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

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

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 65 by nigel, Sat Feb 24 21:40:08 2007 UTC revision 932 by ph10, Fri Feb 24 18:54:43 2012 UTC
# Line 1  Line 1 
1  .TH PCRE 3  .TH PCREAPI 3
2  .SH NAME  .SH NAME
3  PCRE - Perl-compatible regular expressions  PCRE - Perl-compatible regular expressions
 .SH SYNOPSIS OF PCRE API  
 .rs  
4  .sp  .sp
5  .B #include <pcre.h>  .B #include <pcre.h>
6  .PP  .
7    .
8    .SH "PCRE NATIVE API BASIC FUNCTIONS"
9    .rs
10    .sp
11  .SM  .SM
12  .br  .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
13  .B pcre *pcre_compile(const char *\fIpattern\fR, int \fIoptions\fR,  .ti +5n
14    .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
15    .ti +5n
16    .B const unsigned char *\fItableptr\fP);
17    .PP
18    .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
19    .ti +5n
20    .B int *\fIerrorcodeptr\fP,
21  .ti +5n  .ti +5n
22  .B const char **\fIerrptr\fR, int *\fIerroffset\fR,  .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
23  .ti +5n  .ti +5n
24  .B const unsigned char *\fItableptr\fR);  .B const unsigned char *\fItableptr\fP);
25  .PP  .PP
26  .br  .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
 .B pcre_extra *pcre_study(const pcre *\fIcode\fR, int \fIoptions\fR,  
27  .ti +5n  .ti +5n
28  .B const char **\fIerrptr\fR);  .B const char **\fIerrptr\fP);
29  .PP  .PP
30  .br  .B void pcre_free_study(pcre_extra *\fIextra\fP);
31  .B int pcre_exec(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"  .PP
32    .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
33  .ti +5n  .ti +5n
34  .B "const char *\fIsubject\fR," int \fIlength\fR, int \fIstartoffset\fR,  .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
35  .ti +5n  .ti +5n
36  .B int \fIoptions\fR, int *\fIovector\fR, int \fIovecsize\fR);  .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
37  .PP  .PP
38  .br  .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
39  .B int pcre_copy_named_substring(const pcre *\fIcode\fR,  .ti +5n
40    .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
41    .ti +5n
42    .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
43    .ti +5n
44    .B int *\fIworkspace\fP, int \fIwscount\fP);
45    .
46    .
47    .SH "PCRE NATIVE API STRING EXTRACTION FUNCTIONS"
48    .rs
49    .sp
50    .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
51  .ti +5n  .ti +5n
52  .B const char *\fIsubject\fR, int *\fIovector\fR,  .B const char *\fIsubject\fP, int *\fIovector\fP,
53  .ti +5n  .ti +5n
54  .B int \fIstringcount\fR, const char *\fIstringname\fR,  .B int \fIstringcount\fP, const char *\fIstringname\fP,
55  .ti +5n  .ti +5n
56  .B char *\fIbuffer\fR, int \fIbuffersize\fR);  .B char *\fIbuffer\fP, int \fIbuffersize\fP);
57  .PP  .PP
58  .br  .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
 .B int pcre_copy_substring(const char *\fIsubject\fR, int *\fIovector\fR,  
59  .ti +5n  .ti +5n
60  .B int \fIstringcount\fR, int \fIstringnumber\fR, char *\fIbuffer\fR,  .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
61  .ti +5n  .ti +5n
62  .B int \fIbuffersize\fR);  .B int \fIbuffersize\fP);
63  .PP  .PP
64  .br  .B int pcre_get_named_substring(const pcre *\fIcode\fP,
 .B int pcre_get_named_substring(const pcre *\fIcode\fR,  
65  .ti +5n  .ti +5n
66  .B const char *\fIsubject\fR, int *\fIovector\fR,  .B const char *\fIsubject\fP, int *\fIovector\fP,
67  .ti +5n  .ti +5n
68  .B int \fIstringcount\fR, const char *\fIstringname\fR,  .B int \fIstringcount\fP, const char *\fIstringname\fP,
69    .ti +5n
70    .B const char **\fIstringptr\fP);
71    .PP
72    .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
73  .ti +5n  .ti +5n
74  .B const char **\fIstringptr\fR);  .B const char *\fIname\fP);
75  .PP  .PP
76  .br  .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
 .B int pcre_get_stringnumber(const pcre *\fIcode\fR,  
77  .ti +5n  .ti +5n
78  .B const char *\fIname\fR);  .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
79  .PP  .PP
80  .br  .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
 .B int pcre_get_substring(const char *\fIsubject\fR, int *\fIovector\fR,  
81  .ti +5n  .ti +5n
82  .B int \fIstringcount\fR, int \fIstringnumber\fR,  .B int \fIstringcount\fP, int \fIstringnumber\fP,
83  .ti +5n  .ti +5n
84  .B const char **\fIstringptr\fR);  .B const char **\fIstringptr\fP);
85  .PP  .PP
86  .br  .B int pcre_get_substring_list(const char *\fIsubject\fP,
 .B int pcre_get_substring_list(const char *\fIsubject\fR,  
87  .ti +5n  .ti +5n
88  .B int *\fIovector\fR, int \fIstringcount\fR, "const char ***\fIlistptr\fR);"  .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
89  .PP  .PP
90  .br  .B void pcre_free_substring(const char *\fIstringptr\fP);
 .B void pcre_free_substring(const char *\fIstringptr\fR);  
91  .PP  .PP
92  .br  .B void pcre_free_substring_list(const char **\fIstringptr\fP);
93  .B void pcre_free_substring_list(const char **\fIstringptr\fR);  .
94    .
95    .SH "PCRE NATIVE API AUXILIARY FUNCTIONS"
96    .rs
97    .sp
98    .B pcre_jit_stack *pcre_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
99    .PP
100    .B void pcre_jit_stack_free(pcre_jit_stack *\fIstack\fP);
101    .PP
102    .B void pcre_assign_jit_stack(pcre_extra *\fIextra\fP,
103    .ti +5n
104    .B pcre_jit_callback \fIcallback\fP, void *\fIdata\fP);
105  .PP  .PP
 .br  
106  .B const unsigned char *pcre_maketables(void);  .B const unsigned char *pcre_maketables(void);
107  .PP  .PP
108  .br  .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
 .B int pcre_fullinfo(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"  
109  .ti +5n  .ti +5n
110  .B int \fIwhat\fR, void *\fIwhere\fR);  .B int \fIwhat\fP, void *\fIwhere\fP);
111  .PP  .PP
112  .br  .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
 .B int pcre_info(const pcre *\fIcode\fR, int *\fIoptptr\fR, int  
 .B *\fIfirstcharptr\fR);  
113  .PP  .PP
114  .br  .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
 .B int pcre_config(int \fIwhat\fR, void *\fIwhere\fR);  
115  .PP  .PP
116  .br  .B const char *pcre_version(void);
 .B char *pcre_version(void);  
117  .PP  .PP
118  .br  .B int pcre_pattern_to_host_byte_order(pcre *\fIcode\fP,
119    .ti +5n
120    .B pcre_extra *\fIextra\fP, const unsigned char *\fItables\fP);
121    .
122    .
123    .SH "PCRE NATIVE API INDIRECTED FUNCTIONS"
124    .rs
125    .sp
126  .B void *(*pcre_malloc)(size_t);  .B void *(*pcre_malloc)(size_t);
127  .PP  .PP
 .br  
128  .B void (*pcre_free)(void *);  .B void (*pcre_free)(void *);
129  .PP  .PP
130  .br  .B void *(*pcre_stack_malloc)(size_t);
131    .PP
132    .B void (*pcre_stack_free)(void *);
133    .PP
134  .B int (*pcre_callout)(pcre_callout_block *);  .B int (*pcre_callout)(pcre_callout_block *);
135    .
136  .SH PCRE API  .
137    .SH "PCRE 8-BIT AND 16-BIT LIBRARIES"
138    .rs
139    .sp
140    From release 8.30, PCRE can be compiled as a library for handling 16-bit
141    character strings as well as, or instead of, the original library that handles
142    8-bit character strings. To avoid too much complication, this document
143    describes the 8-bit versions of the functions, with only occasional references
144    to the 16-bit library.
145    .P
146    The 16-bit functions operate in the same way as their 8-bit counterparts; they
147    just use different data types for their arguments and results, and their names
148    start with \fBpcre16_\fP instead of \fBpcre_\fP. For every option that has UTF8
149    in its name (for example, PCRE_UTF8), there is a corresponding 16-bit name with
150    UTF8 replaced by UTF16. This facility is in fact just cosmetic; the 16-bit
151    option names define the same bit values.
152    .P
153    References to bytes and UTF-8 in this document should be read as references to
154    16-bit data quantities and UTF-16 when using the 16-bit library, unless
155    specified otherwise. More details of the specific differences for the 16-bit
156    library are given in the
157    .\" HREF
158    \fBpcre16\fP
159    .\"
160    page.
161    .
162    .
163    .SH "PCRE API OVERVIEW"
164  .rs  .rs
165  .sp  .sp
166  PCRE has its own native API, which is described in this document. There is also  PCRE has its own native API, which is described in this document. There are
167  a set of wrapper functions that correspond to the POSIX regular expression API.  also some wrapper functions (for the 8-bit library only) that correspond to the
168  These are described in the \fBpcreposix\fR documentation.  POSIX regular expression API, but they do not give access to all the
169    functionality. They are described in the
170  The native API function prototypes are defined in the header file \fBpcre.h\fR,  .\" HREF
171  and on Unix systems the library itself is called \fBlibpcre.a\fR, so can be  \fBpcreposix\fP
172  accessed by adding \fB-lpcre\fR to the command for linking an application which  .\"
173  calls it. The header file defines the macros PCRE_MAJOR and PCRE_MINOR to  documentation. Both of these APIs define a set of C function calls. A C++
174  contain the major and minor release numbers for the library. Applications can  wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
175  use these to include support for different releases.  documented in the
176    .\" HREF
177  The functions \fBpcre_compile()\fR, \fBpcre_study()\fR, and \fBpcre_exec()\fR  \fBpcrecpp\fP
178  are used for compiling and matching regular expressions. A sample program that  .\"
179  demonstrates the simplest way of using them is given in the file  page.
180  \fIpcredemo.c\fR. The \fBpcresample\fR documentation describes how to run it.  .P
181    The native API C function prototypes are defined in the header file
182  There are convenience functions for extracting captured substrings from a  \fBpcre.h\fP, and on Unix-like systems the (8-bit) library itself is called
183  matched subject string. They are:  \fBlibpcre\fP. It can normally be accessed by adding \fB-lpcre\fP to the
184    command for linking an application that uses PCRE. The header file defines the
185    \fBpcre_copy_substring()\fR  macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
186    \fBpcre_copy_named_substring()\fR  for the library. Applications can use these to include support for different
187    \fBpcre_get_substring()\fR  releases of PCRE.
188    \fBpcre_get_named_substring()\fR  .P
189    \fBpcre_get_substring_list()\fR  In a Windows environment, if you want to statically link an application program
190    against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
191  \fBpcre_free_substring()\fR and \fBpcre_free_substring_list()\fR are also  including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
192    \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
193    \fB__declspec(dllimport)\fP, with unwanted results.
194    .P
195    The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
196    and \fBpcre_exec()\fP are used for compiling and matching regular expressions
197    in a Perl-compatible manner. A sample program that demonstrates the simplest
198    way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
199    source distribution. A listing of this program is given in the
200    .\" HREF
201    \fBpcredemo\fP
202    .\"
203    documentation, and the
204    .\" HREF
205    \fBpcresample\fP
206    .\"
207    documentation describes how to compile and run it.
208    .P
209    Just-in-time compiler support is an optional feature of PCRE that can be built
210    in appropriate hardware environments. It greatly speeds up the matching
211    performance of many patterns. Simple programs can easily request that it be
212    used if available, by setting an option that is ignored when it is not
213    relevant. More complicated programs might need to make use of the functions
214    \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
215    \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
216    These functions are discussed in the
217    .\" HREF
218    \fBpcrejit\fP
219    .\"
220    documentation.
221    .P
222    A second matching function, \fBpcre_dfa_exec()\fP, which is not
223    Perl-compatible, is also provided. This uses a different algorithm for the
224    matching. The alternative algorithm finds all possible matches (at a given
225    point in the subject), and scans the subject just once (unless there are
226    lookbehind assertions). However, this algorithm does not return captured
227    substrings. A description of the two matching algorithms and their advantages
228    and disadvantages is given in the
229    .\" HREF
230    \fBpcrematching\fP
231    .\"
232    documentation.
233    .P
234    In addition to the main compiling and matching functions, there are convenience
235    functions for extracting captured substrings from a subject string that is
236    matched by \fBpcre_exec()\fP. They are:
237    .sp
238      \fBpcre_copy_substring()\fP
239      \fBpcre_copy_named_substring()\fP
240      \fBpcre_get_substring()\fP
241      \fBpcre_get_named_substring()\fP
242      \fBpcre_get_substring_list()\fP
243      \fBpcre_get_stringnumber()\fP
244      \fBpcre_get_stringtable_entries()\fP
245    .sp
246    \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
247  provided, to free the memory used for extracted strings.  provided, to free the memory used for extracted strings.
248    .P
249  The function \fBpcre_maketables()\fR is used (optionally) to build a set of  The function \fBpcre_maketables()\fP is used to build a set of character tables
250  character tables in the current locale for passing to \fBpcre_compile()\fR.  in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
251    or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
252  The function \fBpcre_fullinfo()\fR is used to find out information about a  specialist use. Most commonly, no special tables are passed, in which case
253  compiled pattern; \fBpcre_info()\fR is an obsolete version which returns only  internal tables that are generated when PCRE is built are used.
254  some of the available information, but is retained for backwards compatibility.  .P
255  The function \fBpcre_version()\fR returns a pointer to a string containing the  The function \fBpcre_fullinfo()\fP is used to find out information about a
256  version of PCRE and its date of release.  compiled pattern. The function \fBpcre_version()\fP returns a pointer to a
257    string containing the version of PCRE and its date of release.
258  The global variables \fBpcre_malloc\fR and \fBpcre_free\fR initially contain  .P
259  the entry points of the standard \fBmalloc()\fR and \fBfree()\fR functions  The function \fBpcre_refcount()\fP maintains a reference count in a data block
260    containing a compiled pattern. This is provided for the benefit of
261    object-oriented applications.
262    .P
263    The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
264    the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
265  respectively. PCRE calls the memory management functions via these variables,  respectively. PCRE calls the memory management functions via these variables,
266  so a calling program can replace them if it wishes to intercept the calls. This  so a calling program can replace them if it wishes to intercept the calls. This
267  should be done before calling any PCRE functions.  should be done before calling any PCRE functions.
268    .P
269  The global variable \fBpcre_callout\fR initially contains NULL. It can be set  The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
270    indirections to memory management functions. These special functions are used
271    only when PCRE is compiled to use the heap for remembering data, instead of
272    recursive function calls, when running the \fBpcre_exec()\fP function. See the
273    .\" HREF
274    \fBpcrebuild\fP
275    .\"
276    documentation for details of how to do this. It is a non-standard way of
277    building PCRE, for use in environments that have limited stacks. Because of the
278    greater use of memory management, it runs more slowly. Separate functions are
279    provided so that special-purpose external code can be used for this case. When
280    used, these functions are always called in a stack-like manner (last obtained,
281    first freed), and always for memory blocks of the same size. There is a
282    discussion about PCRE's stack usage in the
283    .\" HREF
284    \fBpcrestack\fP
285    .\"
286    documentation.
287    .P
288    The global variable \fBpcre_callout\fP initially contains NULL. It can be set
289  by the caller to a "callout" function, which PCRE will then call at specified  by the caller to a "callout" function, which PCRE will then call at specified
290  points during a matching operation. Details are given in the \fBpcrecallout\fR  points during a matching operation. Details are given in the
291    .\" HREF
292    \fBpcrecallout\fP
293    .\"
294  documentation.  documentation.
295    .
296    .
297    .\" HTML <a name="newlines"></a>
298    .SH NEWLINES
299    .rs
300    .sp
301    PCRE supports five different conventions for indicating line breaks in
302    strings: a single CR (carriage return) character, a single LF (linefeed)
303    character, the two-character sequence CRLF, any of the three preceding, or any
304    Unicode newline sequence. The Unicode newline sequences are the three just
305    mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
306    U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
307    (paragraph separator, U+2029).
308    .P
309    Each of the first three conventions is used by at least one operating system as
310    its standard newline sequence. When PCRE is built, a default can be specified.
311    The default default is LF, which is the Unix standard. When PCRE is run, the
312    default can be overridden, either when a pattern is compiled, or when it is
313    matched.
314    .P
315    At compile time, the newline convention can be specified by the \fIoptions\fP
316    argument of \fBpcre_compile()\fP, or it can be specified by special text at the
317    start of the pattern itself; this overrides any other settings. See the
318    .\" HREF
319    \fBpcrepattern\fP
320    .\"
321    page for details of the special character sequences.
322    .P
323    In the PCRE documentation the word "newline" is used to mean "the character or
324    pair of characters that indicate a line break". The choice of newline
325    convention affects the handling of the dot, circumflex, and dollar
326    metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
327    recognized line ending sequence, the match position advancement for a
328    non-anchored pattern. There is more detail about this in the
329    .\" HTML <a href="#execoptions">
330    .\" </a>
331    section on \fBpcre_exec()\fP options
332    .\"
333    below.
334    .P
335    The choice of newline convention does not affect the interpretation of
336    the \en or \er escape sequences, nor does it affect what \eR matches, which is
337    controlled in a similar way, but by separate options.
338    .
339    .
340  .SH MULTITHREADING  .SH MULTITHREADING
341  .rs  .rs
342  .sp  .sp
343  The PCRE functions can be used in multi-threading applications, with the  The PCRE functions can be used in multi-threading applications, with the
344  proviso that the memory management functions pointed to by \fBpcre_malloc\fR  proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
345  and \fBpcre_free\fR, and the callout function pointed to by \fBpcre_callout\fR,  \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
346  are shared by all threads.  callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
347    .P
348  The compiled form of a regular expression is not altered during matching, so  The compiled form of a regular expression is not altered during matching, so
349  the same compiled pattern can safely be used by several threads at once.  the same compiled pattern can safely be used by several threads at once.
350    .P
351  .SH CHECKING BUILD-TIME OPTIONS  If the just-in-time optimization feature is being used, it needs separate
352    memory stack areas for each thread. See the
353    .\" HREF
354    \fBpcrejit\fP
355    .\"
356    documentation for more details.
357    .
358    .
359    .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
360  .rs  .rs
361  .sp  .sp
362  .B int pcre_config(int \fIwhat\fR, void *\fIwhere\fR);  The compiled form of a regular expression can be saved and re-used at a later
363    time, possibly by a different program, and even on a host other than the one on
364    which it was compiled. Details are given in the
365    .\" HREF
366    \fBpcreprecompile\fP
367    .\"
368    documentation, which includes a description of the
369    \fBpcre_pattern_to_host_byte_order()\fP function. However, compiling a regular
370    expression with one version of PCRE for use with a different version is not
371    guaranteed to work and may cause crashes.
372    .
373    .
374    .SH "CHECKING BUILD-TIME OPTIONS"
375    .rs
376    .sp
377    .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
378  .PP  .PP
379  The function \fBpcre_config()\fR makes it possible for a PCRE client to  The function \fBpcre_config()\fP makes it possible for a PCRE client to
380  discover which optional features have been compiled into the PCRE library. The  discover which optional features have been compiled into the PCRE library. The
381  .\" HREF  .\" HREF
382  \fBpcrebuild\fR  \fBpcrebuild\fP
383  .\"  .\"
384  documentation has more details about these optional features.  documentation has more details about these optional features.
385    .P
386  The first argument for \fBpcre_config()\fR is an integer, specifying which  The first argument for \fBpcre_config()\fP is an integer, specifying which
387  information is required; the second argument is a pointer to a variable into  information is required; the second argument is a pointer to a variable into
388  which the information is placed. The following information is available:  which the information is placed. The returned value is zero on success, or the
389    negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
390    not recognized. The following information is available:
391    .sp
392    PCRE_CONFIG_UTF8    PCRE_CONFIG_UTF8
393    .sp
394  The output is an integer that is set to one if UTF-8 support is available;  The output is an integer that is set to one if UTF-8 support is available;
395  otherwise it is set to zero.  otherwise it is set to zero. If this option is given to the 16-bit version of
396    this function, \fBpcre16_config()\fP, the result is PCRE_ERROR_BADOPTION.
397    .sp
398      PCRE_CONFIG_UTF16
399    .sp
400    The output is an integer that is set to one if UTF-16 support is available;
401    otherwise it is set to zero. This value should normally be given to the 16-bit
402    version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
403    version of this function, the result is PCRE_ERROR_BADOPTION.
404    .sp
405      PCRE_CONFIG_UNICODE_PROPERTIES
406    .sp
407    The output is an integer that is set to one if support for Unicode character
408    properties is available; otherwise it is set to zero.
409    .sp
410      PCRE_CONFIG_JIT
411    .sp
412    The output is an integer that is set to one if support for just-in-time
413    compiling is available; otherwise it is set to zero.
414    .sp
415      PCRE_CONFIG_JITTARGET
416    .sp
417    The output is a pointer to a zero-terminated "const char *" string. If JIT
418    support is available, the string contains the name of the architecture for
419    which the JIT compiler is configured, for example "x86 32bit (little endian +
420    unaligned)". If JIT support is not available, the result is NULL.
421    .sp
422    PCRE_CONFIG_NEWLINE    PCRE_CONFIG_NEWLINE
423    .sp
424  The output is an integer that is set to the value of the code that is used for  The output is an integer whose value specifies the default character sequence
425  the newline character. It is either linefeed (10) or carriage return (13), and  that is recognized as meaning "newline". The four values that are supported
426  should normally be the standard character for your operating system.  are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF, and -1 for ANY.
427    Though they are derived from ASCII, the same values are returned in EBCDIC
428    environments. The default should normally correspond to the standard sequence
429    for your operating system.
430    .sp
431      PCRE_CONFIG_BSR
432    .sp
433    The output is an integer whose value indicates what character sequences the \eR
434    escape sequence matches by default. A value of 0 means that \eR matches any
435    Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
436    or CRLF. The default can be overridden when a pattern is compiled or matched.
437    .sp
438    PCRE_CONFIG_LINK_SIZE    PCRE_CONFIG_LINK_SIZE
439    .sp
440  The output is an integer that contains the number of bytes used for internal  The output is an integer that contains the number of bytes used for internal
441  linkage in compiled regular expressions. The value is 2, 3, or 4. Larger values  linkage in compiled regular expressions. For the 8-bit library, the value can
442  allow larger regular expressions to be compiled, at the expense of slower  be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
443  matching. The default value of 2 is sufficient for all but the most massive  a number of bytes. The default value of 2 is sufficient for all but the most
444  patterns, since it allows the compiled pattern to be up to 64K in size.  massive patterns, since it allows the compiled pattern to be up to 64K in size.
445    Larger values allow larger regular expressions to be compiled, at the expense
446    of slower matching.
447    .sp
448    PCRE_CONFIG_POSIX_MALLOC_THRESHOLD    PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
449    .sp
450  The output is an integer that contains the threshold above which the POSIX  The output is an integer that contains the threshold above which the POSIX
451  interface uses \fBmalloc()\fR for output vectors. Further details are given in  interface uses \fBmalloc()\fP for output vectors. Further details are given in
452  the \fBpcreposix\fR documentation.  the
453    .\" HREF
454    \fBpcreposix\fP
455    .\"
456    documentation.
457    .sp
458    PCRE_CONFIG_MATCH_LIMIT    PCRE_CONFIG_MATCH_LIMIT
459    .sp
460  The output is an integer that gives the default limit for the number of  The output is a long integer that gives the default limit for the number of
461  internal matching function calls in a \fBpcre_exec()\fR execution. Further  internal matching function calls in a \fBpcre_exec()\fP execution. Further
462  details are given with \fBpcre_exec()\fR below.  details are given with \fBpcre_exec()\fP below.
463    .sp
464  .SH COMPILING A PATTERN    PCRE_CONFIG_MATCH_LIMIT_RECURSION
465  .rs  .sp
466  .sp  The output is a long integer that gives the default limit for the depth of
467  .B pcre *pcre_compile(const char *\fIpattern\fR, int \fIoptions\fR,  recursion when calling the internal matching function in a \fBpcre_exec()\fP
468  .ti +5n  execution. Further details are given with \fBpcre_exec()\fP below.
469  .B const char **\fIerrptr\fR, int *\fIerroffset\fR,  .sp
470  .ti +5n    PCRE_CONFIG_STACKRECURSE
471  .B const unsigned char *\fItableptr\fR);  .sp
472  .PP  The output is an integer that is set to one if internal recursion when running
473    \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
474  The function \fBpcre_compile()\fR is called to compile a pattern into an  to remember their state. This is the usual way that PCRE is compiled. The
475  internal form. The pattern is a C string terminated by a binary zero, and  output is zero if PCRE was compiled to use blocks of data on the heap instead
476  is passed in the argument \fIpattern\fR. A pointer to a single block of memory  of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
477  that is obtained via \fBpcre_malloc\fR is returned. This contains the compiled  \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
478  code and related data. The \fBpcre\fR type is defined for the returned block;  avoiding the use of the stack.
479  this is a typedef for a structure whose contents are not externally defined. It  .
480  is up to the caller to free the memory when it is no longer required.  .
481    .SH "COMPILING A PATTERN"
482    .rs
483    .sp
484    .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
485    .ti +5n
486    .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
487    .ti +5n
488    .B const unsigned char *\fItableptr\fP);
489    .sp
490    .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
491    .ti +5n
492    .B int *\fIerrorcodeptr\fP,
493    .ti +5n
494    .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
495    .ti +5n
496    .B const unsigned char *\fItableptr\fP);
497    .P
498    Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
499    called to compile a pattern into an internal form. The only difference between
500    the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
501    \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
502    too much repetition, we refer just to \fBpcre_compile()\fP below, but the
503    information applies equally to \fBpcre_compile2()\fP.
504    .P
505    The pattern is a C string terminated by a binary zero, and is passed in the
506    \fIpattern\fP argument. A pointer to a single block of memory that is obtained
507    via \fBpcre_malloc\fP is returned. This contains the compiled code and related
508    data. The \fBpcre\fP type is defined for the returned block; this is a typedef
509    for a structure whose contents are not externally defined. It is up to the
510    caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
511    .P
512  Although the compiled code of a PCRE regex is relocatable, that is, it does not  Although the compiled code of a PCRE regex is relocatable, that is, it does not
513  depend on memory location, the complete \fBpcre\fR data block is not  depend on memory location, the complete \fBpcre\fP data block is not
514  fully relocatable, because it contains a copy of the \fItableptr\fR argument,  fully relocatable, because it may contain a copy of the \fItableptr\fP
515  which is an address (see below).  argument, which is an address (see below).
516    .P
517  The \fIoptions\fR argument contains independent bits that affect the  The \fIoptions\fP argument contains various bit settings that affect the
518  compilation. It should be zero if no options are required. Some of the options,  compilation. It should be zero if no options are required. The available
519  in particular, those that are compatible with Perl, can also be set and unset  options are described below. Some of them (in particular, those that are
520  from within the pattern (see the detailed description of regular expressions  compatible with Perl, but some others as well) can also be set and unset from
521  in the \fBpcrepattern\fR documentation). For these options, the contents of the  within the pattern (see the detailed description in the
522  \fIoptions\fR argument specifies their initial settings at the start of  .\" HREF
523  compilation and execution. The PCRE_ANCHORED option can be set at the time of  \fBpcrepattern\fP
524  matching as well as at compile time.  .\"
525    documentation). For those options that can be different in different parts of
526  If \fIerrptr\fR is NULL, \fBpcre_compile()\fR returns NULL immediately.  the pattern, the contents of the \fIoptions\fP argument specifies their
527  Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fR returns  settings at the start of compilation and execution. The PCRE_ANCHORED,
528  NULL, and sets the variable pointed to by \fIerrptr\fR to point to a textual  PCRE_BSR_\fIxxx\fP, PCRE_NEWLINE_\fIxxx\fP, PCRE_NO_UTF8_CHECK, and
529  error message. The offset from the start of the pattern to the character where  PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
530  the error was discovered is placed in the variable pointed to by  compile time.
531  \fIerroffset\fR, which must not be NULL. If it is, an immediate error is given.  .P
532    If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
533  If the final argument, \fItableptr\fR, is NULL, PCRE uses a default set of  Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
534  character tables which are built when it is compiled, using the default C  NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
535  locale. Otherwise, \fItableptr\fR must be the result of a call to  error message. This is a static string that is part of the library. You must
536  \fBpcre_maketables()\fR. See the section on locale support below.  not try to free it. Normally, the offset from the start of the pattern to the
537    byte that was being processed when the error was discovered is placed in the
538  This code fragment shows a typical straightforward call to \fBpcre_compile()\fR:  variable pointed to by \fIerroffset\fP, which must not be NULL (if it is, an
539    immediate error is given). However, for an invalid UTF-8 string, the offset is
540    that of the first byte of the failing character.
541    .P
542    Some errors are not detected until the whole pattern has been scanned; in these
543    cases, the offset passed back is the length of the pattern. Note that the
544    offset is in bytes, not characters, even in UTF-8 mode. It may sometimes point
545    into the middle of a UTF-8 character.
546    .P
547    If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
548    \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
549    returned via this argument in the event of an error. This is in addition to the
550    textual error message. Error codes and messages are listed below.
551    .P
552    If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
553    character tables that are built when PCRE is compiled, using the default C
554    locale. Otherwise, \fItableptr\fP must be an address that is the result of a
555    call to \fBpcre_maketables()\fP. This value is stored with the compiled
556    pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
557    passed to it. For more discussion, see the section on locale support below.
558    .P
559    This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
560    .sp
561    pcre *re;    pcre *re;
562    const char *error;    const char *error;
563    int erroffset;    int erroffset;
# Line 265  This code fragment shows a typical strai Line 567  This code fragment shows a typical strai
567      &error,           /* for error message */      &error,           /* for error message */
568      &erroffset,       /* for error offset */      &erroffset,       /* for error offset */
569      NULL);            /* use default character tables */      NULL);            /* use default character tables */
570    .sp
571  The following option bits are defined:  The following names for option bits are defined in the \fBpcre.h\fP header
572    file:
573    .sp
574    PCRE_ANCHORED    PCRE_ANCHORED
575    .sp
576  If this bit is set, the pattern is forced to be "anchored", that is, it is  If this bit is set, the pattern is forced to be "anchored", that is, it is
577  constrained to match only at the first matching point in the string which is  constrained to match only at the first matching point in the string that is
578  being searched (the "subject string"). This effect can also be achieved by  being searched (the "subject string"). This effect can also be achieved by
579  appropriate constructs in the pattern itself, which is the only way to do it in  appropriate constructs in the pattern itself, which is the only way to do it in
580  Perl.  Perl.
581    .sp
582      PCRE_AUTO_CALLOUT
583    .sp
584    If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
585    all with number 255, before each pattern item. For discussion of the callout
586    facility, see the
587    .\" HREF
588    \fBpcrecallout\fP
589    .\"
590    documentation.
591    .sp
592      PCRE_BSR_ANYCRLF
593      PCRE_BSR_UNICODE
594    .sp
595    These options (which are mutually exclusive) control what the \eR escape
596    sequence matches. The choice is either to match only CR, LF, or CRLF, or to
597    match any Unicode newline sequence. The default is specified when PCRE is
598    built. It can be overridden from within the pattern, or by setting an option
599    when a compiled pattern is matched.
600    .sp
601    PCRE_CASELESS    PCRE_CASELESS
602    .sp
603  If this bit is set, letters in the pattern match both upper and lower case  If this bit is set, letters in the pattern match both upper and lower case
604  letters. It is equivalent to Perl's /i option, and it can be changed within a  letters. It is equivalent to Perl's /i option, and it can be changed within a
605  pattern by a (?i) option setting.  pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
606    concept of case for characters whose values are less than 128, so caseless
607    matching is always possible. For characters with higher values, the concept of
608    case is supported if PCRE is compiled with Unicode property support, but not
609    otherwise. If you want to use caseless matching for characters 128 and above,
610    you must ensure that PCRE is compiled with Unicode property support as well as
611    with UTF-8 support.
612    .sp
613    PCRE_DOLLAR_ENDONLY    PCRE_DOLLAR_ENDONLY
614    .sp
615  If this bit is set, a dollar metacharacter in the pattern matches only at the  If this bit is set, a dollar metacharacter in the pattern matches only at the
616  end of the subject string. Without this option, a dollar also matches  end of the subject string. Without this option, a dollar also matches
617  immediately before the final character if it is a newline (but not before any  immediately before a newline at the end of the string (but not before any other
618  other newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is  newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
619  set. There is no equivalent to this option in Perl, and no way to set it within  There is no equivalent to this option in Perl, and no way to set it within a
620  a pattern.  pattern.
621    .sp
622    PCRE_DOTALL    PCRE_DOTALL
623    .sp
624  If this bit is set, a dot metacharater in the pattern matches all characters,  If this bit is set, a dot metacharacter in the pattern matches a character of
625  including newlines. Without it, newlines are excluded. This option is  any value, including one that indicates a newline. However, it only ever
626    matches one character, even if newlines are coded as CRLF. Without this option,
627    a dot does not match when the current position is at a newline. This option is
628  equivalent to Perl's /s option, and it can be changed within a pattern by a  equivalent to Perl's /s option, and it can be changed within a pattern by a
629  (?s) option setting. A negative class such as [^a] always matches a newline  (?s) option setting. A negative class such as [^a] always matches newline
630  character, independent of the setting of this option.  characters, independent of the setting of this option.
631    .sp
632      PCRE_DUPNAMES
633    .sp
634    If this bit is set, names used to identify capturing subpatterns need not be
635    unique. This can be helpful for certain types of pattern when it is known that
636    only one instance of the named subpattern can ever be matched. There are more
637    details of named subpatterns below; see also the
638    .\" HREF
639    \fBpcrepattern\fP
640    .\"
641    documentation.
642    .sp
643    PCRE_EXTENDED    PCRE_EXTENDED
644    .sp
645  If this bit is set, whitespace data characters in the pattern are totally  If this bit is set, whitespace data characters in the pattern are totally
646  ignored except when escaped or inside a character class. Whitespace does not  ignored except when escaped or inside a character class. Whitespace does not
647  include the VT character (code 11). In addition, characters between an  include the VT character (code 11). In addition, characters between an
648  unescaped # outside a character class and the next newline character,  unescaped # outside a character class and the next newline, inclusive, are also
649  inclusive, are also ignored. This is equivalent to Perl's /x option, and it can  ignored. This is equivalent to Perl's /x option, and it can be changed within a
650  be changed within a pattern by a (?x) option setting.  pattern by a (?x) option setting.
651    .P
652    Which characters are interpreted as newlines is controlled by the options
653    passed to \fBpcre_compile()\fP or by a special sequence at the start of the
654    pattern, as described in the section entitled
655    .\" HTML <a href="pcrepattern.html#newlines">
656    .\" </a>
657    "Newline conventions"
658    .\"
659    in the \fBpcrepattern\fP documentation. Note that the end of this type of
660    comment is a literal newline sequence in the pattern; escape sequences that
661    happen to represent a newline do not count.
662    .P
663  This option makes it possible to include comments inside complicated patterns.  This option makes it possible to include comments inside complicated patterns.
664  Note, however, that this applies only to data characters. Whitespace characters  Note, however, that this applies only to data characters. Whitespace characters
665  may never appear within special character sequences in a pattern, for example  may never appear within special character sequences in a pattern, for example
666  within the sequence (?( which introduces a conditional subpattern.  within the sequence (?( that introduces a conditional subpattern.
667    .sp
668    PCRE_EXTRA    PCRE_EXTRA
669    .sp
670  This option was invented in order to turn on additional functionality of PCRE  This option was invented in order to turn on additional functionality of PCRE
671  that is incompatible with Perl, but it is currently of very little use. When  that is incompatible with Perl, but it is currently of very little use. When
672  set, any backslash in a pattern that is followed by a letter that has no  set, any backslash in a pattern that is followed by a letter that has no
673  special meaning causes an error, thus reserving these combinations for future  special meaning causes an error, thus reserving these combinations for future
674  expansion. By default, as in Perl, a backslash followed by a letter with no  expansion. By default, as in Perl, a backslash followed by a letter with no
675  special meaning is treated as a literal. There are at present no other features  special meaning is treated as a literal. (Perl can, however, be persuaded to
676  controlled by this option. It can also be set by a (?X) option setting within a  give an error for this, by running it with the -w option.) There are at present
677  pattern.  no other features controlled by this option. It can also be set by a (?X)
678    option setting within a pattern.
679    .sp
680      PCRE_FIRSTLINE
681    .sp
682    If this option is set, an unanchored pattern is required to match before or at
683    the first newline in the subject string, though the matched text may continue
684    over the newline.
685    .sp
686      PCRE_JAVASCRIPT_COMPAT
687    .sp
688    If this option is set, PCRE's behaviour is changed in some ways so that it is
689    compatible with JavaScript rather than Perl. The changes are as follows:
690    .P
691    (1) A lone closing square bracket in a pattern causes a compile-time error,
692    because this is illegal in JavaScript (by default it is treated as a data
693    character). Thus, the pattern AB]CD becomes illegal when this option is set.
694    .P
695    (2) At run time, a back reference to an unset subpattern group matches an empty
696    string (by default this causes the current matching alternative to fail). A
697    pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
698    an "a" in the subject), whereas it fails by default, for Perl compatibility.
699    .P
700    (3) \eU matches an upper case "U" character; by default \eU causes a compile
701    time error (Perl uses \eU to upper case subsequent characters).
702    .P
703    (4) \eu matches a lower case "u" character unless it is followed by four
704    hexadecimal digits, in which case the hexadecimal number defines the code point
705    to match. By default, \eu causes a compile time error (Perl uses it to upper
706    case the following character).
707    .P
708    (5) \ex matches a lower case "x" character unless it is followed by two
709    hexadecimal digits, in which case the hexadecimal number defines the code point
710    to match. By default, as in Perl, a hexadecimal number is always expected after
711    \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
712    binary zero character followed by z).
713    .sp
714    PCRE_MULTILINE    PCRE_MULTILINE
715    .sp
716  By default, PCRE treats the subject string as consisting of a single "line" of  By default, PCRE treats the subject string as consisting of a single line of
717  characters (even if it actually contains several newlines). The "start of line"  characters (even if it actually contains newlines). The "start of line"
718  metacharacter (^) matches only at the start of the string, while the "end of  metacharacter (^) matches only at the start of the string, while the "end of
719  line" metacharacter ($) matches only at the end of the string, or before a  line" metacharacter ($) matches only at the end of the string, or before a
720  terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as  terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
721  Perl.  Perl.
722    .P
723  When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs  When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
724  match immediately following or immediately before any newline in the subject  match immediately following or immediately before internal newlines in the
725  string, respectively, as well as at the very start and end. This is equivalent  subject string, respectively, as well as at the very start and end. This is
726  to Perl's /m option, and it can be changed within a pattern by a (?m) option  equivalent to Perl's /m option, and it can be changed within a pattern by a
727  setting. If there are no "\\n" characters in a subject string, or no  (?m) option setting. If there are no newlines in a subject string, or no
728  occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.  occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
729    .sp
730      PCRE_NEWLINE_CR
731      PCRE_NEWLINE_LF
732      PCRE_NEWLINE_CRLF
733      PCRE_NEWLINE_ANYCRLF
734      PCRE_NEWLINE_ANY
735    .sp
736    These options override the default newline definition that was chosen when PCRE
737    was built. Setting the first or the second specifies that a newline is
738    indicated by a single character (CR or LF, respectively). Setting
739    PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
740    CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
741    preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
742    that any Unicode newline sequence should be recognized. The Unicode newline
743    sequences are the three just mentioned, plus the single characters VT (vertical
744    tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
745    separator, U+2028), and PS (paragraph separator, U+2029). For the 8-bit
746    library, the last two are recognized only in UTF-8 mode.
747    .P
748    The newline setting in the options word uses three bits that are treated
749    as a number, giving eight possibilities. Currently only six are used (default
750    plus the five values above). This means that if you set more than one newline
751    option, the combination may or may not be sensible. For example,
752    PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
753    other combinations may yield unused numbers and cause an error.
754    .P
755    The only time that a line break in a pattern is specially recognized when
756    compiling is when PCRE_EXTENDED is set. CR and LF are whitespace characters,
757    and so are ignored in this mode. Also, an unescaped # outside a character class
758    indicates a comment that lasts until after the next line break sequence. In
759    other circumstances, line break sequences in patterns are treated as literal
760    data.
761    .P
762    The newline option that is set at compile time becomes the default that is used
763    for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
764    .sp
765    PCRE_NO_AUTO_CAPTURE    PCRE_NO_AUTO_CAPTURE
766    .sp
767  If this option is set, it disables the use of numbered capturing parentheses in  If this option is set, it disables the use of numbered capturing parentheses in
768  the pattern. Any opening parenthesis that is not followed by ? behaves as if it  the pattern. Any opening parenthesis that is not followed by ? behaves as if it
769  were followed by ?: but named parentheses can still be used for capturing (and  were followed by ?: but named parentheses can still be used for capturing (and
770  they acquire numbers in the usual way). There is no equivalent of this option  they acquire numbers in the usual way). There is no equivalent of this option
771  in Perl.  in Perl.
772    .sp
773      NO_START_OPTIMIZE
774    .sp
775    This is an option that acts at matching time; that is, it is really an option
776    for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
777    it is remembered with the compiled pattern and assumed at matching time. For
778    details see the discussion of PCRE_NO_START_OPTIMIZE
779    .\" HTML <a href="#execoptions">
780    .\" </a>
781    below.
782    .\"
783    .sp
784      PCRE_UCP
785    .sp
786    This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
787    \ew, and some of the POSIX character classes. By default, only ASCII characters
788    are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
789    classify characters. More details are given in the section on
790    .\" HTML <a href="pcre.html#genericchartypes">
791    .\" </a>
792    generic character types
793    .\"
794    in the
795    .\" HREF
796    \fBpcrepattern\fP
797    .\"
798    page. If you set PCRE_UCP, matching one of the items it affects takes much
799    longer. The option is available only if PCRE has been compiled with Unicode
800    property support.
801    .sp
802    PCRE_UNGREEDY    PCRE_UNGREEDY
803    .sp
804  This option inverts the "greediness" of the quantifiers so that they are not  This option inverts the "greediness" of the quantifiers so that they are not
805  greedy by default, but become greedy if followed by "?". It is not compatible  greedy by default, but become greedy if followed by "?". It is not compatible
806  with Perl. It can also be set by a (?U) option setting within the pattern.  with Perl. It can also be set by a (?U) option setting within the pattern.
807    .sp
808    PCRE_UTF8    PCRE_UTF8
809    .sp
810  This option causes PCRE to regard both the pattern and the subject as strings  This option causes PCRE to regard both the pattern and the subject as strings
811  of UTF-8 characters instead of single-byte character strings. However, it is  of UTF-8 characters instead of single-byte strings. However, it is available
812  available only if PCRE has been built to include UTF-8 support. If not, the use  only when PCRE is built to include UTF support. If not, the use of this option
813  of this option provokes an error. Details of how this option changes the  provokes an error. Details of how this option changes the behaviour of PCRE are
814  behaviour of PCRE are given in the  given in the
815  .\" HTML <a href="pcre.html#utf8support">  .\" HREF
816    \fBpcreunicode\fP
817    .\"
818    page.
819    .sp
820      PCRE_NO_UTF8_CHECK
821    .sp
822    When PCRE_UTF8 is set, the validity of the pattern as a UTF-8
823    string is automatically checked. There is a discussion about the
824    .\" HTML <a href="pcreunicode.html#utf8strings">
825  .\" </a>  .\" </a>
826  section on UTF-8 support  validity of UTF-8 strings
827  .\"  .\"
828  in the main  in the
829  .\" HREF  .\" HREF
830  \fBpcre\fR  \fBpcreunicode\fP
831  .\"  .\"
832  page.  page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
833    error. If you already know that your pattern is valid, and you want to skip
834  .SH STUDYING A PATTERN  this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
835    When it is set, the effect of passing an invalid UTF-8 string as a pattern is
836    undefined. It may cause your program to crash. Note that this option can also
837    be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress the
838    validity checking of subject strings.
839    .
840    .
841    .SH "COMPILATION ERROR CODES"
842    .rs
843    .sp
844    The following table lists the error codes than may be returned by
845    \fBpcre_compile2()\fP, along with the error messages that may be returned by
846    both compiling functions. Note that error messages are always 8-bit ASCII
847    strings, even in 16-bit mode. As PCRE has developed, some error codes have
848    fallen out of use. To avoid confusion, they have not been re-used.
849    .sp
850       0  no error
851       1  \e at end of pattern
852       2  \ec at end of pattern
853       3  unrecognized character follows \e
854       4  numbers out of order in {} quantifier
855       5  number too big in {} quantifier
856       6  missing terminating ] for character class
857       7  invalid escape sequence in character class
858       8  range out of order in character class
859       9  nothing to repeat
860      10  [this code is not in use]
861      11  internal error: unexpected repeat
862      12  unrecognized character after (? or (?-
863      13  POSIX named classes are supported only within a class
864      14  missing )
865      15  reference to non-existent subpattern
866      16  erroffset passed as NULL
867      17  unknown option bit(s) set
868      18  missing ) after comment
869      19  [this code is not in use]
870      20  regular expression is too large
871      21  failed to get memory
872      22  unmatched parentheses
873      23  internal error: code overflow
874      24  unrecognized character after (?<
875      25  lookbehind assertion is not fixed length
876      26  malformed number or name after (?(
877      27  conditional group contains more than two branches
878      28  assertion expected after (?(
879      29  (?R or (?[+-]digits must be followed by )
880      30  unknown POSIX class name
881      31  POSIX collating elements are not supported
882      32  this version of PCRE is compiled without UTF support
883      33  [this code is not in use]
884      34  character value in \ex{...} sequence is too large
885      35  invalid condition (?(0)
886      36  \eC not allowed in lookbehind assertion
887      37  PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
888      38  number after (?C is > 255
889      39  closing ) for (?C expected
890      40  recursive call could loop indefinitely
891      41  unrecognized character after (?P
892      42  syntax error in subpattern name (missing terminator)
893      43  two named subpatterns have the same name
894      44  invalid UTF-8 string (specifically UTF-8)
895      45  support for \eP, \ep, and \eX has not been compiled
896      46  malformed \eP or \ep sequence
897      47  unknown property name after \eP or \ep
898      48  subpattern name is too long (maximum 32 characters)
899      49  too many named subpatterns (maximum 10000)
900      50  [this code is not in use]
901      51  octal value is greater than \e377 in 8-bit non-UTF-8 mode
902      52  internal error: overran compiling workspace
903      53  internal error: previously-checked referenced subpattern
904            not found
905      54  DEFINE group contains more than one branch
906      55  repeating a DEFINE group is not allowed
907      56  inconsistent NEWLINE options
908      57  \eg is not followed by a braced, angle-bracketed, or quoted
909            name/number or by a plain number
910      58  a numbered reference must not be zero
911      59  an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
912      60  (*VERB) not recognized
913      61  number is too big
914      62  subpattern name expected
915      63  digit expected after (?+
916      64  ] is an invalid data character in JavaScript compatibility mode
917      65  different names for subpatterns of the same number are
918            not allowed
919      66  (*MARK) must have an argument
920      67  this version of PCRE is not compiled with Unicode property
921            support
922      68  \ec must be followed by an ASCII character
923      69  \ek is not followed by a braced, angle-bracketed, or quoted name
924      70  internal error: unknown opcode in find_fixedlength()
925      71  \eN is not supported in a class
926      72  too many forward references
927      73  disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
928      74  invalid UTF-16 string (specifically UTF-16)
929    .sp
930    The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
931    be used if the limits were changed when PCRE was built.
932    .
933    .
934    .\" HTML <a name="studyingapattern"></a>
935    .SH "STUDYING A PATTERN"
936  .rs  .rs
937  .sp  .sp
938  .B pcre_extra *pcre_study(const pcre *\fIcode\fR, int \fIoptions\fR,  .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
939  .ti +5n  .ti +5n
940  .B const char **\fIerrptr\fR);  .B const char **\fIerrptr\fP);
941  .PP  .PP
942  When a pattern is going to be used several times, it is worth spending more  If a compiled pattern is going to be used several times, it is worth spending
943  time analyzing it in order to speed up the time taken for matching. The  more time analyzing it in order to speed up the time taken for matching. The
944  function \fBpcre_study()\fR takes a pointer to a compiled pattern as its first  function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
945  argument. If studing the pattern produces additional information that will help  argument. If studying the pattern produces additional information that will
946  speed up matching, \fBpcre_study()\fR returns a pointer to a \fBpcre_extra\fR  help speed up matching, \fBpcre_study()\fP returns a pointer to a
947  block, in which the \fIstudy_data\fR field points to the results of the study.  \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
948    results of the study.
949  The returned value from a \fBpcre_study()\fR can be passed directly to  .P
950  \fBpcre_exec()\fR. However, the \fBpcre_extra\fR block also contains other  The returned value from \fBpcre_study()\fP can be passed directly to
951  fields that can be set by the caller before the block is passed; these are  \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
952  described below. If studying the pattern does not produce any additional  also contains other fields that can be set by the caller before the block is
953  information, \fBpcre_study()\fR returns NULL. In that circumstance, if the  passed; these are described
954  calling program wants to pass some of the other fields to \fBpcre_exec()\fR, it  .\" HTML <a href="#extradata">
955  must set up its own \fBpcre_extra\fR block.  .\" </a>
956    below
957  The second argument contains option bits. At present, no options are defined  .\"
958  for \fBpcre_study()\fR, and this argument should always be zero.  in the section on matching a pattern.
959    .P
960  The third argument for \fBpcre_study()\fR is a pointer for an error message. If  If studying the pattern does not produce any useful information,
961    \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program
962    wants to pass any of the other fields to \fBpcre_exec()\fP or
963    \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block.
964    .P
965    The second argument of \fBpcre_study()\fP contains option bits. There are three
966    options:
967    .sp
968      PCRE_STUDY_JIT_COMPILE
969      PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
970      PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
971    .sp
972    If any of these are set, and the just-in-time compiler is available, the
973    pattern is further compiled into machine code that executes much faster than
974    the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
975    compiler is not available, these options are ignored. All other bits in the
976    \fIoptions\fP argument must be zero.
977    .P
978    JIT compilation is a heavyweight optimization. It can take some time for
979    patterns to be analyzed, and for one-off matches and simple patterns the
980    benefit of faster execution might be offset by a much slower study time.
981    Not all patterns can be optimized by the JIT compiler. For those that cannot be
982    handled, matching automatically falls back to the \fBpcre_exec()\fP
983    interpreter. For more details, see the
984    .\" HREF
985    \fBpcrejit\fP
986    .\"
987    documentation.
988    .P
989    The third argument for \fBpcre_study()\fP is a pointer for an error message. If
990  studying succeeds (even if no data is returned), the variable it points to is  studying succeeds (even if no data is returned), the variable it points to is
991  set to NULL. Otherwise it points to a textual error message. You should  set to NULL. Otherwise it is set to point to a textual error message. This is a
992  therefore test the error pointer for NULL after calling \fBpcre_study()\fR, to  static string that is part of the library. You must not try to free it. You
993  be sure that it has run successfully.  should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
994    sure that it has run successfully.
995  This is a typical call to \fBpcre_study\fR():  .P
996    When you are finished with a pattern, you can free the memory used for the
997    pcre_extra *pe;  study data by calling \fBpcre_free_study()\fP. This function was added to the
998    pe = pcre_study(  API for release 8.20. For earlier versions, the memory could be freed with
999    \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1000    where JIT optimization is not used, but it is advisable to change to the new
1001    function when convenient.
1002    .P
1003    This is a typical way in which \fBpcre_study\fP() is used (except that in a
1004    real application there should be tests for errors):
1005    .sp
1006      int rc;
1007      pcre *re;
1008      pcre_extra *sd;
1009      re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1010      sd = pcre_study(
1011      re,             /* result of pcre_compile() */      re,             /* result of pcre_compile() */
1012      0,              /* no options exist */      0,              /* no options */
1013      &error);        /* set to NULL or points to a message */      &error);        /* set to NULL or points to a message */
1014      rc = pcre_exec(   /* see below for details of pcre_exec() options */
1015  At present, studying a pattern is useful only for non-anchored patterns that do      re, sd, "subject", 7, 0, 0, ovector, 30);
1016  not have a single fixed starting character. A bitmap of possible starting    ...
1017  characters is created.    pcre_free_study(sd);
1018      pcre_free(re);
1019    .sp
1020    Studying a pattern does two things: first, a lower bound for the length of
1021    subject string that is needed to match the pattern is computed. This does not
1022    mean that there are any strings of that length that match, but it does
1023    guarantee that no shorter strings match. The value is used by
1024    \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP to avoid wasting time by trying to
1025    match strings that are shorter than the lower bound. You can find out the value
1026    in a calling program via the \fBpcre_fullinfo()\fP function.
1027    .P
1028    Studying a pattern is also useful for non-anchored patterns that do not have a
1029    single fixed starting character. A bitmap of possible starting bytes is
1030    created. This speeds up finding a position in the subject at which to start
1031    matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.)
1032    .P
1033    These two optimizations apply to both \fBpcre_exec()\fP and
1034    \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1035    The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option
1036    when calling \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP, but if this is done,
1037    JIT execution is also disabled. You might want to do this if your pattern
1038    contains callouts or (*MARK) and you want to make use of these facilities in
1039    cases where matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
1040    .\" HTML <a href="#execoptions">
1041    .\" </a>
1042    below.
1043    .\"
1044    .
1045    .
1046  .\" HTML <a name="localesupport"></a>  .\" HTML <a name="localesupport"></a>
1047  .SH LOCALE SUPPORT  .SH "LOCALE SUPPORT"
1048  .rs  .rs
1049  .sp  .sp
1050  PCRE handles caseless matching, and determines whether characters are letters,  PCRE handles caseless matching, and determines whether characters are letters,
1051  digits, or whatever, by reference to a set of tables. When running in UTF-8  digits, or whatever, by reference to a set of tables, indexed by character
1052  mode, this applies only to characters with codes less than 256. The library  value. When running in UTF-8 mode, this applies only to characters
1053  contains a default set of tables that is created in the default C locale when  with codes less than 128. By default, higher-valued codes never match escapes
1054  PCRE is compiled. This is used when the final argument of \fBpcre_compile()\fR  such as \ew or \ed, but they can be tested with \ep if PCRE is built with
1055  is NULL, and is sufficient for many applications.  Unicode character property support. Alternatively, the PCRE_UCP option can be
1056    set at compile time; this causes \ew and friends to use Unicode property
1057  An alternative set of tables can, however, be supplied. Such tables are built  support instead of built-in tables. The use of locales with Unicode is
1058  by calling the \fBpcre_maketables()\fR function, which has no arguments, in the  discouraged. If you are handling characters with codes greater than 128, you
1059  relevant locale. The result can then be passed to \fBpcre_compile()\fR as often  should either use UTF-8 and Unicode, or use locales, but not try to mix the
1060  as necessary. For example, to build and use tables that are appropriate for the  two.
1061  French locale (where accented characters with codes greater than 128 are  .P
1062  treated as letters), the following code could be used:  PCRE contains an internal set of tables that are used when the final argument
1063    of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1064    setlocale(LC_CTYPE, "fr");  Normally, the internal tables recognize only ASCII characters. However, when
1065    PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1066    default "C" locale of the local system, which may cause them to be different.
1067    .P
1068    The internal tables can always be overridden by tables supplied by the
1069    application that calls PCRE. These may be created in a different locale from
1070    the default. As more and more applications change to using Unicode, the need
1071    for this locale support is expected to die away.
1072    .P
1073    External tables are built by calling the \fBpcre_maketables()\fP function,
1074    which has no arguments, in the relevant locale. The result can then be passed
1075    to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
1076    example, to build and use tables that are appropriate for the French locale
1077    (where accented characters with values greater than 128 are treated as letters),
1078    the following code could be used:
1079    .sp
1080      setlocale(LC_CTYPE, "fr_FR");
1081    tables = pcre_maketables();    tables = pcre_maketables();
1082    re = pcre_compile(..., tables);    re = pcre_compile(..., tables);
1083    .sp
1084  The tables are built in memory that is obtained via \fBpcre_malloc\fR. The  The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1085  pointer that is passed to \fBpcre_compile\fR is saved with the compiled  are using Windows, the name for the French locale is "french".
1086  pattern, and the same tables are used via this pointer by \fBpcre_study()\fR  .P
1087  and \fBpcre_exec()\fR. Thus, for any single pattern, compilation, studying and  When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1088  matching all happen in the same locale, but different patterns can be compiled  obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1089  in different locales. It is the caller's responsibility to ensure that the  that the memory containing the tables remains available for as long as it is
1090  memory containing the tables remains available for as long as it is needed.  needed.
1091    .P
1092  .SH INFORMATION ABOUT A PATTERN  The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1093  .rs  pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1094  .sp  and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
1095  .B int pcre_fullinfo(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"  pattern, compilation, studying and matching all happen in the same locale, but
1096  .ti +5n  different patterns can be compiled in different locales.
1097  .B int \fIwhat\fR, void *\fIwhere\fR);  .P
1098  .PP  It is possible to pass a table pointer or NULL (indicating the use of the
1099  The \fBpcre_fullinfo()\fR function returns information about a compiled  internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
1100  pattern. It replaces the obsolete \fBpcre_info()\fR function, which is  this facility could be used to match a pattern in a different locale from the
1101  nevertheless retained for backwards compability (and is documented below).  one in which it was compiled. Passing table pointers at run time is discussed
1102    below in the section on matching a pattern.
1103  The first argument for \fBpcre_fullinfo()\fR is a pointer to the compiled  .
1104  pattern. The second argument is the result of \fBpcre_study()\fR, or NULL if  .
1105    .\" HTML <a name="infoaboutpattern"></a>
1106    .SH "INFORMATION ABOUT A PATTERN"
1107    .rs
1108    .sp
1109    .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1110    .ti +5n
1111    .B int \fIwhat\fP, void *\fIwhere\fP);
1112    .PP
1113    The \fBpcre_fullinfo()\fP function returns information about a compiled
1114    pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1115    library at version 8.30, after more than 10 years of obsolescence.
1116    .P
1117    The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1118    pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1119  the pattern was not studied. The third argument specifies which piece of  the pattern was not studied. The third argument specifies which piece of
1120  information is required, and the fourth argument is a pointer to a variable  information is required, and the fourth argument is a pointer to a variable
1121  to receive the data. The yield of the function is zero for success, or one of  to receive the data. The yield of the function is zero for success, or one of
1122  the following negative numbers:  the following negative numbers:
1123    .sp
1124    PCRE_ERROR_NULL       the argument \fIcode\fR was NULL    PCRE_ERROR_NULL           the argument \fIcode\fP was NULL
1125                          the argument \fIwhere\fR was NULL                              the argument \fIwhere\fP was NULL
1126    PCRE_ERROR_BADMAGIC   the "magic number" was not found    PCRE_ERROR_BADMAGIC       the "magic number" was not found
1127    PCRE_ERROR_BADOPTION  the value of \fIwhat\fR was invalid    PCRE_ERROR_BADENDIANNESS  the pattern was compiled with different
1128                                endianness
1129  Here is a typical call of \fBpcre_fullinfo()\fR, to obtain the length of the    PCRE_ERROR_BADOPTION      the value of \fIwhat\fP was invalid
1130  compiled pattern:  .sp
1131    The "magic number" is placed at the start of each compiled pattern as an simple
1132    check against passing an arbitrary memory pointer. The endianness error can
1133    occur if a compiled pattern is saved and reloaded on a different host. Here is
1134    a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1135    pattern:
1136    .sp
1137    int rc;    int rc;
1138    unsigned long int length;    size_t length;
1139    rc = pcre_fullinfo(    rc = pcre_fullinfo(
1140      re,               /* result of pcre_compile() */      re,               /* result of pcre_compile() */
1141      pe,               /* result of pcre_study(), or NULL */      sd,               /* result of pcre_study(), or NULL */
1142      PCRE_INFO_SIZE,   /* what is required */      PCRE_INFO_SIZE,   /* what is required */
1143      &length);         /* where to put the data */      &length);         /* where to put the data */
1144    .sp
1145  The possible values for the third argument are defined in \fBpcre.h\fR, and are  The possible values for the third argument are defined in \fBpcre.h\fP, and are
1146  as follows:  as follows:
1147    .sp
1148    PCRE_INFO_BACKREFMAX    PCRE_INFO_BACKREFMAX
1149    .sp
1150  Return the number of the highest back reference in the pattern. The fourth  Return the number of the highest back reference in the pattern. The fourth
1151  argument should point to an \fBint\fR variable. Zero is returned if there are  argument should point to an \fBint\fP variable. Zero is returned if there are
1152  no back references.  no back references.
1153    .sp
1154    PCRE_INFO_CAPTURECOUNT    PCRE_INFO_CAPTURECOUNT
1155    .sp
1156  Return the number of capturing subpatterns in the pattern. The fourth argument  Return the number of capturing subpatterns in the pattern. The fourth argument
1157  should point to an \fbint\fR variable.  should point to an \fBint\fP variable.
1158    .sp
1159      PCRE_INFO_DEFAULT_TABLES
1160    .sp
1161    Return a pointer to the internal default character tables within PCRE. The
1162    fourth argument should point to an \fBunsigned char *\fP variable. This
1163    information call is provided for internal use by the \fBpcre_study()\fP
1164    function. External callers can cause PCRE to use its internal tables by passing
1165    a NULL table pointer.
1166    .sp
1167    PCRE_INFO_FIRSTBYTE    PCRE_INFO_FIRSTBYTE
1168    .sp
1169  Return information about the first byte of any matched string, for a  Return information about the first data unit of any matched string, for a
1170  non-anchored pattern. (This option used to be called PCRE_INFO_FIRSTCHAR; the  non-anchored pattern. (The name of this option refers to the 8-bit library,
1171  old name is still recognized for backwards compatibility.)  where data units are bytes.) The fourth argument should point to an \fBint\fP
1172    variable.
1173  If there is a fixed first byte, e.g. from a pattern such as (cat|cow|coyote),  .P
1174  it is returned in the integer pointed to by \fIwhere\fR. Otherwise, if either  If there is a fixed first value, for example, the letter "c" from a pattern
1175    such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1176    value is always less than 256; in the 16-bit library the value can be up to
1177    0xffff.
1178    .P
1179    If there is no fixed first value, and if either
1180    .sp
1181  (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch  (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1182  starts with "^", or  starts with "^", or
1183    .sp
1184  (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set  (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1185  (if it were set, the pattern would be anchored),  (if it were set, the pattern would be anchored),
1186    .sp
1187  -1 is returned, indicating that the pattern matches only at the start of a  -1 is returned, indicating that the pattern matches only at the start of a
1188  subject string or after any newline within the string. Otherwise -2 is  subject string or after any newline within the string. Otherwise -2 is
1189  returned. For anchored patterns, -2 is returned.  returned. For anchored patterns, -2 is returned.
1190    .sp
1191    PCRE_INFO_FIRSTTABLE    PCRE_INFO_FIRSTTABLE
1192    .sp
1193  If the pattern was studied, and this resulted in the construction of a 256-bit  If the pattern was studied, and this resulted in the construction of a 256-bit
1194  table indicating a fixed set of bytes for the first byte in any matching  table indicating a fixed set of values for the first data unit in any matching
1195  string, a pointer to the table is returned. Otherwise NULL is returned. The  string, a pointer to the table is returned. Otherwise NULL is returned. The
1196  fourth argument should point to an \fBunsigned char *\fR variable.  fourth argument should point to an \fBunsigned char *\fP variable.
1197    .sp
1198      PCRE_INFO_HASCRORLF
1199    .sp
1200    Return 1 if the pattern contains any explicit matches for CR or LF characters,
1201    otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1202    explicit match is either a literal CR or LF character, or \er or \en.
1203    .sp
1204      PCRE_INFO_JCHANGED
1205    .sp
1206    Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1207    0. The fourth argument should point to an \fBint\fP variable. (?J) and
1208    (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1209    .sp
1210      PCRE_INFO_JIT
1211    .sp
1212    Return 1 if the pattern was studied with one of the JIT options, and
1213    just-in-time compiling was successful. The fourth argument should point to an
1214    \fBint\fP variable. A return value of 0 means that JIT support is not available
1215    in this version of PCRE, or that the pattern was not studied with a JIT option,
1216    or that the JIT compiler could not handle this particular pattern. See the
1217    .\" HREF
1218    \fBpcrejit\fP
1219    .\"
1220    documentation for details of what can and cannot be handled.
1221    .sp
1222      PCRE_INFO_JITSIZE
1223    .sp
1224    If the pattern was successfully studied with a JIT option, return the size of
1225    the JIT compiled code, otherwise return zero. The fourth argument should point
1226    to a \fBsize_t\fP variable.
1227    .sp
1228    PCRE_INFO_LASTLITERAL    PCRE_INFO_LASTLITERAL
1229    .sp
1230  Return the value of the rightmost literal byte that must exist in any matched  Return the value of the rightmost literal data unit that must exist in any
1231  string, other than at its start, if such a byte has been recorded. The fourth  matched string, other than at its start, if such a value has been recorded. The
1232  argument should point to an \fBint\fR variable. If there is no such byte, -1 is  fourth argument should point to an \fBint\fP variable. If there is no such
1233  returned. For anchored patterns, a last literal byte is recorded only if it  value, -1 is returned. For anchored patterns, a last literal value is recorded
1234  follows something of variable length. For example, for the pattern  only if it follows something of variable length. For example, for the pattern
1235  /^a\\d+z\\d+/ the returned value is "z", but for /^a\\dz\\d/ the returned value  /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1236  is -1.  is -1.
1237    .sp
1238      PCRE_INFO_MAXLOOKBEHIND
1239    .sp
1240    Return the number of characters (NB not bytes) in the longest lookbehind
1241    assertion in the pattern. Note that the simple assertions \eb and \eB require a
1242    one-character lookbehind. This information is useful when doing multi-segment
1243    matching using the partial matching facilities.
1244    .sp
1245      PCRE_INFO_MINLENGTH
1246    .sp
1247    If the pattern was studied and a minimum length for matching subject strings
1248    was computed, its value is returned. Otherwise the returned value is -1. The
1249    value is a number of characters, which in UTF-8 mode may be different from the
1250    number of bytes. The fourth argument should point to an \fBint\fP variable. A
1251    non-negative value is a lower bound to the length of any matching string. There
1252    may not be any strings of that length that do actually match, but every string
1253    that does match is at least that long.
1254    .sp
1255    PCRE_INFO_NAMECOUNT    PCRE_INFO_NAMECOUNT
1256    PCRE_INFO_NAMEENTRYSIZE    PCRE_INFO_NAMEENTRYSIZE
1257    PCRE_INFO_NAMETABLE    PCRE_INFO_NAMETABLE
1258    .sp
1259  PCRE supports the use of named as well as numbered capturing parentheses. The  PCRE supports the use of named as well as numbered capturing parentheses. The
1260  names are just an additional way of identifying the parentheses, which still  names are just an additional way of identifying the parentheses, which still
1261  acquire a number. A caller that wants to extract data from a named subpattern  acquire numbers. Several convenience functions such as
1262  must convert the name to a number in order to access the correct pointers in  \fBpcre_get_named_substring()\fP are provided for extracting captured
1263  the output vector (described with \fBpcre_exec()\fR below). In order to do  substrings by name. It is also possible to extract the data directly, by first
1264  this, it must first use these three values to obtain the name-to-number mapping  converting the name to a number in order to access the correct pointers in the
1265  table for the pattern.  output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1266    you need to use the name-to-number map, which is described by these three
1267    values.
1268    .P
1269  The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives  The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1270  the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each  the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1271  entry; both of these return an \fBint\fR value. The entry size depends on the  entry; both of these return an \fBint\fP value. The entry size depends on the
1272  length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first  length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1273  entry of the table (a pointer to \fBchar\fR). The first two bytes of each entry  entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1274  are the number of the capturing parenthesis, most significant byte first. The  the first two bytes of each entry are the number of the capturing parenthesis,
1275  rest of the entry is the corresponding name, zero terminated. The names are in  most significant byte first. In the 16-bit library, the pointer points to
1276  alphabetical order. For example, consider the following pattern (assume  16-bit data units, the first of which contains the parenthesis number. The rest
1277  PCRE_EXTENDED is set, so white space - including newlines - is ignored):  of the entry is the corresponding name, zero terminated.
1278    .P
1279    (?P<date> (?P<year>(\\d\\d)?\\d\\d) -  The names are in alphabetical order. Duplicate names may appear if (?| is used
1280    (?P<month>\\d\\d) - (?P<day>\\d\\d) )  to create multiple groups with the same number, as described in the
1281    .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1282    .\" </a>
1283    section on duplicate subpattern numbers
1284    .\"
1285    in the
1286    .\" HREF
1287    \fBpcrepattern\fP
1288    .\"
1289    page. Duplicate names for subpatterns with different numbers are permitted only
1290    if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1291    table in the order in which they were found in the pattern. In the absence of
1292    (?| this is the order of increasing number; when (?| is used this is not
1293    necessarily the case because later subpatterns may have lower numbers.
1294    .P
1295    As a simple example of the name/number table, consider the following pattern
1296    after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1297    space - including newlines - is ignored):
1298    .sp
1299    .\" JOIN
1300      (?<date> (?<year>(\ed\ed)?\ed\ed) -
1301      (?<month>\ed\ed) - (?<day>\ed\ed) )
1302    .sp
1303  There are four named subpatterns, so the table has four entries, and each entry  There are four named subpatterns, so the table has four entries, and each entry
1304  in the table is eight bytes long. The table is as follows, with non-printing  in the table is eight bytes long. The table is as follows, with non-printing
1305  bytes shows in hex, and undefined bytes shown as ??:  bytes shows in hexadecimal, and undefined bytes shown as ??:
1306    .sp
1307    00 01 d  a  t  e  00 ??    00 01 d  a  t  e  00 ??
1308    00 05 d  a  y  00 ?? ??    00 05 d  a  y  00 ?? ??
1309    00 04 m  o  n  t  h  00    00 04 m  o  n  t  h  00
1310    00 02 y  e  a  r  00 ??    00 02 y  e  a  r  00 ??
1311    .sp
1312  When writing code to extract data from named subpatterns, remember that the  When writing code to extract data from named subpatterns using the
1313  length of each entry may be different for each compiled pattern.  name-to-number map, remember that the length of the entries is likely to be
1314    different for each compiled pattern.
1315    .sp
1316      PCRE_INFO_OKPARTIAL
1317    .sp
1318    Return 1 if the pattern can be used for partial matching with
1319    \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1320    \fBint\fP variable. From release 8.00, this always returns 1, because the
1321    restrictions that previously applied to partial matching have been lifted. The
1322    .\" HREF
1323    \fBpcrepartial\fP
1324    .\"
1325    documentation gives details of partial matching.
1326    .sp
1327    PCRE_INFO_OPTIONS    PCRE_INFO_OPTIONS
1328    .sp
1329  Return a copy of the options with which the pattern was compiled. The fourth  Return a copy of the options with which the pattern was compiled. The fourth
1330  argument should point to an \fBunsigned long int\fR variable. These option bits  argument should point to an \fBunsigned long int\fP variable. These option bits
1331  are those specified in the call to \fBpcre_compile()\fR, modified by any  are those specified in the call to \fBpcre_compile()\fP, modified by any
1332  top-level option settings within the pattern itself.  top-level option settings at the start of the pattern itself. In other words,
1333    they are the options that will be in force when matching starts. For example,
1334    if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1335    result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1336    .P
1337  A pattern is automatically anchored by PCRE if all of its top-level  A pattern is automatically anchored by PCRE if all of its top-level
1338  alternatives begin with one of the following:  alternatives begin with one of the following:
1339    .sp
1340    ^     unless PCRE_MULTILINE is set    ^     unless PCRE_MULTILINE is set
1341    \\A    always    \eA    always
1342    \\G    always    \eG    always
1343    .\" JOIN
1344    .*    if PCRE_DOTALL is set and there are no back    .*    if PCRE_DOTALL is set and there are no back
1345            references to the subpattern in which .* appears            references to the subpattern in which .* appears
1346    .sp
1347  For such patterns, the PCRE_ANCHORED bit is set in the options returned by  For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1348  \fBpcre_fullinfo()\fR.  \fBpcre_fullinfo()\fP.
1349    .sp
1350    PCRE_INFO_SIZE    PCRE_INFO_SIZE
1351    .sp
1352  Return the size of the compiled pattern, that is, the value that was passed as  Return the size of the compiled pattern in bytes (for both libraries). The
1353  the argument to \fBpcre_malloc()\fR when PCRE was getting memory in which to  fourth argument should point to a \fBsize_t\fP variable. This value does not
1354  place the compiled data. The fourth argument should point to a \fBsize_t\fR  include the size of the \fBpcre\fP structure that is returned by
1355  variable.  \fBpcre_compile()\fP. The value that is passed as the argument to
1356    \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1357    place the compiled data is the value returned by this option plus the size of
1358    the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1359    does not alter the value returned by this option.
1360    .sp
1361    PCRE_INFO_STUDYSIZE    PCRE_INFO_STUDYSIZE
1362    .sp
1363  Returns the size of the data block pointed to by the \fIstudy_data\fR field in  Return the size in bytes of the data block pointed to by the \fIstudy_data\fP
1364  a \fBpcre_extra\fR block. That is, it is the value that was passed to  field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP is NULL, or there is no
1365  \fBpcre_malloc()\fR when PCRE was getting memory into which to place the data  study data, zero is returned. The fourth argument should point to a
1366  created by \fBpcre_study()\fR. The fourth argument should point to a  \fBsize_t\fP variable. The \fIstudy_data\fP field is set by \fBpcre_study()\fP
1367  \fBsize_t\fR variable.  to record information that will speed up matching (see the section entitled
1368    .\" HTML <a href="#studyingapattern">
1369  .SH OBSOLETE INFO FUNCTION  .\" </a>
1370  .rs  "Studying a pattern"
1371  .sp  .\"
1372  .B int pcre_info(const pcre *\fIcode\fR, int *\fIoptptr\fR, int  above). The format of the \fIstudy_data\fP block is private, but its length
1373  .B *\fIfirstcharptr\fR);  is made available via this option so that it can be saved and restored (see the
1374  .PP  .\" HREF
1375  The \fBpcre_info()\fR function is now obsolete because its interface is too  \fBpcreprecompile\fP
1376  restrictive to return all the available data about a compiled pattern. New  .\"
1377  programs should use \fBpcre_fullinfo()\fR instead. The yield of  documentation for details).
1378  \fBpcre_info()\fR is the number of capturing subpatterns, or one of the  .
1379  following negative numbers:  .
1380    .SH "REFERENCE COUNTS"
1381    PCRE_ERROR_NULL       the argument \fIcode\fR was NULL  .rs
1382    PCRE_ERROR_BADMAGIC   the "magic number" was not found  .sp
1383    .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1384  If the \fIoptptr\fR argument is not NULL, a copy of the options with which the  .PP
1385  pattern was compiled is placed in the integer it points to (see  The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1386  PCRE_INFO_OPTIONS above).  data block that contains a compiled pattern. It is provided for the benefit of
1387    applications that operate in an object-oriented manner, where different parts
1388  If the pattern is not anchored and the \fIfirstcharptr\fR argument is not NULL,  of the application may be using the same compiled pattern, but you want to free
1389  it is used to pass back information about the first character of any matched  the block when they are all done.
1390  string (see PCRE_INFO_FIRSTBYTE above).  .P
1391    When a pattern is compiled, the reference count field is initialized to zero.
1392  .SH MATCHING A PATTERN  It is changed only by calling this function, whose action is to add the
1393  .rs  \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1394  .sp  function is the new value. However, the value of the count is constrained to
1395  .B int pcre_exec(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"  lie between 0 and 65535, inclusive. If the new value is outside these limits,
1396  .ti +5n  it is forced to the appropriate limit value.
1397  .B "const char *\fIsubject\fR," int \fIlength\fR, int \fIstartoffset\fR,  .P
1398  .ti +5n  Except when it is zero, the reference count is not correctly preserved if a
1399  .B int \fIoptions\fR, int *\fIovector\fR, int \fIovecsize\fR);  pattern is compiled on one host and then transferred to a host whose byte-order
1400  .PP  is different. (This seems a highly unlikely scenario.)
1401  The function \fBpcre_exec()\fR is called to match a subject string against a  .
1402  pre-compiled pattern, which is passed in the \fIcode\fR argument. If the  .
1403  pattern has been studied, the result of the study should be passed in the  .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1404  \fIextra\fR argument.  .rs
1405    .sp
1406  Here is an example of a simple call to \fBpcre_exec()\fR:  .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1407    .ti +5n
1408    .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1409    .ti +5n
1410    .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1411    .P
1412    The function \fBpcre_exec()\fP is called to match a subject string against a
1413    compiled pattern, which is passed in the \fIcode\fP argument. If the
1414    pattern was studied, the result of the study should be passed in the
1415    \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1416    and \fIextra\fP arguments as many times as you like, in order to match
1417    different subject strings with the same pattern.
1418    .P
1419    This function is the main matching facility of the library, and it operates in
1420    a Perl-like manner. For specialist use there is also an alternative matching
1421    function, which is described
1422    .\" HTML <a href="#dfamatch">
1423    .\" </a>
1424    below
1425    .\"
1426    in the section about the \fBpcre_dfa_exec()\fP function.
1427    .P
1428    In most applications, the pattern will have been compiled (and optionally
1429    studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1430    possible to save compiled patterns and study data, and then use them later
1431    in different processes, possibly even on different hosts. For a discussion
1432    about this, see the
1433    .\" HREF
1434    \fBpcreprecompile\fP
1435    .\"
1436    documentation.
1437    .P
1438    Here is an example of a simple call to \fBpcre_exec()\fP:
1439    .sp
1440    int rc;    int rc;
1441    int ovector[30];    int ovector[30];
1442    rc = pcre_exec(    rc = pcre_exec(
# Line 647  Here is an example of a simple call to \ Line 1446  Here is an example of a simple call to \
1446      11,             /* the length of the subject string */      11,             /* the length of the subject string */
1447      0,              /* start at offset 0 in the subject */      0,              /* start at offset 0 in the subject */
1448      0,              /* default options */      0,              /* default options */
1449      ovector,        /* vector for substring information */      ovector,        /* vector of integers for substring information */
1450      30);            /* number of elements in the vector */      30);            /* number of elements (NOT size in bytes) */
1451    .
1452  If the \fIextra\fR argument is not NULL, it must point to a \fBpcre_extra\fR  .
1453  data block. The \fBpcre_study()\fR function returns such a block (when it  .\" HTML <a name="extradata"></a>
1454    .SS "Extra data for \fBpcre_exec()\fR"
1455    .rs
1456    .sp
1457    If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1458    data block. The \fBpcre_study()\fP function returns such a block (when it
1459  doesn't return NULL), but you can also create one for yourself, and pass  doesn't return NULL), but you can also create one for yourself, and pass
1460  additional information in it. The fields in the block are as follows:  additional information in it. The \fBpcre_extra\fP block contains the following
1461    fields (not necessarily in this order):
1462    unsigned long int \fIflags\fR;  .sp
1463    void *\fIstudy_data\fR;    unsigned long int \fIflags\fP;
1464    unsigned long int \fImatch_limit\fR;    void *\fIstudy_data\fP;
1465    void *\fIcallout_data\fR;    void *\fIexecutable_jit\fP;
1466      unsigned long int \fImatch_limit\fP;
1467  The \fIflags\fR field is a bitmap that specifies which of the other fields    unsigned long int \fImatch_limit_recursion\fP;
1468  are set. The flag bits are:    void *\fIcallout_data\fP;
1469      const unsigned char *\fItables\fP;
1470    PCRE_EXTRA_STUDY_DATA    unsigned char **\fImark\fP;
1471    PCRE_EXTRA_MATCH_LIMIT  .sp
1472    In the 16-bit version of this structure, the \fImark\fP field has type
1473    "PCRE_UCHAR16 **".
1474    .P
1475    The \fIflags\fP field is used to specify which of the other fields are set. The
1476    flag bits are:
1477    .sp
1478    PCRE_EXTRA_CALLOUT_DATA    PCRE_EXTRA_CALLOUT_DATA
1479      PCRE_EXTRA_EXECUTABLE_JIT
1480  Other flag bits should be set to zero. The \fIstudy_data\fR field is set in the    PCRE_EXTRA_MARK
1481  \fBpcre_extra\fR block that is returned by \fBpcre_study()\fR, together with    PCRE_EXTRA_MATCH_LIMIT
1482  the appropriate flag bit. You should not set this yourself, but you can add to    PCRE_EXTRA_MATCH_LIMIT_RECURSION
1483  the block by setting the other fields.    PCRE_EXTRA_STUDY_DATA
1484      PCRE_EXTRA_TABLES
1485  The \fImatch_limit\fR field provides a means of preventing PCRE from using up a  .sp
1486    Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1487    the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1488    returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1489    should not set these yourself, but you may add to the block by setting other
1490    fields and their corresponding flag bits.
1491    .P
1492    The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1493  vast amount of resources when running patterns that are not going to match,  vast amount of resources when running patterns that are not going to match,
1494  but which have a very large number of possibilities in their search trees. The  but which have a very large number of possibilities in their search trees. The
1495  classic example is the use of nested unlimited repeats. Internally, PCRE uses a  classic example is a pattern that uses nested unlimited repeats.
1496  function called \fBmatch()\fR which it calls repeatedly (sometimes  .P
1497  recursively). The limit is imposed on the number of times this function is  Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1498  called during a match, which has the effect of limiting the amount of recursion  calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1499  and backtracking that can take place. For patterns that are not anchored, the  imposed on the number of times this function is called during a match, which
1500  count starts from zero for each position in the subject string.  has the effect of limiting the amount of backtracking that can take place. For
1501    patterns that are not anchored, the count restarts from zero for each position
1502  The default limit for the library can be set when PCRE is built; the default  in the subject string.
1503    .P
1504    When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1505    with a JIT option, the way that the matching is executed is entirely different.
1506    However, there is still the possibility of runaway matching that goes on for a
1507    very long time, and so the \fImatch_limit\fP value is also used in this case
1508    (but in a different way) to limit how long the matching can continue.
1509    .P
1510    The default value for the limit can be set when PCRE is built; the default
1511  default is 10 million, which handles all but the most extreme cases. You can  default is 10 million, which handles all but the most extreme cases. You can
1512  reduce the default by suppling \fBpcre_exec()\fR with a \fRpcre_extra\fR block  override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1513  in which \fImatch_limit\fR is set to a smaller value, and  block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1514  PCRE_EXTRA_MATCH_LIMIT is set in the \fIflags\fR field. If the limit is  the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1515  exceeded, \fBpcre_exec()\fR returns PCRE_ERROR_MATCHLIMIT.  PCRE_ERROR_MATCHLIMIT.
1516    .P
1517  The \fIpcre_callout\fR field is used in conjunction with the "callout" feature,  The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1518  which is described in the \fBpcrecallout\fR documentation.  instead of limiting the total number of times that \fBmatch()\fP is called, it
1519    limits the depth of recursion. The recursion depth is a smaller number than the
1520  The PCRE_ANCHORED option can be passed in the \fIoptions\fR argument, whose  total number of calls, because not all calls to \fBmatch()\fP are recursive.
1521  unused bits must be zero. This limits \fBpcre_exec()\fR to matching at the  This limit is of use only if it is set smaller than \fImatch_limit\fP.
1522  first matching position. However, if a pattern was compiled with PCRE_ANCHORED,  .P
1523  or turned out to be anchored by virtue of its contents, it cannot be made  Limiting the recursion depth limits the amount of machine stack that can be
1524  unachored at matching time.  used, or, when PCRE has been compiled to use memory on the heap instead of the
1525    stack, the amount of heap memory that can be used. This limit is not relevant,
1526  There are also three further options that can be set only at matching time:  and is ignored, when matching is done using JIT compiled code.
1527    .P
1528    The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1529    built; the default default is the same value as the default for
1530    \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1531    with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1532    PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1533    is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1534    .P
1535    The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1536    and is described in the
1537    .\" HREF
1538    \fBpcrecallout\fP
1539    .\"
1540    documentation.
1541    .P
1542    The \fItables\fP field is used to pass a character tables pointer to
1543    \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1544    pattern. A non-NULL value is stored with the compiled pattern only if custom
1545    tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1546    If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1547    internal tables to be used. This facility is helpful when re-using patterns
1548    that have been saved after compiling with an external set of tables, because
1549    the external tables might be at a different address when \fBpcre_exec()\fP is
1550    called. See the
1551    .\" HREF
1552    \fBpcreprecompile\fP
1553    .\"
1554    documentation for a discussion of saving compiled patterns for later use.
1555    .P
1556    If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1557    be set to point to a suitable variable. If the pattern contains any
1558    backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1559    a name to pass back, a pointer to the name string (zero terminated) is placed
1560    in the variable pointed to by the \fImark\fP field. The names are within the
1561    compiled pattern; if you wish to retain such a name you must copy it before
1562    freeing the memory of a compiled pattern. If there is no name to pass back, the
1563    variable pointed to by the \fImark\fP field is set to NULL. For details of the
1564    backtracking control verbs, see the section entitled
1565    .\" HTML <a href="pcrepattern#backtrackcontrol">
1566    .\" </a>
1567    "Backtracking control"
1568    .\"
1569    in the
1570    .\" HREF
1571    \fBpcrepattern\fP
1572    .\"
1573    documentation.
1574    .
1575    .
1576    .\" HTML <a name="execoptions"></a>
1577    .SS "Option bits for \fBpcre_exec()\fP"
1578    .rs
1579    .sp
1580    The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1581    zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1582    PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1583    PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1584    PCRE_PARTIAL_SOFT.
1585    .P
1586    If the pattern was successfully studied with one of the just-in-time (JIT)
1587    compile options, the only supported options for JIT execution are
1588    PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1589    PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1590    unsupported option is used, JIT execution is disabled and the normal
1591    interpretive code in \fBpcre_exec()\fP is run.
1592    .sp
1593      PCRE_ANCHORED
1594    .sp
1595    The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1596    matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1597    to be anchored by virtue of its contents, it cannot be made unachored at
1598    matching time.
1599    .sp
1600      PCRE_BSR_ANYCRLF
1601      PCRE_BSR_UNICODE
1602    .sp
1603    These options (which are mutually exclusive) control what the \eR escape
1604    sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1605    match any Unicode newline sequence. These options override the choice that was
1606    made or defaulted when the pattern was compiled.
1607    .sp
1608      PCRE_NEWLINE_CR
1609      PCRE_NEWLINE_LF
1610      PCRE_NEWLINE_CRLF
1611      PCRE_NEWLINE_ANYCRLF
1612      PCRE_NEWLINE_ANY
1613    .sp
1614    These options override the newline definition that was chosen or defaulted when
1615    the pattern was compiled. For details, see the description of
1616    \fBpcre_compile()\fP above. During matching, the newline choice affects the
1617    behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1618    the way the match position is advanced after a match failure for an unanchored
1619    pattern.
1620    .P
1621    When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1622    match attempt for an unanchored pattern fails when the current position is at a
1623    CRLF sequence, and the pattern contains no explicit matches for CR or LF
1624    characters, the match position is advanced by two characters instead of one, in
1625    other words, to after the CRLF.
1626    .P
1627    The above rule is a compromise that makes the most common cases work as
1628    expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1629    set), it does not match the string "\er\enA" because, after failing at the
1630    start, it skips both the CR and the LF before retrying. However, the pattern
1631    [\er\en]A does match that string, because it contains an explicit CR or LF
1632    reference, and so advances only by one character after the first failure.
1633    .P
1634    An explicit match for CR of LF is either a literal appearance of one of those
1635    characters, or one of the \er or \en escape sequences. Implicit matches such as
1636    [^X] do not count, nor does \es (which includes CR and LF in the characters
1637    that it matches).
1638    .P
1639    Notwithstanding the above, anomalous effects may still occur when CRLF is a
1640    valid newline sequence and explicit \er or \en escapes appear in the pattern.
1641    .sp
1642    PCRE_NOTBOL    PCRE_NOTBOL
1643    .sp
1644  The first character of the string is not the beginning of a line, so the  This option specifies that first character of the subject string is not the
1645  circumflex metacharacter should not match before it. Setting this without  beginning of a line, so the circumflex metacharacter should not match before
1646  PCRE_MULTILINE (at compile time) causes circumflex never to match.  it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1647    never to match. This option affects only the behaviour of the circumflex
1648    metacharacter. It does not affect \eA.
1649    .sp
1650    PCRE_NOTEOL    PCRE_NOTEOL
1651    .sp
1652  The end of the string is not the end of a line, so the dollar metacharacter  This option specifies that the end of the subject string is not the end of a
1653  should not match it nor (except in multiline mode) a newline immediately before  line, so the dollar metacharacter should not match it nor (except in multiline
1654  it. Setting this without PCRE_MULTILINE (at compile time) causes dollar never  mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1655  to match.  compile time) causes dollar never to match. This option affects only the
1656    behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1657    .sp
1658    PCRE_NOTEMPTY    PCRE_NOTEMPTY
1659    .sp
1660  An empty string is not considered to be a valid match if this option is set. If  An empty string is not considered to be a valid match if this option is set. If
1661  there are alternatives in the pattern, they are tried. If all the alternatives  there are alternatives in the pattern, they are tried. If all the alternatives
1662  match the empty string, the entire match fails. For example, if the pattern  match the empty string, the entire match fails. For example, if the pattern
1663    .sp
1664    a?b?    a?b?
1665    .sp
1666  is applied to a string not beginning with "a" or "b", it matches the empty  is applied to a string not beginning with "a" or "b", it matches an empty
1667  string at the start of the subject. With PCRE_NOTEMPTY set, this match is not  string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1668  valid, so PCRE searches further into the string for occurrences of "a" or "b".  valid, so PCRE searches further into the string for occurrences of "a" or "b".
1669    .sp
1670  Perl has no direct equivalent of PCRE_NOTEMPTY, but it does make a special case    PCRE_NOTEMPTY_ATSTART
1671  of a pattern match of the empty string within its \fBsplit()\fR function, and  .sp
1672  when using the /g modifier. It is possible to emulate Perl's behaviour after  This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1673  matching a null string by first trying the match again at the same offset with  the start of the subject is permitted. If the pattern is anchored, such a match
1674  PCRE_NOTEMPTY set, and then if that fails by advancing the starting offset (see  can occur only if the pattern contains \eK.
1675  below) and trying an ordinary match again.  .P
1676    Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1677  The subject string is passed to \fBpcre_exec()\fR as a pointer in  does make a special case of a pattern match of the empty string within its
1678  \fIsubject\fR, a length in \fIlength\fR, and a starting offset in  \fBsplit()\fP function, and when using the /g modifier. It is possible to
1679  \fIstartoffset\fR. Unlike the pattern string, the subject may contain binary  emulate Perl's behaviour after matching a null string by first trying the match
1680  zero bytes. When the starting offset is zero, the search for a match starts at  again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1681  the beginning of the subject, and this is by far the most common case.  if that fails, by advancing the starting offset (see below) and trying an
1682    ordinary match again. There is some code that demonstrates how to do this in
1683  If the pattern was compiled with the PCRE_UTF8 option, the subject must be a  the
1684  sequence of bytes that is a valid UTF-8 string. If an invalid UTF-8 string is  .\" HREF
1685  passed, PCRE's behaviour is not defined.  \fBpcredemo\fP
1686    .\"
1687    sample program. In the most general case, you have to check to see if the
1688    newline convention recognizes CRLF as a newline, and if so, and the current
1689    character is CR followed by LF, advance the starting offset by two characters
1690    instead of one.
1691    .sp
1692      PCRE_NO_START_OPTIMIZE
1693    .sp
1694    There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1695    a match, in order to speed up the process. For example, if it is known that an
1696    unanchored match must start with a specific character, it searches the subject
1697    for that character, and fails immediately if it cannot find it, without
1698    actually running the main matching function. This means that a special item
1699    such as (*COMMIT) at the start of a pattern is not considered until after a
1700    suitable starting point for the match has been found. When callouts or (*MARK)
1701    items are in use, these "start-up" optimizations can cause them to be skipped
1702    if the pattern is never actually used. The start-up optimizations are in effect
1703    a pre-scan of the subject that takes place before the pattern is run.
1704    .P
1705    The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1706    causing performance to suffer, but ensuring that in cases where the result is
1707    "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1708    are considered at every possible starting position in the subject string. If
1709    PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1710    time. The use of PCRE_NO_START_OPTIMIZE disables JIT execution; when it is set,
1711    matching is always done using interpretively.
1712    .P
1713    Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1714    Consider the pattern
1715    .sp
1716      (*COMMIT)ABC
1717    .sp
1718    When this is compiled, PCRE records the fact that a match must start with the
1719    character "A". Suppose the subject string is "DEFABC". The start-up
1720    optimization scans along the subject, finds "A" and runs the first match
1721    attempt from there. The (*COMMIT) item means that the pattern must match the
1722    current starting position, which in this case, it does. However, if the same
1723    match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1724    subject string does not happen. The first match attempt is run starting from
1725    "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1726    the overall result is "no match". If the pattern is studied, more start-up
1727    optimizations may be used. For example, a minimum length for the subject may be
1728    recorded. Consider the pattern
1729    .sp
1730      (*MARK:A)(X|Y)
1731    .sp
1732    The minimum length for a match is one character. If the subject is "ABC", there
1733    will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1734    If the pattern is studied, the final attempt does not take place, because PCRE
1735    knows that the subject is too short, and so the (*MARK) is never encountered.
1736    In this case, studying the pattern does not affect the overall match result,
1737    which is still "no match", but it does affect the auxiliary information that is
1738    returned.
1739    .sp
1740      PCRE_NO_UTF8_CHECK
1741    .sp
1742    When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1743    string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1744    The value of \fIstartoffset\fP is also checked to ensure that it points to the
1745    start of a UTF-8 character. There is a discussion about the validity of UTF-8
1746    strings in the
1747    .\" HREF
1748    \fBpcreunicode\fP
1749    .\"
1750    page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1751    error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1752    truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1753    cases, information about the precise nature of the error may also be returned
1754    (see the descriptions of these errors in the section entitled \fIError return
1755    values from\fP \fBpcre_exec()\fP
1756    .\" HTML <a href="#errorlist">
1757    .\" </a>
1758    below).
1759    .\"
1760    If \fIstartoffset\fP contains a value that does not point to the start of a
1761    UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1762    returned.
1763    .P
1764    If you already know that your subject is valid, and you want to skip these
1765    checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1766    calling \fBpcre_exec()\fP. You might want to do this for the second and
1767    subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1768    all the matches in a single subject string. However, you should be sure that
1769    the value of \fIstartoffset\fP points to the start of a character (or the end
1770    of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1771    invalid string as a subject or an invalid value of \fIstartoffset\fP is
1772    undefined. Your program may crash.
1773    .sp
1774      PCRE_PARTIAL_HARD
1775      PCRE_PARTIAL_SOFT
1776    .sp
1777    These options turn on the partial matching feature. For backwards
1778    compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1779    occurs if the end of the subject string is reached successfully, but there are
1780    not enough subject characters to complete the match. If this happens when
1781    PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1782    testing any remaining alternatives. Only if no complete match can be found is
1783    PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1784    PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1785    but only if no complete match can be found.
1786    .P
1787    If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1788    partial match is found, \fBpcre_exec()\fP immediately returns
1789    PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1790    when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1791    important that an alternative complete match.
1792    .P
1793    In both cases, the portion of the string that was inspected when the partial
1794    match was found is set as the first matching string. There is a more detailed
1795    discussion of partial and multi-segment matching, with examples, in the
1796    .\" HREF
1797    \fBpcrepartial\fP
1798    .\"
1799    documentation.
1800    .
1801    .
1802    .SS "The string to be matched by \fBpcre_exec()\fP"
1803    .rs
1804    .sp
1805    The subject string is passed to \fBpcre_exec()\fP as a pointer in
1806    \fIsubject\fP, a length in bytes in \fIlength\fP, and a starting byte offset
1807    in \fIstartoffset\fP. If this is negative or greater than the length of the
1808    subject, \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting
1809    offset is zero, the search for a match starts at the beginning of the subject,
1810    and this is by far the most common case. In UTF-8 mode, the byte offset must
1811    point to the start of a UTF-8 character (or the end of the subject). Unlike the
1812    pattern string, the subject may contain binary zero bytes.
1813    .P
1814  A non-zero starting offset is useful when searching for another match in the  A non-zero starting offset is useful when searching for another match in the
1815  same subject by calling \fBpcre_exec()\fR again after a previous success.  same subject by calling \fBpcre_exec()\fP again after a previous success.
1816  Setting \fIstartoffset\fR differs from just passing over a shortened string and  Setting \fIstartoffset\fP differs from just passing over a shortened string and
1817  setting PCRE_NOTBOL in the case of a pattern that begins with any kind of  setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1818  lookbehind. For example, consider the pattern  lookbehind. For example, consider the pattern
1819    .sp
1820    \\Biss\\B    \eBiss\eB
1821    .sp
1822  which finds occurrences of "iss" in the middle of words. (\\B matches only if  which finds occurrences of "iss" in the middle of words. (\eB matches only if
1823  the current position in the subject is not a word boundary.) When applied to  the current position in the subject is not a word boundary.) When applied to
1824  the string "Mississipi" the first call to \fBpcre_exec()\fR finds the first  the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1825  occurrence. If \fBpcre_exec()\fR is called again with just the remainder of the  occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1826  subject, namely "issipi", it does not match, because \\B is always false at the  subject, namely "issipi", it does not match, because \eB is always false at the
1827  start of the subject, which is deemed to be a word boundary. However, if  start of the subject, which is deemed to be a word boundary. However, if
1828  \fBpcre_exec()\fR is passed the entire string again, but with \fIstartoffset\fR  \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1829  set to 4, it finds the second occurrence of "iss" because it is able to look  set to 4, it finds the second occurrence of "iss" because it is able to look
1830  behind the starting point to discover that it is preceded by a letter.  behind the starting point to discover that it is preceded by a letter.
1831    .P
1832    Finding all the matches in a subject is tricky when the pattern can match an
1833    empty string. It is possible to emulate Perl's /g behaviour by first trying the
1834    match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1835    PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1836    and trying an ordinary match again. There is some code that demonstrates how to
1837    do this in the
1838    .\" HREF
1839    \fBpcredemo\fP
1840    .\"
1841    sample program. In the most general case, you have to check to see if the
1842    newline convention recognizes CRLF as a newline, and if so, and the current
1843    character is CR followed by LF, advance the starting offset by two characters
1844    instead of one.
1845    .P
1846  If a non-zero starting offset is passed when the pattern is anchored, one  If a non-zero starting offset is passed when the pattern is anchored, one
1847  attempt to match at the given offset is tried. This can only succeed if the  attempt to match at the given offset is made. This can only succeed if the
1848  pattern does not require the match to be at the start of the subject.  pattern does not require the match to be at the start of the subject.
1849    .
1850    .
1851    .SS "How \fBpcre_exec()\fP returns captured substrings"
1852    .rs
1853    .sp
1854  In general, a pattern matches a certain portion of the subject, and in  In general, a pattern matches a certain portion of the subject, and in
1855  addition, further substrings from the subject may be picked out by parts of the  addition, further substrings from the subject may be picked out by parts of the
1856  pattern. Following the usage in Jeffrey Friedl's book, this is called  pattern. Following the usage in Jeffrey Friedl's book, this is called
1857  "capturing" in what follows, and the phrase "capturing subpattern" is used for  "capturing" in what follows, and the phrase "capturing subpattern" is used for
1858  a fragment of a pattern that picks out a substring. PCRE supports several other  a fragment of a pattern that picks out a substring. PCRE supports several other
1859  kinds of parenthesized subpattern that do not cause substrings to be captured.  kinds of parenthesized subpattern that do not cause substrings to be captured.
1860    .P
1861  Captured substrings are returned to the caller via a vector of integer offsets  Captured substrings are returned to the caller via a vector of integers whose
1862  whose address is passed in \fIovector\fR. The number of elements in the vector  address is passed in \fIovector\fP. The number of elements in the vector is
1863  is passed in \fIovecsize\fR. The first two-thirds of the vector is used to pass  passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
1864  back captured substrings, each substring using a pair of integers. The  argument is NOT the size of \fIovector\fP in bytes.
1865  remaining third of the vector is used as workspace by \fBpcre_exec()\fR while  .P
1866  matching capturing subpatterns, and is not available for passing back  The first two-thirds of the vector is used to pass back captured substrings,
1867  information. The length passed in \fIovecsize\fR should always be a multiple of  each substring using a pair of integers. The remaining third of the vector is
1868  three. If it is not, it is rounded down.  used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1869    and is not available for passing back information. The number passed in
1870  When a match has been successful, information about captured substrings is  \fIovecsize\fP should always be a multiple of three. If it is not, it is
1871  returned in pairs of integers, starting at the beginning of \fIovector\fR, and  rounded down.
1872  continuing up to two-thirds of its length at the most. The first element of a  .P
1873  pair is set to the offset of the first character in a substring, and the second  When a match is successful, information about captured substrings is returned
1874  is set to the offset of the first character after the end of a substring. The  in pairs of integers, starting at the beginning of \fIovector\fP, and
1875  first pair, \fIovector[0]\fR and \fIovector[1]\fR, identify the portion of the  continuing up to two-thirds of its length at the most. The first element of
1876  subject string matched by the entire pattern. The next pair is used for the  each pair is set to the byte offset of the first character in a substring, and
1877  first capturing subpattern, and so on. The value returned by \fBpcre_exec()\fR  the second is set to the byte offset of the first character after the end of a
1878  is the number of pairs that have been set. If there are no capturing  substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
1879  subpatterns, the return value from a successful match is 1, indicating that  mode. They are not character counts.
1880  just the first pair of offsets has been set.  .P
1881    The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
1882  Some convenience functions are provided for extracting the captured substrings  portion of the subject string matched by the entire pattern. The next pair is
1883  as separate strings. These are described in the following section.  used for the first capturing subpattern, and so on. The value returned by
1884    \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
1885  It is possible for an capturing subpattern number \fIn+1\fR to match some  For example, if two substrings have been captured, the returned value is 3. If
1886  part of the subject when subpattern \fIn\fR has not been used at all. For  there are no capturing subpatterns, the return value from a successful match is
1887  example, if the string "abc" is matched against the pattern (a|(z))(bc)  1, indicating that just the first pair of offsets has been set.
1888  subpatterns 1 and 3 are matched, but 2 is not. When this happens, both offset  .P
 values corresponding to the unused subpattern are set to -1.  
   
1889  If a capturing subpattern is matched repeatedly, it is the last portion of the  If a capturing subpattern is matched repeatedly, it is the last portion of the
1890  string that it matched that gets returned.  string that it matched that is returned.
1891    .P
1892  If the vector is too small to hold all the captured substrings, it is used as  If the vector is too small to hold all the captured substring offsets, it is
1893  far as possible (up to two-thirds of its length), and the function returns a  used as far as possible (up to two-thirds of its length), and the function
1894  value of zero. In particular, if the substring offsets are not of interest,  returns a value of zero. If neither the actual string matched not any captured
1895  \fBpcre_exec()\fR may be called with \fIovector\fR passed as NULL and  substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
1896  \fIovecsize\fR as zero. However, if the pattern contains back references and  passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
1897  the \fIovector\fR isn't big enough to remember the related substrings, PCRE has  back references and the \fIovector\fP is not big enough to remember the related
1898  to get additional memory for use during matching. Thus it is usually advisable  substrings, PCRE has to get additional memory for use during matching. Thus it
1899  to supply an \fIovector\fR.  is usually advisable to supply an \fIovector\fP of reasonable size.
1900    .P
1901  Note that \fBpcre_info()\fR can be used to find out how many capturing  There are some cases where zero is returned (indicating vector overflow) when
1902    in fact the vector is exactly the right size for the final match. For example,
1903    consider the pattern
1904    .sp
1905      (a)(?:(b)c|bd)
1906    .sp
1907    If a vector of 6 elements (allowing for only 1 captured substring) is given
1908    with subject string "abd", \fBpcre_exec()\fP will try to set the second
1909    captured string, thereby recording a vector overflow, before failing to match
1910    "c" and backing up to try the second alternative. The zero return, however,
1911    does correctly indicate that the maximum number of slots (namely 2) have been
1912    filled. In similar cases where there is temporary overflow, but the final
1913    number of used slots is actually less than the maximum, a non-zero value is
1914    returned.
1915    .P
1916    The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
1917  subpatterns there are in a compiled pattern. The smallest size for  subpatterns there are in a compiled pattern. The smallest size for
1918  \fIovector\fR that will allow for \fIn\fR captured substrings, in addition to  \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
1919  the offsets of the substring matched by the whole pattern, is (\fIn\fR+1)*3.  the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
1920    .P
1921  If \fBpcre_exec()\fR fails, it returns a negative number. The following are  It is possible for capturing subpattern number \fIn+1\fP to match some part of
1922    the subject when subpattern \fIn\fP has not been used at all. For example, if
1923    the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1924    function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1925    happens, both values in the offset pairs corresponding to unused subpatterns
1926    are set to -1.
1927    .P
1928    Offset values that correspond to unused subpatterns at the end of the
1929    expression are also set to -1. For example, if the string "abc" is matched
1930    against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1931    return from the function is 2, because the highest used capturing subpattern
1932    number is 1, and the offsets for for the second and third capturing subpatterns
1933    (assuming the vector is large enough, of course) are set to -1.
1934    .P
1935    \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
1936    correspond to capturing parentheses in the pattern are never changed. That is,
1937    if a pattern contains \fIn\fP capturing parentheses, no more than
1938    \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
1939    elements (in the first two-thirds) retain whatever values they previously had.
1940    .P
1941    Some convenience functions are provided for extracting the captured substrings
1942    as separate strings. These are described below.
1943    .
1944    .
1945    .\" HTML <a name="errorlist"></a>
1946    .SS "Error return values from \fBpcre_exec()\fP"
1947    .rs
1948    .sp
1949    If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1950  defined in the header file:  defined in the header file:
1951    .sp
1952    PCRE_ERROR_NOMATCH        (-1)    PCRE_ERROR_NOMATCH        (-1)
1953    .sp
1954  The subject string did not match the pattern.  The subject string did not match the pattern.
1955    .sp
1956    PCRE_ERROR_NULL           (-2)    PCRE_ERROR_NULL           (-2)
1957    .sp
1958  Either \fIcode\fR or \fIsubject\fR was passed as NULL, or \fIovector\fR was  Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
1959  NULL and \fIovecsize\fR was not zero.  NULL and \fIovecsize\fP was not zero.
1960    .sp
1961    PCRE_ERROR_BADOPTION      (-3)    PCRE_ERROR_BADOPTION      (-3)
1962    .sp
1963  An unrecognized bit was set in the \fIoptions\fR argument.  An unrecognized bit was set in the \fIoptions\fP argument.
1964    .sp
1965    PCRE_ERROR_BADMAGIC       (-4)    PCRE_ERROR_BADMAGIC       (-4)
1966    .sp
1967  PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch  PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1968  the case when it is passed a junk pointer. This is the error it gives when the  the case when it is passed a junk pointer and to detect when a pattern that was
1969  magic number isn't present.  compiled in an environment of one endianness is run in an environment with the
1970    other endianness. This is the error that PCRE gives when the magic number is
1971    PCRE_ERROR_UNKNOWN_NODE   (-5)  not present.
1972    .sp
1973      PCRE_ERROR_UNKNOWN_OPCODE (-5)
1974    .sp
1975  While running the pattern match, an unknown item was encountered in the  While running the pattern match, an unknown item was encountered in the
1976  compiled pattern. This error could be caused by a bug in PCRE or by overwriting  compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1977  of the compiled pattern.  of the compiled pattern.
1978    .sp
1979    PCRE_ERROR_NOMEMORY       (-6)    PCRE_ERROR_NOMEMORY       (-6)
1980    .sp
1981  If a pattern contains back references, but the \fIovector\fR that is passed to  If a pattern contains back references, but the \fIovector\fP that is passed to
1982  \fBpcre_exec()\fR is not big enough to remember the referenced substrings, PCRE  \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
1983  gets a block of memory at the start of matching to use for this purpose. If the  gets a block of memory at the start of matching to use for this purpose. If the
1984  call via \fBpcre_malloc()\fR fails, this error is given. The memory is freed at  call via \fBpcre_malloc()\fP fails, this error is given. The memory is
1985  the end of matching.  automatically freed at the end of matching.
1986    .P
1987    This error is also given if \fBpcre_stack_malloc()\fP fails in
1988    \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
1989    \fB--disable-stack-for-recursion\fP.
1990    .sp
1991    PCRE_ERROR_NOSUBSTRING    (-7)    PCRE_ERROR_NOSUBSTRING    (-7)
1992    .sp
1993  This error is used by the \fBpcre_copy_substring()\fR,  This error is used by the \fBpcre_copy_substring()\fP,
1994  \fBpcre_get_substring()\fR, and \fBpcre_get_substring_list()\fR functions (see  \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
1995  below). It is never returned by \fBpcre_exec()\fR.  below). It is never returned by \fBpcre_exec()\fP.
1996    .sp
1997    PCRE_ERROR_MATCHLIMIT     (-8)    PCRE_ERROR_MATCHLIMIT     (-8)
1998    .sp
1999  The recursion and backtracking limit, as specified by the \fImatch_limit\fR  The backtracking limit, as specified by the \fImatch_limit\fP field in a
2000  field in a \fBpcre_extra\fR structure (or defaulted) was reached. See the  \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2001  description above.  above.
2002    .sp
2003    PCRE_ERROR_CALLOUT        (-9)    PCRE_ERROR_CALLOUT        (-9)
2004    .sp
2005  This error is never generated by \fBpcre_exec()\fR itself. It is provided for  This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2006  use by callout functions that want to yield a distinctive error code. See the  use by callout functions that want to yield a distinctive error code. See the
2007  \fBpcrecallout\fR documentation for details.  .\" HREF
2008    \fBpcrecallout\fP
2009  .SH EXTRACTING CAPTURED SUBSTRINGS BY NUMBER  .\"
2010    documentation for details.
2011    .sp
2012      PCRE_ERROR_BADUTF8        (-10)
2013    .sp
2014    A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2015    and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2016    (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2017    UTF-8 character is placed in the first element, and a reason code is placed in
2018    the second element. The reason codes are listed in the
2019    .\" HTML <a href="#badutf8reasons">
2020    .\" </a>
2021    following section.
2022    .\"
2023    For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2024    truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2025    PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2026    .sp
2027      PCRE_ERROR_BADUTF8_OFFSET (-11)
2028    .sp
2029    The UTF-8 byte sequence that was passed as a subject was checked and found to
2030    be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2031    \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2032    end of the subject.
2033    .sp
2034      PCRE_ERROR_PARTIAL        (-12)
2035    .sp
2036    The subject string did not match, but it did match partially. See the
2037    .\" HREF
2038    \fBpcrepartial\fP
2039    .\"
2040    documentation for details of partial matching.
2041    .sp
2042      PCRE_ERROR_BADPARTIAL     (-13)
2043    .sp
2044    This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2045    option was used with a compiled pattern containing items that were not
2046    supported for partial matching. From release 8.00 onwards, there are no
2047    restrictions on partial matching.
2048    .sp
2049      PCRE_ERROR_INTERNAL       (-14)
2050    .sp
2051    An unexpected internal error has occurred. This error could be caused by a bug
2052    in PCRE or by overwriting of the compiled pattern.
2053    .sp
2054      PCRE_ERROR_BADCOUNT       (-15)
2055    .sp
2056    This error is given if the value of the \fIovecsize\fP argument is negative.
2057    .sp
2058      PCRE_ERROR_RECURSIONLIMIT (-21)
2059    .sp
2060    The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2061    field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2062    description above.
2063    .sp
2064      PCRE_ERROR_BADNEWLINE     (-23)
2065    .sp
2066    An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2067    .sp
2068      PCRE_ERROR_BADOFFSET      (-24)
2069    .sp
2070    The value of \fIstartoffset\fP was negative or greater than the length of the
2071    subject, that is, the value in \fIlength\fP.
2072    .sp
2073      PCRE_ERROR_SHORTUTF8      (-25)
2074    .sp
2075    This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2076    ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2077    Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2078    fact sufficient to detect this case, but this special error code for
2079    PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2080    retained for backwards compatibility.
2081    .sp
2082      PCRE_ERROR_RECURSELOOP    (-26)
2083    .sp
2084    This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2085    the pattern. Specifically, it means that either the whole pattern or a
2086    subpattern has been called recursively for the second time at the same position
2087    in the subject string. Some simple patterns that might do this are detected and
2088    faulted at compile time, but more complicated cases, in particular mutual
2089    recursions between two different subpatterns, cannot be detected until run
2090    time.
2091    .sp
2092      PCRE_ERROR_JIT_STACKLIMIT (-27)
2093    .sp
2094    This error is returned when a pattern that was successfully studied using a
2095    JIT compile option is being matched, but the memory available for the
2096    just-in-time processing stack is not large enough. See the
2097    .\" HREF
2098    \fBpcrejit\fP
2099    .\"
2100    documentation for more details.
2101    .sp
2102      PCRE_ERROR_BADMODE (-28)
2103    .sp
2104    This error is given if a pattern that was compiled by the 8-bit library is
2105    passed to a 16-bit library function, or vice versa.
2106    .sp
2107      PCRE_ERROR_BADENDIANNESS (-29)
2108    .sp
2109    This error is given if a pattern that was compiled and saved is reloaded on a
2110    host with different endianness. The utility function
2111    \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2112    so that it runs on the new host.
2113    .P
2114    Error numbers -16 to -20 and -22 are not used by \fBpcre_exec()\fP.
2115    .
2116    .
2117    .\" HTML <a name="badutf8reasons"></a>
2118    .SS "Reason codes for invalid UTF-8 strings"
2119  .rs  .rs
2120  .sp  .sp
2121  .B int pcre_copy_substring(const char *\fIsubject\fR, int *\fIovector\fR,  This section applies only to the 8-bit library. The corresponding information
2122    for the 16-bit library is given in the
2123    .\" HREF
2124    \fBpcre16\fP
2125    .\"
2126    page.
2127    .P
2128    When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2129    PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2130    least 2, the offset of the start of the invalid UTF-8 character is placed in
2131    the first output vector element (\fIovector[0]\fP) and a reason code is placed
2132    in the second element (\fIovector[1]\fP). The reason codes are given names in
2133    the \fBpcre.h\fP header file:
2134    .sp
2135      PCRE_UTF8_ERR1
2136      PCRE_UTF8_ERR2
2137      PCRE_UTF8_ERR3
2138      PCRE_UTF8_ERR4
2139      PCRE_UTF8_ERR5
2140    .sp
2141    The string ends with a truncated UTF-8 character; the code specifies how many
2142    bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2143    no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2144    allows for up to 6 bytes, and this is checked first; hence the possibility of
2145    4 or 5 missing bytes.
2146    .sp
2147      PCRE_UTF8_ERR6
2148      PCRE_UTF8_ERR7
2149      PCRE_UTF8_ERR8
2150      PCRE_UTF8_ERR9
2151      PCRE_UTF8_ERR10
2152    .sp
2153    The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2154    character do not have the binary value 0b10 (that is, either the most
2155    significant bit is 0, or the next bit is 1).
2156    .sp
2157      PCRE_UTF8_ERR11
2158      PCRE_UTF8_ERR12
2159    .sp
2160    A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2161    these code points are excluded by RFC 3629.
2162    .sp
2163      PCRE_UTF8_ERR13
2164    .sp
2165    A 4-byte character has a value greater than 0x10fff; these code points are
2166    excluded by RFC 3629.
2167    .sp
2168      PCRE_UTF8_ERR14
2169    .sp
2170    A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2171    code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2172    from UTF-8.
2173    .sp
2174      PCRE_UTF8_ERR15
2175      PCRE_UTF8_ERR16
2176      PCRE_UTF8_ERR17
2177      PCRE_UTF8_ERR18
2178      PCRE_UTF8_ERR19
2179    .sp
2180    A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2181    value that can be represented by fewer bytes, which is invalid. For example,
2182    the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2183    one byte.
2184    .sp
2185      PCRE_UTF8_ERR20
2186    .sp
2187    The two most significant bits of the first byte of a character have the binary
2188    value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2189    byte can only validly occur as the second or subsequent byte of a multi-byte
2190    character.
2191    .sp
2192      PCRE_UTF8_ERR21
2193    .sp
2194    The first byte of a character has the value 0xfe or 0xff. These values can
2195    never occur in a valid UTF-8 string.
2196    .
2197    .
2198    .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2199    .rs
2200    .sp
2201    .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2202  .ti +5n  .ti +5n
2203  .B int \fIstringcount\fR, int \fIstringnumber\fR, char *\fIbuffer\fR,  .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
2204  .ti +5n  .ti +5n
2205  .B int \fIbuffersize\fR);  .B int \fIbuffersize\fP);
2206  .PP  .PP
2207  .br  .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
 .B int pcre_get_substring(const char *\fIsubject\fR, int *\fIovector\fR,  
2208  .ti +5n  .ti +5n
2209  .B int \fIstringcount\fR, int \fIstringnumber\fR,  .B int \fIstringcount\fP, int \fIstringnumber\fP,
2210  .ti +5n  .ti +5n
2211  .B const char **\fIstringptr\fR);  .B const char **\fIstringptr\fP);
2212  .PP  .PP
2213  .br  .B int pcre_get_substring_list(const char *\fIsubject\fP,
 .B int pcre_get_substring_list(const char *\fIsubject\fR,  
2214  .ti +5n  .ti +5n
2215  .B int *\fIovector\fR, int \fIstringcount\fR, "const char ***\fIlistptr\fR);"  .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
2216  .PP  .PP
2217  Captured substrings can be accessed directly by using the offsets returned by  Captured substrings can be accessed directly by using the offsets returned by
2218  \fBpcre_exec()\fR in \fIovector\fR. For convenience, the functions  \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2219  \fBpcre_copy_substring()\fR, \fBpcre_get_substring()\fR, and  \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2220  \fBpcre_get_substring_list()\fR are provided for extracting captured substrings  \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2221  as new, separate, zero-terminated strings. These functions identify substrings  as new, separate, zero-terminated strings. These functions identify substrings
2222  by number. The next section describes functions for extracting named  by number. The next section describes functions for extracting named
2223  substrings. A substring that contains a binary zero is correctly extracted and  substrings.
2224  has a further zero added on the end, but the result is not, of course,  .P
2225  a C string.  A substring that contains a binary zero is correctly extracted and has a
2226    further zero added on the end, but the result is not, of course, a C string.
2227    However, you can process such a string by referring to the length that is
2228    returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2229    Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2230    for handling strings containing binary zeros, because the end of the final
2231    string is not independently indicated.
2232    .P
2233  The first three arguments are the same for all three of these functions:  The first three arguments are the same for all three of these functions:
2234  \fIsubject\fR is the subject string which has just been successfully matched,  \fIsubject\fP is the subject string that has just been successfully matched,
2235  \fIovector\fR is a pointer to the vector of integer offsets that was passed to  \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2236  \fBpcre_exec()\fR, and \fIstringcount\fR is the number of substrings that were  \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2237  captured by the match, including the substring that matched the entire regular  captured by the match, including the substring that matched the entire regular
2238  expression. This is the value returned by \fBpcre_exec\fR if it is greater than  expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2239  zero. If \fBpcre_exec()\fR returned zero, indicating that it ran out of space  than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2240  in \fIovector\fR, the value passed as \fIstringcount\fR should be the size of  space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2241  the vector divided by three.  number of elements in the vector divided by three.
2242    .P
2243  The functions \fBpcre_copy_substring()\fR and \fBpcre_get_substring()\fR  The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2244  extract a single substring, whose number is given as \fIstringnumber\fR. A  extract a single substring, whose number is given as \fIstringnumber\fP. A
2245  value of zero extracts the substring that matched the entire pattern, while  value of zero extracts the substring that matched the entire pattern, whereas
2246  higher values extract the captured substrings. For \fBpcre_copy_substring()\fR,  higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2247  the string is placed in \fIbuffer\fR, whose length is given by  the string is placed in \fIbuffer\fP, whose length is given by
2248  \fIbuffersize\fR, while for \fBpcre_get_substring()\fR a new block of memory is  \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2249  obtained via \fBpcre_malloc\fR, and its address is returned via  obtained via \fBpcre_malloc\fP, and its address is returned via
2250  \fIstringptr\fR. The yield of the function is the length of the string, not  \fIstringptr\fP. The yield of the function is the length of the string, not
2251  including the terminating zero, or one of  including the terminating zero, or one of these error codes:
2252    .sp
2253    PCRE_ERROR_NOMEMORY       (-6)    PCRE_ERROR_NOMEMORY       (-6)
2254    .sp
2255  The buffer was too small for \fBpcre_copy_substring()\fR, or the attempt to get  The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2256  memory failed for \fBpcre_get_substring()\fR.  memory failed for \fBpcre_get_substring()\fP.
2257    .sp
2258    PCRE_ERROR_NOSUBSTRING    (-7)    PCRE_ERROR_NOSUBSTRING    (-7)
2259    .sp
2260  There is no substring whose number is \fIstringnumber\fR.  There is no substring whose number is \fIstringnumber\fP.
2261    .P
2262  The \fBpcre_get_substring_list()\fR function extracts all available substrings  The \fBpcre_get_substring_list()\fP function extracts all available substrings
2263  and builds a list of pointers to them. All this is done in a single block of  and builds a list of pointers to them. All this is done in a single block of
2264  memory which is obtained via \fBpcre_malloc\fR. The address of the memory block  memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2265  is returned via \fIlistptr\fR, which is also the start of the list of string  is returned via \fIlistptr\fP, which is also the start of the list of string
2266  pointers. The end of the list is marked by a NULL pointer. The yield of the  pointers. The end of the list is marked by a NULL pointer. The yield of the
2267  function is zero if all went well, or  function is zero if all went well, or the error code
2268    .sp
2269    PCRE_ERROR_NOMEMORY       (-6)    PCRE_ERROR_NOMEMORY       (-6)
2270    .sp
2271  if the attempt to get the memory block failed.  if the attempt to get the memory block failed.
2272    .P
2273  When any of these functions encounter a substring that is unset, which can  When any of these functions encounter a substring that is unset, which can
2274  happen when capturing subpattern number \fIn+1\fR matches some part of the  happen when capturing subpattern number \fIn+1\fP matches some part of the
2275  subject, but subpattern \fIn\fR has not been used at all, they return an empty  subject, but subpattern \fIn\fP has not been used at all, they return an empty
2276  string. This can be distinguished from a genuine zero-length substring by  string. This can be distinguished from a genuine zero-length substring by
2277  inspecting the appropriate offset in \fIovector\fR, which is negative for unset  inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2278  substrings.  substrings.
2279    .P
2280  The two convenience functions \fBpcre_free_substring()\fR and  The two convenience functions \fBpcre_free_substring()\fP and
2281  \fBpcre_free_substring_list()\fR can be used to free the memory returned by  \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2282  a previous call of \fBpcre_get_substring()\fR or  a previous call of \fBpcre_get_substring()\fP or
2283  \fBpcre_get_substring_list()\fR, respectively. They do nothing more than call  \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2284  the function pointed to by \fBpcre_free\fR, which of course could be called  the function pointed to by \fBpcre_free\fP, which of course could be called
2285  directly from a C program. However, PCRE is used in some situations where it is  directly from a C program. However, PCRE is used in some situations where it is
2286  linked via a special interface to another programming language which cannot use  linked via a special interface to another programming language that cannot use
2287  \fBpcre_free\fR directly; it is for these cases that the functions are  \fBpcre_free\fP directly; it is for these cases that the functions are
2288  provided.  provided.
2289    .
2290  .SH EXTRACTING CAPTURED SUBSTRINGS BY NAME  .
2291    .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2292  .rs  .rs
2293  .sp  .sp
2294  .B int pcre_copy_named_substring(const pcre *\fIcode\fR,  .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2295  .ti +5n  .ti +5n
2296  .B const char *\fIsubject\fR, int *\fIovector\fR,  .B const char *\fIname\fP);
2297    .PP
2298    .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2299  .ti +5n  .ti +5n
2300  .B int \fIstringcount\fR, const char *\fIstringname\fR,  .B const char *\fIsubject\fP, int *\fIovector\fP,
2301  .ti +5n  .ti +5n
2302  .B char *\fIbuffer\fR, int \fIbuffersize\fR);  .B int \fIstringcount\fP, const char *\fIstringname\fP,
 .PP  
 .br  
 .B int pcre_get_stringnumber(const pcre *\fIcode\fR,  
2303  .ti +5n  .ti +5n
2304  .B const char *\fIname\fR);  .B char *\fIbuffer\fP, int \fIbuffersize\fP);
2305  .PP  .PP
2306  .br  .B int pcre_get_named_substring(const pcre *\fIcode\fP,
 .B int pcre_get_named_substring(const pcre *\fIcode\fR,  
2307  .ti +5n  .ti +5n
2308  .B const char *\fIsubject\fR, int *\fIovector\fR,  .B const char *\fIsubject\fP, int *\fIovector\fP,
2309  .ti +5n  .ti +5n
2310  .B int \fIstringcount\fR, const char *\fIstringname\fR,  .B int \fIstringcount\fP, const char *\fIstringname\fP,
2311  .ti +5n  .ti +5n
2312  .B const char **\fIstringptr\fR);  .B const char **\fIstringptr\fP);
2313  .PP  .PP
2314  To extract a substring by name, you first have to find associated number. This  To extract a substring by name, you first have to find associated number.
2315  can be done by calling \fBpcre_get_stringnumber()\fR. The first argument is the  For example, for this pattern
2316  compiled pattern, and the second is the name. For example, for this pattern  .sp
2317      (a+)b(?<xxx>\ed+)...
2318    ab(?<xxx>\\d+)...  .sp
2319    the number of the subpattern called "xxx" is 2. If the name is known to be
2320  the number of the subpattern called "xxx" is 1. Given the number, you can then  unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2321  extract the substring directly, or use one of the functions described in the  calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2322  previous section. For convenience, there are also two functions that do the  pattern, and the second is the name. The yield of the function is the
2323  whole job.  subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2324    that name.
2325  Most of the arguments of \fIpcre_copy_named_substring()\fR and  .P
2326  \fIpcre_get_named_substring()\fR are the same as those for the functions that  Given the number, you can extract the substring directly, or use one of the
2327  extract by number, and so are not re-described here. There are just two  functions described in the previous section. For convenience, there are also
2328  differences.  two functions that do the whole job.
2329    .P
2330    Most of the arguments of \fBpcre_copy_named_substring()\fP and
2331    \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2332    functions that extract by number. As these are described in the previous
2333    section, they are not re-described here. There are just two differences:
2334    .P
2335  First, instead of a substring number, a substring name is given. Second, there  First, instead of a substring number, a substring name is given. Second, there
2336  is an extra argument, given at the start, which is a pointer to the compiled  is an extra argument, given at the start, which is a pointer to the compiled
2337  pattern. This is needed in order to gain access to the name-to-number  pattern. This is needed in order to gain access to the name-to-number
2338  translation table.  translation table.
2339    .P
2340  These functions call \fBpcre_get_stringnumber()\fR, and if it succeeds, they  These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2341  then call \fIpcre_copy_substring()\fR or \fIpcre_get_substring()\fR, as  then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2342  appropriate.  appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2343    the behaviour may not be what you want (see the next section).
2344  .in 0  .P
2345  Last updated: 03 February 2003  \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2346  .br  subpatterns with the same number, as described in the
2347  Copyright (c) 1997-2003 University of Cambridge.  .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2348    .\" </a>
2349    section on duplicate subpattern numbers
2350    .\"
2351    in the
2352    .\" HREF
2353    \fBpcrepattern\fP
2354    .\"
2355    page, you cannot use names to distinguish the different subpatterns, because
2356    names are not included in the compiled code. The matching process uses only
2357    numbers. For this reason, the use of different names for subpatterns of the
2358    same number causes an error at compile time.
2359    .
2360    .
2361    .SH "DUPLICATE SUBPATTERN NAMES"
2362    .rs
2363    .sp
2364    .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2365    .ti +5n
2366    .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
2367    .PP
2368    When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2369    are not required to be unique. (Duplicate names are always allowed for
2370    subpatterns with the same number, created by using the (?| feature. Indeed, if
2371    such subpatterns are named, they are required to use the same names.)
2372    .P
2373    Normally, patterns with duplicate names are such that in any one match, only
2374    one of the named subpatterns participates. An example is shown in the
2375    .\" HREF
2376    \fBpcrepattern\fP
2377    .\"
2378    documentation.
2379    .P
2380    When duplicates are present, \fBpcre_copy_named_substring()\fP and
2381    \fBpcre_get_named_substring()\fP return the first substring corresponding to
2382    the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2383    returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2384    returns one of the numbers that are associated with the name, but it is not
2385    defined which it is.
2386    .P
2387    If you want to get full details of all captured substrings for a given name,
2388    you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2389    argument is the compiled pattern, and the second is the name. The third and
2390    fourth are pointers to variables which are updated by the function. After it
2391    has run, they point to the first and last entries in the name-to-number table
2392    for the given name. The function itself returns the length of each entry, or
2393    PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2394    described above in the section entitled \fIInformation about a pattern\fP
2395    .\" HTML <a href="#infoaboutpattern">
2396    .\" </a>
2397    above.
2398    .\"
2399    Given all the relevant entries for the name, you can extract each of their
2400    numbers, and hence the captured data, if any.
2401    .
2402    .
2403    .SH "FINDING ALL POSSIBLE MATCHES"
2404    .rs
2405    .sp
2406    The traditional matching function uses a similar algorithm to Perl, which stops
2407    when it finds the first match, starting at a given point in the subject. If you
2408    want to find all possible matches, or the longest possible match, consider
2409    using the alternative matching function (see below) instead. If you cannot use
2410    the alternative function, but still need to find all possible matches, you
2411    can kludge it up by making use of the callout facility, which is described in
2412    the
2413    .\" HREF
2414    \fBpcrecallout\fP
2415    .\"
2416    documentation.
2417    .P
2418    What you have to do is to insert a callout right at the end of the pattern.
2419    When your callout function is called, extract and save the current matched
2420    substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2421    other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2422    will yield PCRE_ERROR_NOMATCH.
2423    .
2424    .
2425    .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2426    .rs
2427    .sp
2428    Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2429    stack, which in certain environments can be rather limited in size. Some users
2430    find it helpful to have an estimate of the amount of stack that is used by
2431    \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2432    .\" HREF
2433    \fBpcrestack\fP
2434    .\"
2435    documentation. The estimate that is output by \fBpcretest\fP when called with
2436    the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2437    the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2438    .P
2439    Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2440    the negative error code PCRE_ERROR_NULL, but with this special combination of
2441    arguments, it returns instead a negative number whose absolute value is the
2442    approximate stack frame size in bytes. (A negative number is used so that it is
2443    clear that no match has happened.) The value is approximate because in some
2444    cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2445    additional variables on the stack.
2446    .P
2447    If PCRE has been compiled to use the heap instead of the stack for recursion,
2448    the value returned is the size of each block that is obtained from the heap.
2449    .
2450    .
2451    .\" HTML <a name="dfamatch"></a>
2452    .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2453    .rs
2454    .sp
2455    .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2456    .ti +5n
2457    .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2458    .ti +5n
2459    .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2460    .ti +5n
2461    .B int *\fIworkspace\fP, int \fIwscount\fP);
2462    .P
2463    The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2464    a compiled pattern, using a matching algorithm that scans the subject string
2465    just once, and does not backtrack. This has different characteristics to the
2466    normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2467    patterns are not supported. Nevertheless, there are times when this kind of
2468    matching can be useful. For a discussion of the two matching algorithms, and a
2469    list of features that \fBpcre_dfa_exec()\fP does not support, see the
2470    .\" HREF
2471    \fBpcrematching\fP
2472    .\"
2473    documentation.
2474    .P
2475    The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2476    \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2477    different way, and this is described below. The other common arguments are used
2478    in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2479    here.
2480    .P
2481    The two additional arguments provide workspace for the function. The workspace
2482    vector should contain at least 20 elements. It is used for keeping track of
2483    multiple paths through the pattern tree. More workspace will be needed for
2484    patterns and subjects where there are a lot of potential matches.
2485    .P
2486    Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2487    .sp
2488      int rc;
2489      int ovector[10];
2490      int wspace[20];
2491      rc = pcre_dfa_exec(
2492        re,             /* result of pcre_compile() */
2493        NULL,           /* we didn't study the pattern */
2494        "some string",  /* the subject string */
2495        11,             /* the length of the subject string */
2496        0,              /* start at offset 0 in the subject */
2497        0,              /* default options */
2498        ovector,        /* vector of integers for substring information */
2499        10,             /* number of elements (NOT size in bytes) */
2500        wspace,         /* working space vector */
2501        20);            /* number of elements (NOT size in bytes) */
2502    .
2503    .SS "Option bits for \fBpcre_dfa_exec()\fP"
2504    .rs
2505    .sp
2506    The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2507    zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2508    PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2509    PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2510    PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2511    All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2512    so their description is not repeated here.
2513    .sp
2514      PCRE_PARTIAL_HARD
2515      PCRE_PARTIAL_SOFT
2516    .sp
2517    These have the same general effect as they do for \fBpcre_exec()\fP, but the
2518    details are slightly different. When PCRE_PARTIAL_HARD is set for
2519    \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2520    is reached and there is still at least one matching possibility that requires
2521    additional characters. This happens even if some complete matches have also
2522    been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2523    is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2524    there have been no complete matches, but there is still at least one matching
2525    possibility. The portion of the string that was inspected when the longest
2526    partial match was found is set as the first matching string in both cases.
2527    There is a more detailed discussion of partial and multi-segment matching, with
2528    examples, in the
2529    .\" HREF
2530    \fBpcrepartial\fP
2531    .\"
2532    documentation.
2533    .sp
2534      PCRE_DFA_SHORTEST
2535    .sp
2536    Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2537    soon as it has found one match. Because of the way the alternative algorithm
2538    works, this is necessarily the shortest possible match at the first possible
2539    matching point in the subject string.
2540    .sp
2541      PCRE_DFA_RESTART
2542    .sp
2543    When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2544    again, with additional subject characters, and have it continue with the same
2545    match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2546    \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2547    before because data about the match so far is left in them after a partial
2548    match. There is more discussion of this facility in the
2549    .\" HREF
2550    \fBpcrepartial\fP
2551    .\"
2552    documentation.
2553    .
2554    .
2555    .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2556    .rs
2557    .sp
2558    When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2559    substring in the subject. Note, however, that all the matches from one run of
2560    the function start at the same point in the subject. The shorter matches are
2561    all initial substrings of the longer matches. For example, if the pattern
2562    .sp
2563      <.*>
2564    .sp
2565    is matched against the string
2566    .sp
2567      This is <something> <something else> <something further> no more
2568    .sp
2569    the three matched strings are
2570    .sp
2571      <something>
2572      <something> <something else>
2573      <something> <something else> <something further>
2574    .sp
2575    On success, the yield of the function is a number greater than zero, which is
2576    the number of matched substrings. The substrings themselves are returned in
2577    \fIovector\fP. Each string uses two elements; the first is the offset to the
2578    start, and the second is the offset to the end. In fact, all the strings have
2579    the same start offset. (Space could have been saved by giving this only once,
2580    but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2581    returns data, even though the meaning of the strings is different.)
2582    .P
2583    The strings are returned in reverse order of length; that is, the longest
2584    matching string is given first. If there were too many matches to fit into
2585    \fIovector\fP, the yield of the function is zero, and the vector is filled with
2586    the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2587    the entire \fIovector\fP for returning matched strings.
2588    .
2589    .
2590    .SS "Error returns from \fBpcre_dfa_exec()\fP"
2591    .rs
2592    .sp
2593    The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2594    Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2595    described
2596    .\" HTML <a href="#errorlist">
2597    .\" </a>
2598    above.
2599    .\"
2600    There are in addition the following errors that are specific to
2601    \fBpcre_dfa_exec()\fP:
2602    .sp
2603      PCRE_ERROR_DFA_UITEM      (-16)
2604    .sp
2605    This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2606    that it does not support, for instance, the use of \eC or a back reference.
2607    .sp
2608      PCRE_ERROR_DFA_UCOND      (-17)
2609    .sp
2610    This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2611    uses a back reference for the condition, or a test for recursion in a specific
2612    group. These are not supported.
2613    .sp
2614      PCRE_ERROR_DFA_UMLIMIT    (-18)
2615    .sp
2616    This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2617    block that contains a setting of the \fImatch_limit\fP or
2618    \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2619    meaningless for DFA matching).
2620    .sp
2621      PCRE_ERROR_DFA_WSSIZE     (-19)
2622    .sp
2623    This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2624    \fIworkspace\fP vector.
2625    .sp
2626      PCRE_ERROR_DFA_RECURSE    (-20)
2627    .sp
2628    When a recursive subpattern is processed, the matching function calls itself
2629    recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2630    error is given if the output vector is not large enough. This should be
2631    extremely rare, as a vector of size 1000 is used.
2632    .
2633    .
2634    .SH "SEE ALSO"
2635    .rs
2636    .sp
2637    \fBpcre16\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
2638    \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
2639    \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
2640    .
2641    .
2642    .SH AUTHOR
2643    .rs
2644    .sp
2645    .nf
2646    Philip Hazel
2647    University Computing Service
2648    Cambridge CB2 3QH, England.
2649    .fi
2650    .
2651    .
2652    .SH REVISION
2653    .rs
2654    .sp
2655    .nf
2656    Last updated: 24 February 2012
2657    Copyright (c) 1997-2012 University of Cambridge.
2658    .fi

Legend:
Removed from v.65  
changed lines
  Added in v.932

  ViewVC Help
Powered by ViewVC 1.1.5