     used at compile time, it must be provided again when the reloaded pattern is
1198     matched. Attempting to use this facility to match a pattern in a different
1199     locale from the one in which it was compiled is likely to lead to anomalous
1200     (usually incorrect) results.
1201 nigel 75 .
1202     .
1203 ph10 598 .\" HTML <a name="infoaboutpattern"></a>
1204 nigel 75 .SH "INFORMATION ABOUT A PATTERN"
1205 nigel 63 .rs
1206     .sp
1207 ph10 1339 .nf
1208 nigel 75 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1209 ph10 1339 .B " int \fIwhat\fP, void *\fIwhere\fP);"
1210     .fi
1211 nigel 63 .PP
1212 nigel 75 The \fBpcre_fullinfo()\fP function returns information about a compiled
1213 ph10 856 pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1214     library at version 8.30, after more than 10 years of obsolescence.
1215 nigel 75 .P
1216     The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1217     pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1218 nigel 63 the pattern was not studied. The third argument specifies which piece of
1219     information is required, and the fourth argument is a pointer to a variable
1220     to receive the data. The yield of the function is zero for success, or one of
1221     the following negative numbers:
1222 nigel 75 .sp
1223 ph10 856 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1224     the argument \fIwhere\fP was NULL
1225     PCRE_ERROR_BADMAGIC the "magic number" was not found
1226 ph10 903 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1227 ph10 856 endianness
1228     PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1229 ph10 1314 PCRE_ERROR_UNSET the requested field is not set
1230 nigel 75 .sp
1231     The "magic number" is placed at the start of each compiled pattern as an simple
1232 ph10 903 check against passing an arbitrary memory pointer. The endianness error can
1233 ph10 856 occur if a compiled pattern is saved and reloaded on a different host. Here is
1234     a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1235     pattern:
1236 nigel 75 .sp
1237 nigel 63 int rc;
1238 nigel 91 size_t length;
1239 nigel 63 rc = pcre_fullinfo(
1240     re, /* result of pcre_compile() */
1241 ph10 678 sd, /* result of pcre_study(), or NULL */
1242 nigel 63 PCRE_INFO_SIZE, /* what is required */
1243     &length); /* where to put the data */
1244 nigel 75 .sp
1245     The possible values for the third argument are defined in \fBpcre.h\fP, and are
1246 nigel 63 as follows:
1247 nigel 75 .sp
1248 nigel 63 PCRE_INFO_BACKREFMAX
1249 nigel 75 .sp
1250 nigel 63 Return the number of the highest back reference in the pattern. The fourth
1251 nigel 75 argument should point to an \fBint\fP variable. Zero is returned if there are
1252 nigel 63 no back references.
1253 nigel 75 .sp
1254 nigel 63 PCRE_INFO_CAPTURECOUNT
1255 nigel 75 .sp
1256 nigel 63 Return the number of capturing subpatterns in the pattern. The fourth argument
1257 nigel 75 should point to an \fBint\fP variable.
1258     .sp
1259 nigel 77 PCRE_INFO_DEFAULT_TABLES
1260 nigel 75 .sp
1261     Return a pointer to the internal default character tables within PCRE. The
1262     fourth argument should point to an \fBunsigned char *\fP variable. This
1263     information call is provided for internal use by the \fBpcre_study()\fP
1264     function. External callers can cause PCRE to use its internal tables by passing
1265     a NULL table pointer.
1266     .sp
1267 ph10 1433 PCRE_INFO_FIRSTBYTE (deprecated)
1268 nigel 75 .sp
1269 ph10 856 Return information about the first data unit of any matched string, for a
1270 ph10 1433 non-anchored pattern. The name of this option refers to the 8-bit library,
1271     where data units are bytes. The fourth argument should point to an \fBint\fP
1272     variable. Negative values are used for special cases. However, this means that
1273     when the 32-bit library is in non-UTF-32 mode, the full 32-bit range of
1274     characters cannot be returned. For this reason, this value is deprecated; use
1275     PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER instead.
1276 nigel 75 .P
1277 ph10 856 If there is a fixed first value, for example, the letter "c" from a pattern
1278 ph10 903 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1279 chpe 1055 value is always less than 256. In the 16-bit library the value can be up to
1280     0xffff. In the 32-bit library the value can be up to 0x10ffff.
1281 ph10 856 .P
1282     If there is no fixed first value, and if either
1283 nigel 75 .sp
1284 nigel 63 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1285     starts with "^", or
1286 nigel 75 .sp
1287 nigel 63 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1288     (if it were set, the pattern would be anchored),
1289 nigel 75 .sp
1290 nigel 63 -1 is returned, indicating that the pattern matches only at the start of a
1291     subject string or after any newline within the string. Otherwise -2 is
1292     returned. For anchored patterns, -2 is returned.
1293 ph10 1433 .sp
1294     PCRE_INFO_FIRSTCHARACTER
1295     .sp
1296     Return the value of the first data unit (non-UTF character) of any matched
1297     string in the situation where PCRE_INFO_FIRSTCHARACTERFLAGS returns 1;
1298     otherwise return 0. The fourth argument should point to an \fBuint_t\fP
1299     variable.
1300 chpe 1080 .P
1301 ph10 1433 In the 8-bit library, the value is always less than 256. In the 16-bit library
1302     the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1303     can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1304 nigel 75 .sp
1305 ph10 1433 PCRE_INFO_FIRSTCHARACTERFLAGS
1306     .sp
1307     Return information about the first data unit of any matched string, for a
1308     non-anchored pattern. The fourth argument should point to an \fBint\fP
1309     variable.
1310     .P
1311     If there is a fixed first value, for example, the letter "c" from a pattern
1312     such as (cat|cow|coyote), 1 is returned, and the character value can be
1313     retrieved using PCRE_INFO_FIRSTCHARACTER. If there is no fixed first value, and
1314     if either
1315     .sp
1316     (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1317     starts with "^", or
1318     .sp
1319     (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1320     (if it were set, the pattern would be anchored),
1321     .sp
1322     2 is returned, indicating that the pattern matches only at the start of a
1323     subject string or after any newline within the string. Otherwise 0 is
1324     returned. For anchored patterns, 0 is returned.
1325     .sp
1326 nigel 63 PCRE_INFO_FIRSTTABLE
1327 nigel 75 .sp
1328 nigel 63 If the pattern was studied, and this resulted in the construction of a 256-bit
1329 ph10 856 table indicating a fixed set of values for the first data unit in any matching
1330 nigel 63 string, a pointer to the table is returned. Otherwise NULL is returned. The
1331 nigel 75 fourth argument should point to an \fBunsigned char *\fP variable.
1332     .sp
1333 ph10 226 PCRE_INFO_HASCRORLF
1334     .sp
1335 ph10 227 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1336 ph10 243 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1337 ph10 231 explicit match is either a literal CR or LF character, or \er or \en.
1338 ph10 226 .sp
1339 ph10 169 PCRE_INFO_JCHANGED
1340     .sp
1341 ph10 278 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1342     0. The fourth argument should point to an \fBint\fP variable. (?J) and
1343     (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1344 ph10 169 .sp
1345 ph10 678 PCRE_INFO_JIT
1346     .sp
1347 ph10 921 Return 1 if the pattern was studied with one of the JIT options, and
1348 ph10 678 just-in-time compiling was successful. The fourth argument should point to an
1349     \fBint\fP variable. A return value of 0 means that JIT support is not available
1350 ph10 921 in this version of PCRE, or that the pattern was not studied with a JIT option,
1351     or that the JIT compiler could not handle this particular pattern. See the
1352 ph10 678 .\" HREF
1353     \fBpcrejit\fP
1354     .\"
1355     documentation for details of what can and cannot be handled.
1356     .sp
1357 ph10 836 PCRE_INFO_JITSIZE
1358     .sp
1359 ph10 921 If the pattern was successfully studied with a JIT option, return the size of
1360     the JIT compiled code, otherwise return zero. The fourth argument should point
1361     to a \fBsize_t\fP variable.
1362 ph10 836 .sp
1363 nigel 63 PCRE_INFO_LASTLITERAL
1364 nigel 75 .sp
1365 ph10 856 Return the value of the rightmost literal data unit that must exist in any
1366     matched string, other than at its start, if such a value has been recorded. The
1367     fourth argument should point to an \fBint\fP variable. If there is no such
1368     value, -1 is returned. For anchored patterns, a last literal value is recorded
1369     only if it follows something of variable length. For example, for the pattern
1370 nigel 75 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1371 nigel 65 is -1.
1372 chpe 1080 .P
1373     Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1374 ph10 1348 to return the full 32-bit range of characters, this value is deprecated;
1375 chpe 1113 instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1376 chpe 1080 be used.
1377 nigel 75 .sp
1378 ph10 1348 PCRE_INFO_MATCH_EMPTY
1379     .sp
1380 ph10 1404 Return 1 if the pattern can match an empty string, otherwise 0. The fourth
1381     argument should point to an \fBint\fP variable.
1382 ph10 1348 .sp
1383 ph10 1314 PCRE_INFO_MATCHLIMIT
1384     .sp
1385     If the pattern set a match limit by including an item of the form
1386     (*LIMIT_MATCH=nnnn) at the start, the value is returned. The fourth argument
1387     should point to an unsigned 32-bit integer. If no such value has been set, the
1388     call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1389     .sp
1390 ph10 932 PCRE_INFO_MAXLOOKBEHIND
1391     .sp
1392 ph10 1328 Return the number of characters (NB not data units) in the longest lookbehind
1393 ph10 1253 assertion in the pattern. This information is useful when doing multi-segment
1394     matching using the partial matching facilities. Note that the simple assertions
1395     \eb and \eB require a one-character lookbehind. \eA also registers a
1396     one-character lookbehind, though it does not actually inspect the previous
1397     character. This is to ensure that at least one character from the old segment
1398 ph10 1314 is retained when a new segment is processed. Otherwise, if there are no
1399     lookbehinds in the pattern, \eA might match incorrectly at the start of a new
1400 ph10 1253 segment.
1401 ph10 932 .sp
1402 ph10 455 PCRE_INFO_MINLENGTH
1403     .sp
1404     If the pattern was studied and a minimum length for matching subject strings
1405     was computed, its value is returned. Otherwise the returned value is -1. The
1406 ph10 1328 value is a number of characters, which in UTF mode may be different from the
1407     number of data units. The fourth argument should point to an \fBint\fP
1408     variable. A non-negative value is a lower bound to the length of any matching
1409     string. There may not be any strings of that length that do actually match, but
1410     every string that does match is at least that long.
1411 ph10 455 .sp
1412 nigel 63 PCRE_INFO_NAMECOUNT
1413     PCRE_INFO_NAMEENTRYSIZE
1414     PCRE_INFO_NAMETABLE
1415 nigel 75 .sp
1416 nigel 63 PCRE supports the use of named as well as numbered capturing parentheses. The
1417     names are just an additional way of identifying the parentheses, which still
1418 nigel 91 acquire numbers. Several convenience functions such as
1419     \fBpcre_get_named_substring()\fP are provided for extracting captured
1420     substrings by name. It is also possible to extract the data directly, by first
1421     converting the name to a number in order to access the correct pointers in the
1422     output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1423     you need to use the name-to-number map, which is described by these three
1424     values.
1425 nigel 75 .P
1426 nigel 63 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1427     the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1428 nigel 75 entry; both of these return an \fBint\fP value. The entry size depends on the
1429 nigel 63 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1430 ph10 856 entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1431     the first two bytes of each entry are the number of the capturing parenthesis,
1432     most significant byte first. In the 16-bit library, the pointer points to
1433 ph10 1328 16-bit data units, the first of which contains the parenthesis number. In the
1434     32-bit library, the pointer points to 32-bit data units, the first of which
1435     contains the parenthesis number. The rest of the entry is the corresponding
1436     name, zero terminated.
1437 ph10 457 .P
1438 ph10 1359 The names are in alphabetical order. If (?| is used to create multiple groups
1439     with the same number, as described in the
1440 ph10 457 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1441     .\" </a>
1442     section on duplicate subpattern numbers
1443     .\"
1444     in the
1445     .\" HREF
1446     \fBpcrepattern\fP
1447     .\"
1448 ph10 1359 page, the groups may be given the same name, but there is only one entry in the
1449     table. Different names for groups of the same number are not permitted.
1450     Duplicate names for subpatterns with different numbers are permitted,
1451     but only if PCRE_DUPNAMES is set. They appear in the table in the order in
1452     which they were found in the pattern. In the absence of (?| this is the order
1453     of increasing number; when (?| is used this is not necessarily the case because
1454     later subpatterns may have lower numbers.
1455 ph10 457 .P
1456     As a simple example of the name/number table, consider the following pattern
1457 ph10 856 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1458     space - including newlines - is ignored):
1459 nigel 75 .sp
1460     .\" JOIN
1461 nigel 93 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1462     (?<month>\ed\ed) - (?<day>\ed\ed) )
1463 nigel 75 .sp
1464 nigel 63 There are four named subpatterns, so the table has four entries, and each entry
1465     in the table is eight bytes long. The table is as follows, with non-printing
1466 nigel 75 bytes shows in hexadecimal, and undefined bytes shown as ??:
1467     .sp
1468 nigel 63 00 01 d a t e 00 ??
1469     00 05 d a y 00 ?? ??
1470     00 04 m o n t h 00
1471     00 02 y e a r 00 ??
1472 nigel 75 .sp
1473     When writing code to extract data from named subpatterns using the
1474 nigel 91 name-to-number map, remember that the length of the entries is likely to be
1475 nigel 75 different for each compiled pattern.
1476     .sp
1477 ph10 169 PCRE_INFO_OKPARTIAL
1478     .sp
1479 ph10 435 Return 1 if the pattern can be used for partial matching with
1480     \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1481     \fBint\fP variable. From release 8.00, this always returns 1, because the
1482     restrictions that previously applied to partial matching have been lifted. The
1483 ph10 169 .\" HREF
1484     \fBpcrepartial\fP
1485     .\"
1486 ph10 426 documentation gives details of partial matching.
1487 ph10 169 .sp
1488 nigel 63 PCRE_INFO_OPTIONS
1489 nigel 75 .sp
1490 nigel 63 Return a copy of the options with which the pattern was compiled. The fourth
1491 nigel 75 argument should point to an \fBunsigned long int\fP variable. These option bits
1492     are those specified in the call to \fBpcre_compile()\fP, modified by any
1493 ph10 196 top-level option settings at the start of the pattern itself. In other words,
1494     they are the options that will be in force when matching starts. For example,
1495     if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1496     result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1497 nigel 75 .P
1498 nigel 63 A pattern is automatically anchored by PCRE if all of its top-level
1499     alternatives begin with one of the following:
1500 nigel 75 .sp
1501 nigel 63 ^ unless PCRE_MULTILINE is set
1502 nigel 75 \eA always
1503     \eG always
1504     .\" JOIN
1505 nigel 63 .* if PCRE_DOTALL is set and there are no back
1506     references to the subpattern in which .* appears
1507 nigel 75 .sp
1508 nigel 63 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1509 nigel 75 \fBpcre_fullinfo()\fP.
1510     .sp
1511 ph10 1314 PCRE_INFO_RECURSIONLIMIT
1512     .sp
1513     If the pattern set a recursion limit by including an item of the form
1514     (*LIMIT_RECURSION=nnnn) at the start, the value is returned. The fourth
1515     argument should point to an unsigned 32-bit integer. If no such value has been
1516     set, the call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1517     .sp
1518 nigel 63 PCRE_INFO_SIZE
1519 nigel 75 .sp
1520 ph10 1328 Return the size of the compiled pattern in bytes (for all three libraries). The
1521 ph10 856 fourth argument should point to a \fBsize_t\fP variable. This value does not
1522     include the size of the \fBpcre\fP structure that is returned by
1523     \fBpcre_compile()\fP. The value that is passed as the argument to
1524     \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1525     place the compiled data is the value returned by this option plus the size of
1526     the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1527     does not alter the value returned by this option.
1528 nigel 75 .sp
1529 nigel 63 PCRE_INFO_STUDYSIZE
1530 nigel 75 .sp
1531 ph10 1328 Return the size in bytes (for all three libraries) of the data block pointed to
1532     by the \fIstudy_data\fP field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP
1533     is NULL, or there is no study data, zero is returned. The fourth argument
1534     should point to a \fBsize_t\fP variable. The \fIstudy_data\fP field is set by
1535     \fBpcre_study()\fP to record information that will speed up matching (see the
1536     section entitled
1537 ph10 656 .\" HTML <a href="#studyingapattern">
1538     .\" </a>
1539 ph10 659 "Studying a pattern"
1540 ph10 656 .\"
1541     above). The format of the \fIstudy_data\fP block is private, but its length
1542     is made available via this option so that it can be saved and restored (see the
1543     .\" HREF
1544     \fBpcreprecompile\fP
1545     .\"
1546     documentation for details).
1547 chpe 1080 .sp
1548 chpe 1113 PCRE_INFO_REQUIREDCHARFLAGS
1549 chpe 1080 .sp
1550 ph10 1191 Returns 1 if there is a rightmost literal data unit that must exist in any
1551     matched string, other than at its start. The fourth argument should point to
1552     an \fBint\fP variable. If there is no such value, 0 is returned. If returning
1553     1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1554 chpe 1080 .P
1555 ph10 1191 For anchored patterns, a last literal value is recorded only if it follows
1556     something of variable length. For example, for the pattern /^a\ed+z\ed+/ the
1557     returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1558     /^a\edz\ed/ the returned value is 0.
1559 chpe 1080 .sp
1560 chpe 1113 PCRE_INFO_REQUIREDCHAR
1561 chpe 1080 .sp
1562     Return the value of the rightmost literal data unit that must exist in any
1563     matched string, other than at its start, if such a value has been recorded. The
1564     fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1565     value, 0 is returned.
1566 nigel 75 .
1567     .
1568 nigel 77 .SH "REFERENCE COUNTS"
1569 nigel 63 .rs
1570     .sp
1571 nigel 77 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1572     .PP
1573     The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1574     data block that contains a compiled pattern. It is provided for the benefit of
1575     applications that operate in an object-oriented manner, where different parts
1576     of the application may be using the same compiled pattern, but you want to free
1577     the block when they are all done.
1578     .P
1579     When a pattern is compiled, the reference count field is initialized to zero.
1580     It is changed only by calling this function, whose action is to add the
1581     \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1582     function is the new value. However, the value of the count is constrained to
1583     lie between 0 and 65535, inclusive. If the new value is outside these limits,
1584     it is forced to the appropriate limit value.
1585     .P
1586     Except when it is zero, the reference count is not correctly preserved if a
1587     pattern is compiled on one host and then transferred to a host whose byte-order
1588     is different. (This seems a highly unlikely scenario.)
1589     .
1590     .
1591     .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1592     .rs
1593     .sp
1594 ph10 1339 .nf
1595 nigel 75 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1596 ph10 1339 .B " const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1597     .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);"
1598     .fi
1599 nigel 75 .P
1600     The function \fBpcre_exec()\fP is called to match a subject string against a
1601     compiled pattern, which is passed in the \fIcode\fP argument. If the
1602 ph10 455 pattern was studied, the result of the study should be passed in the
1603 ph10 707 \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1604     and \fIextra\fP arguments as many times as you like, in order to match
1605     different subject strings with the same pattern.
1606     .P
1607     This function is the main matching facility of the library, and it operates in
1608     a Perl-like manner. For specialist use there is also an alternative matching
1609     function, which is described
1610 nigel 77 .\" HTML <a href="#dfamatch">
1611     .\" </a>
1612     below
1613     .\"
1614     in the section about the \fBpcre_dfa_exec()\fP function.
1615 nigel 75 .P
1616     In most applications, the pattern will have been compiled (and optionally
1617     studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1618     possible to save compiled patterns and study data, and then use them later
1619     in different processes, possibly even on different hosts. For a discussion
1620     about this, see the
1621     .\" HREF
1622     \fBpcreprecompile\fP
1623     .\"
1624     documentation.
1625     .P
1626     Here is an example of a simple call to \fBpcre_exec()\fP:
1627     .sp
1628 nigel 63 int rc;
1629     int ovector[30];
1630     rc = pcre_exec(
1631     re, /* result of pcre_compile() */
1632     NULL, /* we didn't study the pattern */
1633     "some string", /* the subject string */
1634     11, /* the length of the subject string */
1635     0, /* start at offset 0 in the subject */
1636     0, /* default options */
1637 nigel 75 ovector, /* vector of integers for substring information */
1638 nigel 77 30); /* number of elements (NOT size in bytes) */
1639 nigel 75 .
1640 ph10 656 .
1641 nigel 75 .\" HTML <a name="extradata"></a>
1642     .SS "Extra data for \fBpcre_exec()\fR"
1643     .rs
1644     .sp
1645     If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1646     data block. The \fBpcre_study()\fP function returns such a block (when it
1647 nigel 63 doesn't return NULL), but you can also create one for yourself, and pass
1648 nigel 87 additional information in it. The \fBpcre_extra\fP block contains the following
1649     fields (not necessarily in this order):
1650 nigel 75 .sp
1651     unsigned long int \fIflags\fP;
1652     void *\fIstudy_data\fP;
1653 ph10 691 void *\fIexecutable_jit\fP;
1654 nigel 75 unsigned long int \fImatch_limit\fP;
1655 nigel 87 unsigned long int \fImatch_limit_recursion\fP;
1656 nigel 75 void *\fIcallout_data\fP;
1657     const unsigned char *\fItables\fP;
1658 ph10 512 unsigned char **\fImark\fP;
1659 nigel 75 .sp
1660 ph10 903 In the 16-bit version of this structure, the \fImark\fP field has type
1661 ph10 859 "PCRE_UCHAR16 **".
1662 chpe 1055 .sp
1663     In the 32-bit version of this structure, the \fImark\fP field has type
1664     "PCRE_UCHAR32 **".
1665 ph10 856 .P
1666 ph10 926 The \fIflags\fP field is used to specify which of the other fields are set. The
1667     flag bits are:
1668 nigel 75 .sp
1669 ph10 922 PCRE_EXTRA_CALLOUT_DATA
1670 ph10 691 PCRE_EXTRA_EXECUTABLE_JIT
1671 ph10 922 PCRE_EXTRA_MARK
1672 nigel 63 PCRE_EXTRA_MATCH_LIMIT
1673 nigel 87 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1674 ph10 922 PCRE_EXTRA_STUDY_DATA
1675 nigel 75 PCRE_EXTRA_TABLES
1676     .sp
1677 ph10 678 Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1678     the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1679     returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1680 ph10 922 should not set these yourself, but you may add to the block by setting other
1681     fields and their corresponding flag bits.
1682 nigel 75 .P
1683     The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1684 nigel 63 vast amount of resources when running patterns that are not going to match,
1685     but which have a very large number of possibilities in their search trees. The
1686 ph10 456 classic example is a pattern that uses nested unlimited repeats.
1687 nigel 75 .P
1688 ph10 678 Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1689     calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1690     imposed on the number of times this function is called during a match, which
1691     has the effect of limiting the amount of backtracking that can take place. For
1692     patterns that are not anchored, the count restarts from zero for each position
1693 ph10 691 in the subject string.
1694 nigel 75 .P
1695 ph10 691 When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1696 ph10 921 with a JIT option, the way that the matching is executed is entirely different.
1697     However, there is still the possibility of runaway matching that goes on for a
1698     very long time, and so the \fImatch_limit\fP value is also used in this case
1699     (but in a different way) to limit how long the matching can continue.
1700 ph10 678 .P
1701 nigel 87 The default value for the limit can be set when PCRE is built; the default
1702 nigel 63 default is 10 million, which handles all but the most extreme cases. You can
1703 nigel 87 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1704     block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1705     the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1706     PCRE_ERROR_MATCHLIMIT.
1707 nigel 75 .P
1708 ph10 1314 A value for the match limit may also be supplied by an item at the start of a
1709     pattern of the form
1710     .sp
1711     (*LIMIT_MATCH=d)
1712     .sp
1713     where d is a decimal number. However, such a setting is ignored unless d is
1714     less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1715     is set, less than the default.
1716     .P
1717 nigel 87 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1718     instead of limiting the total number of times that \fBmatch()\fP is called, it
1719     limits the depth of recursion. The recursion depth is a smaller number than the
1720     total number of calls, because not all calls to \fBmatch()\fP are recursive.
1721 ph10 691 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1722 nigel 87 .P
1723 ph10 678 Limiting the recursion depth limits the amount of machine stack that can be
1724     used, or, when PCRE has been compiled to use memory on the heap instead of the
1725 ph10 686 stack, the amount of heap memory that can be used. This limit is not relevant,
1726 ph10 975 and is ignored, when matching is done using JIT compiled code.
1727 nigel 87 .P
1728     The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1729     built; the default default is the same value as the default for
1730     \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1731     with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1732     PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1733     is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1734     .P
1735 ph10 1314 A value for the recursion limit may also be supplied by an item at the start of
1736     a pattern of the form
1737     .sp
1738     (*LIMIT_RECURSION=d)
1739     .sp
1740     where d is a decimal number. However, such a setting is ignored unless d is
1741     less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1742     is set, less than the default.
1743     .P
1744 ph10 440 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1745     and is described in the
1746 nigel 75 .\" HREF
1747     \fBpcrecallout\fP
1748     .\"
1749     documentation.
1750     .P
1751 ph10 1400 The \fItables\fP field is provided for use with patterns that have been
1752     pre-compiled using custom character tables, saved to disc or elsewhere, and
1753     then reloaded, because the tables that were used to compile a pattern are not
1754     saved with it. See the
1755 nigel 75 .\" HREF
1756     \fBpcreprecompile\fP
1757     .\"
1758 ph10 1400 documentation for a discussion of saving compiled patterns for later use. If
1759     NULL is passed using this mechanism, it forces PCRE's internal tables to be
1760     used.
1761 ph10 510 .P
1762 ph10 1400 \fBWarning:\fP The tables that \fBpcre_exec()\fP uses must be the same as those
1763     that were used when the pattern was compiled. If this is not the case, the
1764     behaviour of \fBpcre_exec()\fP is undefined. Therefore, when a pattern is
1765     compiled and matched in the same process, this field should never be set. In
1766     this (the most common) case, the correct table pointer is automatically passed
1767     with the compiled pattern from \fBpcre_compile()\fP to \fBpcre_exec()\fP.
1768     .P
1769 ph10 510 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1770 ph10 856 be set to point to a suitable variable. If the pattern contains any
1771 ph10 510 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1772     a name to pass back, a pointer to the name string (zero terminated) is placed
1773     in the variable pointed to by the \fImark\fP field. The names are within the
1774     compiled pattern; if you wish to retain such a name you must copy it before
1775     freeing the memory of a compiled pattern. If there is no name to pass back, the
1776 ph10 856 variable pointed to by the \fImark\fP field is set to NULL. For details of the
1777 ph10 510 backtracking control verbs, see the section entitled
1778     .\" HTML <a href="pcrepattern#backtrackcontrol">
1779     .\" </a>
1780     "Backtracking control"
1781     .\"
1782     in the
1783     .\" HREF
1784     \fBpcrepattern\fP
1785     .\"
1786     documentation.
1787 nigel 75 .
1788 ph10 510 .
1789 ph10 226 .\" HTML <a name="execoptions"></a>
1790 nigel 75 .SS "Option bits for \fBpcre_exec()\fP"
1791     .rs
1792     .sp
1793     The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1794 nigel 91 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1795 ph10 442 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1796 ph10 921 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1797     PCRE_PARTIAL_SOFT.
1798 ph10 686 .P
1799 ph10 921 If the pattern was successfully studied with one of the just-in-time (JIT)
1800     compile options, the only supported options for JIT execution are
1801     PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1802     PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1803     unsupported option is used, JIT execution is disabled and the normal
1804     interpretive code in \fBpcre_exec()\fP is run.
1805 nigel 75 .sp
1806     PCRE_ANCHORED
1807     .sp
1808     The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1809     matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1810     to be anchored by virtue of its contents, it cannot be made unachored at
1811     matching time.
1812     .sp
1813 ph10 231 PCRE_BSR_ANYCRLF
1814     PCRE_BSR_UNICODE
1815     .sp
1816     These options (which are mutually exclusive) control what the \eR escape
1817     sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1818     match any Unicode newline sequence. These options override the choice that was
1819     made or defaulted when the pattern was compiled.
1820     .sp
1821 nigel 91 PCRE_NEWLINE_CR
1822     PCRE_NEWLINE_LF
1823     PCRE_NEWLINE_CRLF
1824 ph10 150 PCRE_NEWLINE_ANYCRLF
1825 nigel 93 PCRE_NEWLINE_ANY
1826 nigel 91 .sp
1827     These options override the newline definition that was chosen or defaulted when
1828 nigel 93 the pattern was compiled. For details, see the description of
1829     \fBpcre_compile()\fP above. During matching, the newline choice affects the
1830     behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1831     the way the match position is advanced after a match failure for an unanchored
1832 ph10 227 pattern.
1833 ph10 225 .P
1834 ph10 226 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1835     match attempt for an unanchored pattern fails when the current position is at a
1836 ph10 230 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1837 ph10 226 characters, the match position is advanced by two characters instead of one, in
1838     other words, to after the CRLF.
1839     .P
1840 ph10 227 The above rule is a compromise that makes the most common cases work as
1841     expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1842     set), it does not match the string "\er\enA" because, after failing at the
1843     start, it skips both the CR and the LF before retrying. However, the pattern
1844     [\er\en]A does match that string, because it contains an explicit CR or LF
1845 ph10 226 reference, and so advances only by one character after the first failure.
1846     .P
1847 ph10 231 An explicit match for CR of LF is either a literal appearance of one of those
1848     characters, or one of the \er or \en escape sequences. Implicit matches such as
1849 ph10 230 [^X] do not count, nor does \es (which includes CR and LF in the characters
1850     that it matches).
1851     .P
1852 ph10 226 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1853     valid newline sequence and explicit \er or \en escapes appear in the pattern.
1854 nigel 91 .sp
1855 nigel 63 PCRE_NOTBOL
1856 nigel 75 .sp
1857     This option specifies that first character of the subject string is not the
1858     beginning of a line, so the circumflex metacharacter should not match before
1859     it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1860     never to match. This option affects only the behaviour of the circumflex
1861     metacharacter. It does not affect \eA.
1862     .sp
1863 nigel 63 PCRE_NOTEOL
1864 nigel 75 .sp
1865     This option specifies that the end of the subject string is not the end of a
1866     line, so the dollar metacharacter should not match it nor (except in multiline
1867     mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1868     compile time) causes dollar never to match. This option affects only the
1869     behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1870     .sp
1871 nigel 63 PCRE_NOTEMPTY
1872 nigel 75 .sp
1873 nigel 63 An empty string is not considered to be a valid match if this option is set. If
1874     there are alternatives in the pattern, they are tried. If all the alternatives
1875     match the empty string, the entire match fails. For example, if the pattern
1876 nigel 75 .sp
1877 nigel 63 a?b?
1878 nigel 75 .sp
1879 ph10 442 is applied to a string not beginning with "a" or "b", it matches an empty
1880 nigel 63 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1881     valid, so PCRE searches further into the string for occurrences of "a" or "b".
1882 ph10 442 .sp
1883     PCRE_NOTEMPTY_ATSTART
1884     .sp
1885 ph10 461 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1886 ph10 442 the start of the subject is permitted. If the pattern is anchored, such a match
1887     can occur only if the pattern contains \eK.
1888 nigel 75 .P
1889 ph10 442 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1890     does make a special case of a pattern match of the empty string within its
1891     \fBsplit()\fP function, and when using the /g modifier. It is possible to
1892     emulate Perl's behaviour after matching a null string by first trying the match
1893     again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1894     if that fails, by advancing the starting offset (see below) and trying an
1895     ordinary match again. There is some code that demonstrates how to do this in
1896     the
1897 ph10 429 .\" HREF
1898     \fBpcredemo\fP
1899     .\"
1900 ph10 572 sample program. In the most general case, you have to check to see if the
1901     newline convention recognizes CRLF as a newline, and if so, and the current
1902 ph10 566 character is CR followed by LF, advance the starting offset by two characters
1903     instead of one.
1904 nigel 75 .sp
1905 ph10 389 PCRE_NO_START_OPTIMIZE
1906     .sp
1907 ph10 392 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1908 ph10 542 a match, in order to speed up the process. For example, if it is known that an
1909     unanchored match must start with a specific character, it searches the subject
1910     for that character, and fails immediately if it cannot find it, without
1911 ph10 545 actually running the main matching function. This means that a special item
1912 ph10 542 such as (*COMMIT) at the start of a pattern is not considered until after a
1913 ph10 1301 suitable starting point for the match has been found. Also, when callouts or
1914     (*MARK) items are in use, these "start-up" optimizations can cause them to be
1915     skipped if the pattern is never actually used. The start-up optimizations are
1916     in effect a pre-scan of the subject that takes place before the pattern is run.
1917 ph10 546 .P
1918     The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1919     causing performance to suffer, but ensuring that in cases where the result is
1920     "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1921 ph10 579 are considered at every possible starting position in the subject string. If
1922     PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1923 ph10 1301 time. The use of PCRE_NO_START_OPTIMIZE at matching time (that is, passing it
1924     to \fBpcre_exec()\fP) disables JIT execution; in this situation, matching is
1925     always done using interpretively.
1926 ph10 576 .P
1927 ph10 546 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1928     Consider the pattern
1929 ph10 389 .sp
1930 ph10 546 (*COMMIT)ABC
1931     .sp
1932     When this is compiled, PCRE records the fact that a match must start with the
1933 ph10 548 character "A". Suppose the subject string is "DEFABC". The start-up
1934     optimization scans along the subject, finds "A" and runs the first match
1935 ph10 546 attempt from there. The (*COMMIT) item means that the pattern must match the
1936 ph10 548 current starting position, which in this case, it does. However, if the same
1937     match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1938     subject string does not happen. The first match attempt is run starting from
1939 ph10 546 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1940     the overall result is "no match". If the pattern is studied, more start-up
1941     optimizations may be used. For example, a minimum length for the subject may be
1942     recorded. Consider the pattern
1943     .sp
1944     (*MARK:A)(X|Y)
1945     .sp
1946 ph10 548 The minimum length for a match is one character. If the subject is "ABC", there
1947     will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1948     If the pattern is studied, the final attempt does not take place, because PCRE
1949     knows that the subject is too short, and so the (*MARK) is never encountered.
1950     In this case, studying the pattern does not affect the overall match result,
1951     which is still "no match", but it does affect the auxiliary information that is
1952 ph10 546 returned.
1953     .sp
1954 nigel 75 PCRE_NO_UTF8_CHECK
1955     .sp
1956     When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1957     string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1958 ph10 959 The entire string is checked before any other processing takes place. The value
1959     of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1960     UTF-8 character. There is a discussion about the
1961     .\" HTML <a href="pcreunicode.html#utf8strings">
1962     .\" </a>
1963     validity of UTF-8 strings
1964     .\"
1965     in the
1966 ph10 211 .\" HREF
1967 ph10 856 \fBpcreunicode\fP
1968 ph10 211 .\"
1969 ph10 856 page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1970     error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1971     truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1972     cases, information about the precise nature of the error may also be returned
1973     (see the descriptions of these errors in the section entitled \fIError return
1974     values from\fP \fBpcre_exec()\fP
1975 ph10 598 .\" HTML <a href="#errorlist">
1976     .\" </a>
1977     below).
1978     .\"
1979     If \fIstartoffset\fP contains a value that does not point to the start of a
1980     UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1981 ph10 569 returned.
1982 nigel 75 .P
1983     If you already know that your subject is valid, and you want to skip these
1984     checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1985     calling \fBpcre_exec()\fP. You might want to do this for the second and
1986     subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1987     all the matches in a single subject string. However, you should be sure that
1988 ph10 856 the value of \fIstartoffset\fP points to the start of a character (or the end
1989     of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1990     invalid string as a subject or an invalid value of \fIstartoffset\fP is
1991 ph10 1363 undefined. Your program may crash or loop.
1992 nigel 75 .sp
1993 ph10 461 PCRE_PARTIAL_HARD
1994 ph10 428 PCRE_PARTIAL_SOFT
1995 nigel 75 .sp
1996 ph10 428 These options turn on the partial matching feature. For backwards
1997     compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1998     occurs if the end of the subject string is reached successfully, but there are
1999     not enough subject characters to complete the match. If this happens when
2000 ph10 553 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
2001     testing any remaining alternatives. Only if no complete match can be found is
2002     PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
2003     PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
2004     but only if no complete match can be found.
2005     .P
2006     If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
2007     partial match is found, \fBpcre_exec()\fP immediately returns
2008     PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
2009 ph10 572 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
2010 ph10 553 important that an alternative complete match.
2011     .P
2012     In both cases, the portion of the string that was inspected when the partial
2013     match was found is set as the first matching string. There is a more detailed
2014     discussion of partial and multi-segment matching, with examples, in the
2015 nigel 75 .\" HREF
2016     \fBpcrepartial\fP
2017     .\"
2018     documentation.
2019     .
2020 ph10 567 .
2021 nigel 75 .SS "The string to be matched by \fBpcre_exec()\fP"
2022     .rs
2023     .sp
2024     The subject string is passed to \fBpcre_exec()\fP as a pointer in
2025 ph10 1328 \fIsubject\fP, a length in \fIlength\fP, and a starting offset in
2026     \fIstartoffset\fP. The units for \fIlength\fP and \fIstartoffset\fP are bytes
2027     for the 8-bit library, 16-bit data items for the 16-bit library, and 32-bit
2028     data items for the 32-bit library.
2029 nigel 75 .P
2030 ph10 1328 If \fIstartoffset\fP is negative or greater than the length of the subject,
2031     \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting offset is
2032     zero, the search for a match starts at the beginning of the subject, and this
2033     is by far the most common case. In UTF-8 or UTF-16 mode, the offset must point
2034 ph10 1335 to the start of a character, or the end of the subject (in UTF-32 mode, one
2035 ph10 1328 data unit equals one character, so all offsets are valid). Unlike the pattern
2036     string, the subject may contain binary zeroes.
2037     .P
2038 nigel 63 A non-zero starting offset is useful when searching for another match in the
2039 nigel 75 same subject by calling \fBpcre_exec()\fP again after a previous success.
2040     Setting \fIstartoffset\fP differs from just passing over a shortened string and
2041 nigel 63 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
2042     lookbehind. For example, consider the pattern
2043 nigel 75 .sp
2044     \eBiss\eB
2045     .sp
2046     which finds occurrences of "iss" in the middle of words. (\eB matches only if
2047 nigel 63 the current position in the subject is not a word boundary.) When applied to
2048 nigel 75 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
2049     occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
2050     subject, namely "issipi", it does not match, because \eB is always false at the
2051 nigel 63 start of the subject, which is deemed to be a word boundary. However, if
2052 nigel 75 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
2053 nigel 63 set to 4, it finds the second occurrence of "iss" because it is able to look
2054     behind the starting point to discover that it is preceded by a letter.
2055 nigel 75 .P
2056 ph10 567 Finding all the matches in a subject is tricky when the pattern can match an
2057     empty string. It is possible to emulate Perl's /g behaviour by first trying the
2058     match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
2059     PCRE_ANCHORED options, and then if that fails, advancing the starting offset
2060     and trying an ordinary match again. There is some code that demonstrates how to
2061     do this in the
2062     .\" HREF
2063     \fBpcredemo\fP
2064     .\"
2065 ph10 572 sample program. In the most general case, you have to check to see if the
2066     newline convention recognizes CRLF as a newline, and if so, and the current
2067 ph10 567 character is CR followed by LF, advance the starting offset by two characters
2068     instead of one.
2069     .P
2070 nigel 63 If a non-zero starting offset is passed when the pattern is anchored, one
2071 nigel 75 attempt to match at the given offset is made. This can only succeed if the
2072 nigel 63 pattern does not require the match to be at the start of the subject.
2073 nigel 75 .
2074 ph10 567 .
2075 nigel 75 .SS "How \fBpcre_exec()\fP returns captured substrings"
2076     .rs
2077     .sp
2078 nigel 63 In general, a pattern matches a certain portion of the subject, and in
2079     addition, further substrings from the subject may be picked out by parts of the
2080     pattern. Following the usage in Jeffrey Friedl's book, this is called
2081     "capturing" in what follows, and the phrase "capturing subpattern" is used for
2082     a fragment of a pattern that picks out a substring. PCRE supports several other
2083     kinds of parenthesized subpattern that do not cause substrings to be captured.
2084 nigel 75 .P
2085 ph10 368 Captured substrings are returned to the caller via a vector of integers whose
2086     address is passed in \fIovector\fP. The number of elements in the vector is
2087     passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
2088     argument is NOT the size of \fIovector\fP in bytes.
2089 nigel 75 .P
2090     The first two-thirds of the vector is used to pass back captured substrings,
2091     each substring using a pair of integers. The remaining third of the vector is
2092     used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
2093 ph10 368 and is not available for passing back information. The number passed in
2094 nigel 75 \fIovecsize\fP should always be a multiple of three. If it is not, it is
2095     rounded down.
2096     .P
2097     When a match is successful, information about captured substrings is returned
2098     in pairs of integers, starting at the beginning of \fIovector\fP, and
2099 ph10 371 continuing up to two-thirds of its length at the most. The first element of
2100 ph10 1328 each pair is set to the offset of the first character in a substring, and the
2101     second is set to the offset of the first character after the end of a
2102     substring. These values are always data unit offsets, even in UTF mode. They
2103     are byte offsets in the 8-bit library, 16-bit data item offsets in the 16-bit
2104     library, and 32-bit data item offsets in the 32-bit library. \fBNote\fP: they
2105     are not character counts.
2106 nigel 75 .P
2107 ph10 368 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
2108     portion of the subject string matched by the entire pattern. The next pair is
2109     used for the first capturing subpattern, and so on. The value returned by
2110     \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
2111     For example, if two substrings have been captured, the returned value is 3. If
2112     there are no capturing subpatterns, the return value from a successful match is
2113     1, indicating that just the first pair of offsets has been set.
2114     .P
2115 nigel 63 If a capturing subpattern is matched repeatedly, it is the last portion of the
2116 nigel 75 string that it matched that is returned.
2117     .P
2118     If the vector is too small to hold all the captured substring offsets, it is
2119     used as far as possible (up to two-thirds of its length), and the function
2120 ph10 950 returns a value of zero. If neither the actual string matched nor any captured
2121 ph10 686 substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2122     passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2123     back references and the \fIovector\fP is not big enough to remember the related
2124     substrings, PCRE has to get additional memory for use during matching. Thus it
2125     is usually advisable to supply an \fIovector\fP of reasonable size.
2126 nigel 75 .P
2127 ph10 686 There are some cases where zero is returned (indicating vector overflow) when
2128     in fact the vector is exactly the right size for the final match. For example,
2129     consider the pattern
2130     .sp
2131     (a)(?:(b)c|bd)
2132     .sp
2133     If a vector of 6 elements (allowing for only 1 captured substring) is given
2134     with subject string "abd", \fBpcre_exec()\fP will try to set the second
2135     captured string, thereby recording a vector overflow, before failing to match
2136 ph10 691 "c" and backing up to try the second alternative. The zero return, however,
2137 ph10 686 does correctly indicate that the maximum number of slots (namely 2) have been
2138     filled. In similar cases where there is temporary overflow, but the final
2139     number of used slots is actually less than the maximum, a non-zero value is
2140     returned.
2141     .P
2142 ph10 456 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
2143 nigel 63 subpatterns there are in a compiled pattern. The smallest size for
2144 nigel 75 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
2145     the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
2146 nigel 91 .P
2147     It is possible for capturing subpattern number \fIn+1\fP to match some part of
2148     the subject when subpattern \fIn\fP has not been used at all. For example, if
2149     the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2150     function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2151     happens, both values in the offset pairs corresponding to unused subpatterns
2152     are set to -1.
2153     .P
2154     Offset values that correspond to unused subpatterns at the end of the
2155     expression are also set to -1. For example, if the string "abc" is matched
2156     against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2157     return from the function is 2, because the highest used capturing subpattern
2158 ph10 568 number is 1, and the offsets for for the second and third capturing subpatterns
2159     (assuming the vector is large enough, of course) are set to -1.
2160 nigel 91 .P
2161 ph10 686 \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
2162 ph10 683 correspond to capturing parentheses in the pattern are never changed. That is,
2163     if a pattern contains \fIn\fP capturing parentheses, no more than
2164     \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
2165     elements (in the first two-thirds) retain whatever values they previously had.
2166 ph10 568 .P
2167 nigel 91 Some convenience functions are provided for extracting the captured substrings
2168     as separate strings. These are described below.
2169 nigel 75 .
2170 ph10 598 .
2171 nigel 77 .\" HTML <a name="errorlist"></a>
2172 nigel 91 .SS "Error return values from \fBpcre_exec()\fP"
2173 nigel 75 .rs
2174     .sp
2175     If \fBpcre_exec()\fP fails, it returns a negative number. The following are
2176 nigel 63 defined in the header file:
2177 nigel 75 .sp
2178 nigel 63 PCRE_ERROR_NOMATCH (-1)
2179 nigel 75 .sp
2180 nigel 63 The subject string did not match the pattern.
2181 nigel 75 .sp
2182 nigel 63 PCRE_ERROR_NULL (-2)
2183 nigel 75 .sp
2184     Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
2185     NULL and \fIovecsize\fP was not zero.
2186     .sp
2187 nigel 63 PCRE_ERROR_BADOPTION (-3)
2188 nigel 75 .sp
2189     An unrecognized bit was set in the \fIoptions\fP argument.
2190     .sp
2191 nigel 63 PCRE_ERROR_BADMAGIC (-4)
2192 nigel 75 .sp
2193 nigel 63 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2194 nigel 75 the case when it is passed a junk pointer and to detect when a pattern that was
2195     compiled in an environment of one endianness is run in an environment with the
2196     other endianness. This is the error that PCRE gives when the magic number is
2197     not present.
2198     .sp
2199 nigel 93 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2200 nigel 75 .sp
2201 nigel 63 While running the pattern match, an unknown item was encountered in the
2202     compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2203     of the compiled pattern.
2204 nigel 75 .sp
2205 nigel 63 PCRE_ERROR_NOMEMORY (-6)
2206 nigel 75 .sp
2207     If a pattern contains back references, but the \fIovector\fP that is passed to
2208     \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2209 nigel 63 gets a block of memory at the start of matching to use for this purpose. If the
2210 nigel 75 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2211     automatically freed at the end of matching.
2212 ph10 531 .P
2213 ph10 535 This error is also given if \fBpcre_stack_malloc()\fP fails in
2214 ph10 531 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2215     \fB--disable-stack-for-recursion\fP.
2216 nigel 75 .sp
2217 nigel 63 PCRE_ERROR_NOSUBSTRING (-7)
2218 nigel 75 .sp
2219     This error is used by the \fBpcre_copy_substring()\fP,
2220     \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2221     below). It is never returned by \fBpcre_exec()\fP.
2222     .sp
2223 nigel 63 PCRE_ERROR_MATCHLIMIT (-8)
2224 nigel 75 .sp
2225 nigel 87 The backtracking limit, as specified by the \fImatch_limit\fP field in a
2226     \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2227     above.
2228     .sp
2229 nigel 63 PCRE_ERROR_CALLOUT (-9)
2230 nigel 75 .sp
2231     This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2232 nigel 63 use by callout functions that want to yield a distinctive error code. See the
2233 nigel 75 .\" HREF
2234     \fBpcrecallout\fP
2235     .\"
2236     documentation for details.
2237     .sp
2238 nigel 73 PCRE_ERROR_BADUTF8 (-10)
2239 nigel 75 .sp
2240 ph10 598 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2241     and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2242     (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2243     UTF-8 character is placed in the first element, and a reason code is placed in
2244     the second element. The reason codes are listed in the
2245     .\" HTML <a href="#badutf8reasons">
2246     .\" </a>
2247     following section.
2248     .\"
2249     For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2250     truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2251     PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2252 nigel 75 .sp
2253 nigel 73 PCRE_ERROR_BADUTF8_OFFSET (-11)
2254 nigel 75 .sp
2255 ph10 654 The UTF-8 byte sequence that was passed as a subject was checked and found to
2256 ph10 598 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2257     \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2258 ph10 569 end of the subject.
2259 nigel 75 .sp
2260 nigel 77 PCRE_ERROR_PARTIAL (-12)
2261 nigel 75 .sp
2262     The subject string did not match, but it did match partially. See the
2263     .\" HREF
2264     \fBpcrepartial\fP
2265     .\"
2266     documentation for details of partial matching.
2267     .sp
2268 nigel 77 PCRE_ERROR_BADPARTIAL (-13)
2269 nigel 75 .sp
2270 ph10 426 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2271     option was used with a compiled pattern containing items that were not
2272 ph10 461 supported for partial matching. From release 8.00 onwards, there are no
2273 ph10 426 restrictions on partial matching.
2274 nigel 75 .sp
2275 nigel 77 PCRE_ERROR_INTERNAL (-14)
2276 nigel 75 .sp
2277     An unexpected internal error has occurred. This error could be caused by a bug
2278     in PCRE or by overwriting of the compiled pattern.
2279     .sp
2280 nigel 77 PCRE_ERROR_BADCOUNT (-15)
2281 nigel 75 .sp
2282     This error is given if the value of the \fIovecsize\fP argument is negative.
2283 nigel 93 .sp
2284     PCRE_ERROR_RECURSIONLIMIT (-21)
2285     .sp
2286     The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2287     field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2288     description above.
2289     .sp
2290     PCRE_ERROR_BADNEWLINE (-23)
2291     .sp
2292     An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2293 ph10 567 .sp
2294     PCRE_ERROR_BADOFFSET (-24)
2295     .sp
2296 ph10 572 The value of \fIstartoffset\fP was negative or greater than the length of the
2297 ph10 567 subject, that is, the value in \fIlength\fP.
2298 ph10 569 .sp
2299     PCRE_ERROR_SHORTUTF8 (-25)
2300     .sp
2301 ph10 598 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2302     ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2303     Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2304     fact sufficient to detect this case, but this special error code for
2305     PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2306     retained for backwards compatibility.
2307 ph10 642 .sp
2308     PCRE_ERROR_RECURSELOOP (-26)
2309     .sp
2310 ph10 654 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2311     the pattern. Specifically, it means that either the whole pattern or a
2312     subpattern has been called recursively for the second time at the same position
2313 ph10 642 in the subject string. Some simple patterns that might do this are detected and
2314     faulted at compile time, but more complicated cases, in particular mutual
2315     recursions between two different subpatterns, cannot be detected until run
2316     time.
2317 ph10 678 .sp
2318     PCRE_ERROR_JIT_STACKLIMIT (-27)
2319     .sp
2320 ph10 922 This error is returned when a pattern that was successfully studied using a
2321     JIT compile option is being matched, but the memory available for the
2322     just-in-time processing stack is not large enough. See the
2323 ph10 678 .\" HREF
2324     \fBpcrejit\fP
2325     .\"
2326 ph10 691 documentation for more details.
2327 ph10 856 .sp
2328 ph10 960 PCRE_ERROR_BADMODE (-28)
2329 ph10 856 .sp
2330 ph10 903 This error is given if a pattern that was compiled by the 8-bit library is
2331 chpe 1055 passed to a 16-bit or 32-bit library function, or vice versa.
2332 ph10 856 .sp
2333 ph10 960 PCRE_ERROR_BADENDIANNESS (-29)
2334 ph10 903 .sp
2335     This error is given if a pattern that was compiled and saved is reloaded on a
2336     host with different endianness. The utility function
2337     \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2338 ph10 856 so that it runs on the new host.
2339 ph10 1191 .sp
2340 ph10 1194 PCRE_ERROR_JIT_BADOPTION
2341     .sp
2342     This error is returned when a pattern that was successfully studied using a JIT
2343     compile option is being matched, but the matching mode (partial or complete
2344 ph10 1221 match) does not correspond to any JIT compilation mode. When the JIT fast path
2345 ph10 1194 function is used, this error may be also given for invalid options. See the
2346     .\" HREF
2347     \fBpcrejit\fP
2348     .\"
2349     documentation for more details.
2350     .sp
2351 ph10 1191 PCRE_ERROR_BADLENGTH (-32)
2352     .sp
2353 ph10 1221 This error is given if \fBpcre_exec()\fP is called with a negative value for
2354     the \fIlength\fP argument.
2355 nigel 93 .P
2356 ph10 1194 Error numbers -16 to -20, -22, and 30 are not used by \fBpcre_exec()\fP.
2357 nigel 75 .
2358     .
2359 ph10 598 .\" HTML <a name="badutf8reasons"></a>
2360     .SS "Reason codes for invalid UTF-8 strings"
2361     .rs
2362     .sp
2363 ph10 903 This section applies only to the 8-bit library. The corresponding information
2364 ph10 1214 for the 16-bit and 32-bit libraries is given in the
2365 ph10 856 .\" HREF
2366     \fBpcre16\fP
2367     .\"
2368 ph10 1214 and
2369 chpe 1055 .\" HREF
2370     \fBpcre32\fP
2371     .\"
2372 ph10 1221 pages.
2373 ph10 856 .P
2374 ph10 654 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2375     PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2376     least 2, the offset of the start of the invalid UTF-8 character is placed in
2377     the first output vector element (\fIovector[0]\fP) and a reason code is placed
2378 ph10 598 in the second element (\fIovector[1]\fP). The reason codes are given names in
2379     the \fBpcre.h\fP header file:
2380     .sp
2381     PCRE_UTF8_ERR1
2382     PCRE_UTF8_ERR2
2383     PCRE_UTF8_ERR3
2384     PCRE_UTF8_ERR4
2385     PCRE_UTF8_ERR5
2386     .sp
2387 ph10 654 The string ends with a truncated UTF-8 character; the code specifies how many
2388 ph10 598 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2389     no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2390 ph10 654 allows for up to 6 bytes, and this is checked first; hence the possibility of
2391 ph10 598 4 or 5 missing bytes.
2392     .sp
2393     PCRE_UTF8_ERR6
2394     PCRE_UTF8_ERR7
2395     PCRE_UTF8_ERR8
2396     PCRE_UTF8_ERR9
2397     PCRE_UTF8_ERR10
2398     .sp
2399 ph10 654 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2400 ph10 598 character do not have the binary value 0b10 (that is, either the most
2401     significant bit is 0, or the next bit is 1).
2402 ph10 654 .sp
2403 ph10 598 PCRE_UTF8_ERR11
2404     PCRE_UTF8_ERR12
2405     .sp
2406 ph10 654 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2407     these code points are excluded by RFC 3629.
2408     .sp
2409 ph10 598 PCRE_UTF8_ERR13
2410     .sp
2411 ph10 654 A 4-byte character has a value greater than 0x10fff; these code points are
2412 ph10 598 excluded by RFC 3629.
2413 ph10 654 .sp
2414 ph10 598 PCRE_UTF8_ERR14
2415     .sp
2416     A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2417 ph10 654 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2418 ph10 598 from UTF-8.
2419 ph10 654 .sp
2420 ph10 598 PCRE_UTF8_ERR15
2421     PCRE_UTF8_ERR16
2422     PCRE_UTF8_ERR17
2423     PCRE_UTF8_ERR18
2424     PCRE_UTF8_ERR19
2425     .sp
2426 ph10 654 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2427     value that can be represented by fewer bytes, which is invalid. For example,
2428 ph10 598 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2429     one byte.
2430     .sp
2431     PCRE_UTF8_ERR20
2432     .sp
2433 ph10 654 The two most significant bits of the first byte of a character have the binary
2434     value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2435 ph10 598 byte can only validly occur as the second or subsequent byte of a multi-byte
2436     character.
2437     .sp
2438     PCRE_UTF8_ERR21
2439     .sp
2440     The first byte of a character has the value 0xfe or 0xff. These values can
2441     never occur in a valid UTF-8 string.
2442 chpe 1098 .sp
2443 chpe 1262 PCRE_UTF8_ERR22
2444 chpe 1098 .sp
2445 ph10 1261 This error code was formerly used when the presence of a so-called
2446 ph10 1314 "non-character" caused an error. Unicode corrigendum #9 makes it clear that
2447     such characters should not cause a string to be rejected, and so this code is
2448 ph10 1261 no longer in use and is never returned.
2449 ph10 598 .
2450     .
2451 nigel 75 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2452 nigel 63 .rs
2453     .sp
2454 ph10 1339 .nf
2455 nigel 75 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2456 ph10 1339 .B " int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,"
2457     .B " int \fIbuffersize\fP);"
2458     .sp
2459 nigel 75 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2460 ph10 1339 .B " int \fIstringcount\fP, int \fIstringnumber\fP,"
2461     .B " const char **\fIstringptr\fP);"
2462     .sp
2463 nigel 75 .B int pcre_get_substring_list(const char *\fIsubject\fP,
2464 ph10 1339 .B " int *\fIovector\fP, int \fIstringcount\fP, const char ***\fIlistptr\fP);"
2465     .fi
2466 nigel 63 .PP
2467     Captured substrings can be accessed directly by using the offsets returned by
2468 nigel 75 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2469     \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2470     \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2471 nigel 63 as new, separate, zero-terminated strings. These functions identify substrings
2472     by number. The next section describes functions for extracting named
2473 nigel 91 substrings.
2474 nigel 75 .P
2475 nigel 91 A substring that contains a binary zero is correctly extracted and has a
2476     further zero added on the end, but the result is not, of course, a C string.
2477     However, you can process such a string by referring to the length that is
2478     returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2479     Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2480     for handling strings containing binary zeros, because the end of the final
2481     string is not independently indicated.
2482     .P
2483 nigel 63 The first three arguments are the same for all three of these functions:
2484 nigel 75 \fIsubject\fP is the subject string that has just been successfully matched,
2485     \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2486     \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2487 nigel 63 captured by the match, including the substring that matched the entire regular
2488 nigel 75 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2489     than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2490     space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2491     number of elements in the vector divided by three.
2492     .P
2493     The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2494     extract a single substring, whose number is given as \fIstringnumber\fP. A
2495     value of zero extracts the substring that matched the entire pattern, whereas
2496     higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2497     the string is placed in \fIbuffer\fP, whose length is given by
2498     \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2499     obtained via \fBpcre_malloc\fP, and its address is returned via
2500     \fIstringptr\fP. The yield of the function is the length of the string, not
2501 nigel 93 including the terminating zero, or one of these error codes:
2502 nigel 75 .sp
2503 nigel 63 PCRE_ERROR_NOMEMORY (-6)
2504 nigel 75 .sp
2505     The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2506     memory failed for \fBpcre_get_substring()\fP.
2507     .sp
2508 nigel 63 PCRE_ERROR_NOSUBSTRING (-7)
2509 nigel 75 .sp
2510     There is no substring whose number is \fIstringnumber\fP.
2511     .P
2512     The \fBpcre_get_substring_list()\fP function extracts all available substrings
2513 nigel 63 and builds a list of pointers to them. All this is done in a single block of
2514 nigel 75 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2515     is returned via \fIlistptr\fP, which is also the start of the list of string
2516 nigel 63 pointers. The end of the list is marked by a NULL pointer. The yield of the
2517 nigel 93 function is zero if all went well, or the error code
2518 nigel 75 .sp
2519 nigel 63 PCRE_ERROR_NOMEMORY (-6)
2520 nigel 75 .sp
2521 nigel 63 if the attempt to get the memory block failed.
2522 nigel 75 .P
2523 nigel 63 When any of these functions encounter a substring that is unset, which can
2524 nigel 75 happen when capturing subpattern number \fIn+1\fP matches some part of the
2525     subject, but subpattern \fIn\fP has not been used at all, they return an empty
2526 nigel 63 string. This can be distinguished from a genuine zero-length substring by
2527 nigel 75 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2528 nigel 63 substrings.
2529 nigel 75 .P
2530     The two convenience functions \fBpcre_free_substring()\fP and
2531     \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2532     a previous call of \fBpcre_get_substring()\fP or
2533     \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2534     the function pointed to by \fBpcre_free\fP, which of course could be called
2535 nigel 63 directly from a C program. However, PCRE is used in some situations where it is
2536 nigel 91 linked via a special interface to another programming language that cannot use
2537 nigel 75 \fBpcre_free\fP directly; it is for these cases that the functions are
2538 nigel 63 provided.
2539 nigel 75 .
2540     .
2541     .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2542 nigel 63 .rs
2543     .sp
2544 ph10 1339 .nf
2545 nigel 75 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2546 ph10 1339 .B " const char *\fIname\fP);"
2547     .sp
2548 nigel 75 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2549 ph10 1339 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
2550     .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
2551     .B " char *\fIbuffer\fP, int \fIbuffersize\fP);"
2552     .sp
2553 nigel 75 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2554 ph10 1339 .B " const char *\fIsubject\fP, int *\fIovector\fP,"
2555     .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
2556     .B " const char **\fIstringptr\fP);"
2557     .fi
2558 nigel 63 .PP
2559 nigel 75 To extract a substring by name, you first have to find associated number.
2560     For example, for this pattern
2561     .sp
2562 nigel 93 (a+)b(?<xxx>\ed+)...
2563 nigel 75 .sp
2564 nigel 91 the number of the subpattern called "xxx" is 2. If the name is known to be
2565     unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2566     calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2567     pattern, and the second is the name. The yield of the function is the
2568 nigel 75 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2569     that name.
2570     .P
2571     Given the number, you can extract the substring directly, or use one of the
2572     functions described in the previous section. For convenience, there are also
2573     two functions that do the whole job.
2574     .P
2575 nigel 91 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2576     \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2577 nigel 75 functions that extract by number. As these are described in the previous
2578     section, they are not re-described here. There are just two differences:
2579     .P
2580 nigel 63 First, instead of a substring number, a substring name is given. Second, there
2581     is an extra argument, given at the start, which is a pointer to the compiled
2582     pattern. This is needed in order to gain access to the name-to-number
2583     translation table.
2584 nigel 75 .P
2585     These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2586 ph10 127 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2587     appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2588 ph10 128 the behaviour may not be what you want (see the next section).
2589 ph10 385 .P
2590 ph10 457 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2591     subpatterns with the same number, as described in the
2592     .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2593     .\" </a>
2594     section on duplicate subpattern numbers
2595     .\"
2596     in the
2597     .\" HREF
2598     \fBpcrepattern\fP
2599     .\"
2600     page, you cannot use names to distinguish the different subpatterns, because
2601     names are not included in the compiled code. The matching process uses only
2602     numbers. For this reason, the use of different names for subpatterns of the
2603     same number causes an error at compile time.
2604 nigel 77 .
2605 ph10 686 .
2606 nigel 91 .SH "DUPLICATE SUBPATTERN NAMES"
2607     .rs
2608     .sp
2609 ph10 1339 .nf
2610 nigel 91 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2611 ph10 1339 .B " const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);"
2612     .fi
2613 nigel 91 .PP
2614     When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2615 ph10 457 are not required to be unique. (Duplicate names are always allowed for
2616     subpatterns with the same number, created by using the (?| feature. Indeed, if
2617     such subpatterns are named, they are required to use the same names.)
2618     .P
2619     Normally, patterns with duplicate names are such that in any one match, only
2620     one of the named subpatterns participates. An example is shown in the
2621 nigel 91 .\" HREF
2622     \fBpcrepattern\fP
2623     .\"
2624 ph10 208 documentation.
2625 ph10 203 .P
2626     When duplicates are present, \fBpcre_copy_named_substring()\fP and
2627     \fBpcre_get_named_substring()\fP return the first substring corresponding to
2628     the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2629     returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2630     returns one of the numbers that are associated with the name, but it is not
2631     defined which it is.
2632     .P
2633 nigel 91 If you want to get full details of all captured substrings for a given name,
2634     you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2635     argument is the compiled pattern, and the second is the name. The third and
2636     fourth are pointers to variables which are updated by the function. After it
2637     has run, they point to the first and last entries in the name-to-number table
2638     for the given name. The function itself returns the length of each entry, or
2639 nigel 93 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2640 ph10 598 described above in the section entitled \fIInformation about a pattern\fP
2641     .\" HTML <a href="#infoaboutpattern">
2642     .\" </a>
2643     above.
2644     .\"
2645 nigel 93 Given all the relevant entries for the name, you can extract each of their
2646     numbers, and hence the captured data, if any.
2647 nigel 91 .
2648     .
2649 nigel 77 .SH "FINDING ALL POSSIBLE MATCHES"
2650     .rs
2651     .sp
2652     The traditional matching function uses a similar algorithm to Perl, which stops
2653     when it finds the first match, starting at a given point in the subject. If you
2654     want to find all possible matches, or the longest possible match, consider
2655     using the alternative matching function (see below) instead. If you cannot use
2656     the alternative function, but still need to find all possible matches, you
2657     can kludge it up by making use of the callout facility, which is described in
2658     the
2659     .\" HREF
2660     \fBpcrecallout\fP
2661     .\"
2662     documentation.
2663 nigel 75 .P
2664 nigel 77 What you have to do is to insert a callout right at the end of the pattern.
2665     When your callout function is called, extract and save the current matched
2666     substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2667     other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2668     will yield PCRE_ERROR_NOMATCH.
2669     .
2670     .
2671 ph10 901 .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2672     .rs
2673     .sp
2674     Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2675 ph10 903 stack, which in certain environments can be rather limited in size. Some users
2676     find it helpful to have an estimate of the amount of stack that is used by
2677 ph10 901 \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2678     .\" HREF
2679     \fBpcrestack\fP
2680     .\"
2681 ph10 903 documentation. The estimate that is output by \fBpcretest\fP when called with
2682     the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2683 ph10 901 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2684     .P
2685     Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2686     the negative error code PCRE_ERROR_NULL, but with this special combination of
2687     arguments, it returns instead a negative number whose absolute value is the
2688     approximate stack frame size in bytes. (A negative number is used so that it is
2689     clear that no match has happened.) The value is approximate because in some
2690 ph10 903 cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2691 ph10 901 additional variables on the stack.
2692     .P
2693 ph10 903 If PCRE has been compiled to use the heap instead of the stack for recursion,
2694 ph10 901 the value returned is the size of each block that is obtained from the heap.
2695     .
2696     .
2697 nigel 77 .\" HTML <a name="dfamatch"></a>
2698     .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2699     .rs
2700     .sp
2701 ph10 1339 .nf
2702 nigel 77 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2703 ph10 1339 .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
2704     .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
2705     .B " int *\fIworkspace\fP, int \fIwscount\fP);"
2706     .fi
2707 nigel 77 .P
2708     The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2709 nigel 93 a compiled pattern, using a matching algorithm that scans the subject string
2710     just once, and does not backtrack. This has different characteristics to the
2711     normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2712     patterns are not supported. Nevertheless, there are times when this kind of
2713 ph10 461 matching can be useful. For a discussion of the two matching algorithms, and a
2714 ph10 435 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2715 nigel 77 .\" HREF
2716     \fBpcrematching\fP
2717     .\"
2718     documentation.
2719     .P
2720     The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2721     \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2722     different way, and this is described below. The other common arguments are used
2723     in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2724     here.
2725     .P
2726     The two additional arguments provide workspace for the function. The workspace
2727     vector should contain at least 20 elements. It is used for keeping track of
2728     multiple paths through the pattern tree. More workspace will be needed for
2729 nigel 91 patterns and subjects where there are a lot of potential matches.
2730 nigel 77 .P
2731 nigel 87 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2732 nigel 77 .sp
2733     int rc;
2734     int ovector[10];
2735     int wspace[20];
2736 nigel 87 rc = pcre_dfa_exec(
2737 nigel 77 re, /* result of pcre_compile() */
2738     NULL, /* we didn't study the pattern */
2739     "some string", /* the subject string */
2740     11, /* the length of the subject string */
2741     0, /* start at offset 0 in the subject */
2742     0, /* default options */
2743     ovector, /* vector of integers for substring information */
2744     10, /* number of elements (NOT size in bytes) */
2745     wspace, /* working space vector */
2746     20); /* number of elements (NOT size in bytes) */
2747     .
2748     .SS "Option bits for \fBpcre_dfa_exec()\fP"
2749     .rs
2750     .sp
2751     The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2752 nigel 91 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2753 ph10 442 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2754 ph10 542 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2755     PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2756     All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2757     so their description is not repeated here.
2758 nigel 77 .sp
2759 ph10 428 PCRE_PARTIAL_HARD
2760 ph10 461 PCRE_PARTIAL_SOFT
2761 nigel 77 .sp
2762 ph10 428 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2763     details are slightly different. When PCRE_PARTIAL_HARD is set for
2764     \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2765     is reached and there is still at least one matching possibility that requires
2766     additional characters. This happens even if some complete matches have also
2767     been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2768     is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2769     there have been no complete matches, but there is still at least one matching
2770 ph10 435 possibility. The portion of the string that was inspected when the longest
2771     partial match was found is set as the first matching string in both cases.
2772 ph10 553 There is a more detailed discussion of partial and multi-segment matching, with
2773     examples, in the
2774     .\" HREF
2775     \fBpcrepartial\fP
2776     .\"
2777     documentation.
2778 nigel 77 .sp
2779     PCRE_DFA_SHORTEST
2780     .sp
2781     Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2782 nigel 93 soon as it has found one match. Because of the way the alternative algorithm
2783     works, this is necessarily the shortest possible match at the first possible
2784     matching point in the subject string.
2785 nigel 77 .sp
2786     PCRE_DFA_RESTART
2787     .sp
2788 ph10 428 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2789     again, with additional subject characters, and have it continue with the same
2790     match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2791     \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2792     before because data about the match so far is left in them after a partial
2793     match. There is more discussion of this facility in the
2794 nigel 77 .\" HREF
2795     \fBpcrepartial\fP
2796     .\"
2797     documentation.
2798     .
2799 ph10 598 .
2800 nigel 77 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2801     .rs
2802     .sp
2803     When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2804     substring in the subject. Note, however, that all the matches from one run of
2805     the function start at the same point in the subject. The shorter matches are
2806     all initial substrings of the longer matches. For example, if the pattern
2807     .sp
2808     <.*>
2809     .sp
2810     is matched against the string
2811     .sp
2812     This is <something> <something else> <something further> no more
2813     .sp
2814     the three matched strings are
2815     .sp
2816     <something>
2817     <something> <something else>
2818     <something> <something else> <something further>
2819     .sp
2820     On success, the yield of the function is a number greater than zero, which is
2821     the number of matched substrings. The substrings themselves are returned in
2822     \fIovector\fP. Each string uses two elements; the first is the offset to the
2823 nigel 93 start, and the second is the offset to the end. In fact, all the strings have
2824     the same start offset. (Space could have been saved by giving this only once,
2825     but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2826     returns data, even though the meaning of the strings is different.)
2827 nigel 77 .P
2828     The strings are returned in reverse order of length; that is, the longest
2829     matching string is given first. If there were too many matches to fit into
2830     \fIovector\fP, the yield of the function is zero, and the vector is filled with
2831 ph10 691 the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2832 ph10 683 the entire \fIovector\fP for returning matched strings.
2833 ph10 1404 .P
2834 ph10 1363 NOTE: PCRE's "auto-possessification" optimization usually applies to character
2835     repeats at the end of a pattern (as well as internally). For example, the
2836     pattern "a\ed+" is compiled as if it were "a\ed++" because there is no point
2837     even considering the possibility of backtracking into the repeated digits. For
2838     DFA matching, this means that only one possible match is found. If you really
2839     do want multiple matches in such cases, either use an ungreedy repeat
2840 ph10 1398 ("a\ed+?") or set the PCRE_NO_AUTO_POSSESS option when compiling.
2841 nigel 77 .
2842 ph10 598 .
2843 nigel 77 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2844     .rs
2845     .sp
2846     The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2847     Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2848     described
2849     .\" HTML <a href="#errorlist">
2850     .\" </a>
2851     above.
2852     .\"
2853     There are in addition the following errors that are specific to
2854     \fBpcre_dfa_exec()\fP:
2855     .sp
2856     PCRE_ERROR_DFA_UITEM (-16)
2857     .sp
2858     This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2859     that it does not support, for instance, the use of \eC or a back reference.
2860     .sp
2861     PCRE_ERROR_DFA_UCOND (-17)
2862     .sp
2863 nigel 93 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2864     uses a back reference for the condition, or a test for recursion in a specific
2865     group. These are not supported.
2866 nigel 77 .sp
2867     PCRE_ERROR_DFA_UMLIMIT (-18)
2868     .sp
2869     This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2870 ph10 678 block that contains a setting of the \fImatch_limit\fP or
2871     \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2872     meaningless for DFA matching).
2873 nigel 77 .sp
2874     PCRE_ERROR_DFA_WSSIZE (-19)
2875     .sp
2876     This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2877     \fIworkspace\fP vector.
2878     .sp
2879     PCRE_ERROR_DFA_RECURSE (-20)
2880     .sp
2881     When a recursive subpattern is processed, the matching function calls itself
2882     recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2883     error is given if the output vector is not large enough. This should be
2884     extremely rare, as a vector of size 1000 is used.
2885 ph10 960 .sp
2886     PCRE_ERROR_DFA_BADRESTART (-30)
2887     .sp
2888     When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2889 ph10 975 some plausibility checks are made on the contents of the workspace, which
2890     should contain data about the previous partial match. If any of these checks
2891     fail, this error is given.
2892 nigel 93 .
2893     .
2894     .SH "SEE ALSO"
2895     .rs
2896     .sp
2897 chpe 1055 \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2898     \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2899     \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2900     \fBpcrestack\fP(3).
2901 ph10 99 .
2902     .
2903     .SH AUTHOR
2904     .rs
2905     .sp
2906     .nf
2907     Philip Hazel
2908     University Computing Service
2909     Cambridge CB2 3QH, England.
2910     .fi
2911     .
2912     .
2913     .SH REVISION
2914     .rs
2915     .sp
2916     .nf
2917 ph10 1454 Last updated: 09 February 2014
2918 ph10 1433 Copyright (c) 1997-2014 University of Cambridge.
2919 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