/[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 73 by nigel, Sat Feb 24 21:40:30 2007 UTC revision 776 by ph10, Thu Dec 1 11:02:56 2011 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
4  .SH SYNOPSIS OF PCRE API  .SH "PCRE NATIVE API BASIC FUNCTIONS"
5  .rs  .rs
6  .sp  .sp
7  .B #include <pcre.h>  .B #include <pcre.h>
8  .PP  .PP
9  .SM  .SM
10  .br  .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
 .B pcre *pcre_compile(const char *\fIpattern\fR, int \fIoptions\fR,  
11  .ti +5n  .ti +5n
12  .B const char **\fIerrptr\fR, int *\fIerroffset\fR,  .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
13  .ti +5n  .ti +5n
14  .B const unsigned char *\fItableptr\fR);  .B const unsigned char *\fItableptr\fP);
15  .PP  .PP
16  .br  .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
 .B pcre_extra *pcre_study(const pcre *\fIcode\fR, int \fIoptions\fR,  
17  .ti +5n  .ti +5n
18  .B const char **\fIerrptr\fR);  .B int *\fIerrorcodeptr\fP,
19    .ti +5n
20    .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
21    .ti +5n
22    .B const unsigned char *\fItableptr\fP);
23    .PP
24    .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
25    .ti +5n
26    .B const char **\fIerrptr\fP);
27  .PP  .PP
28  .br  .B void pcre_free_study(pcre_extra *\fIextra\fP);
29  .B int pcre_exec(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"  .PP
30    .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
31  .ti +5n  .ti +5n
32  .B "const char *\fIsubject\fR," int \fIlength\fR, int \fIstartoffset\fR,  .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
33  .ti +5n  .ti +5n
34  .B int \fIoptions\fR, int *\fIovector\fR, int \fIovecsize\fR);  .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
35    .
36    .
37    .SH "PCRE NATIVE API AUXILIARY FUNCTIONS"
38    .rs
39    .sp
40    .B pcre_jit_stack *pcre_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
41    .PP
42    .B void pcre_jit_stack_free(pcre_jit_stack *\fIstack\fP);
43  .PP  .PP
44  .br  .B void pcre_assign_jit_stack(pcre_extra *\fIextra\fP,
 .B int pcre_copy_named_substring(const pcre *\fIcode\fR,  
45  .ti +5n  .ti +5n
46  .B const char *\fIsubject\fR, int *\fIovector\fR,  .B pcre_jit_callback \fIcallback\fP, void *\fIdata\fP);
47    .PP
48    .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
49    .ti +5n
50    .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
51  .ti +5n  .ti +5n
52  .B int \fIstringcount\fR, const char *\fIstringname\fR,  .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
53  .ti +5n  .ti +5n
54  .B char *\fIbuffer\fR, int \fIbuffersize\fR);  .B int *\fIworkspace\fP, int \fIwscount\fP);
55  .PP  .PP
56  .br  .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
 .B int pcre_copy_substring(const char *\fIsubject\fR, int *\fIovector\fR,  
57  .ti +5n  .ti +5n
58  .B int \fIstringcount\fR, int \fIstringnumber\fR, char *\fIbuffer\fR,  .B const char *\fIsubject\fP, int *\fIovector\fP,
59  .ti +5n  .ti +5n
60  .B int \fIbuffersize\fR);  .B int \fIstringcount\fP, const char *\fIstringname\fP,
61    .ti +5n
62    .B char *\fIbuffer\fP, int \fIbuffersize\fP);
63  .PP  .PP
64  .br  .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
65  .B int pcre_get_named_substring(const pcre *\fIcode\fR,  .ti +5n
66    .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
67    .ti +5n
68    .B int \fIbuffersize\fP);
69    .PP
70    .B int pcre_get_named_substring(const pcre *\fIcode\fP,
71    .ti +5n
72    .B const char *\fIsubject\fP, int *\fIovector\fP,
73  .ti +5n  .ti +5n
74  .B const char *\fIsubject\fR, int *\fIovector\fR,  .B int \fIstringcount\fP, const char *\fIstringname\fP,
75  .ti +5n  .ti +5n
76  .B int \fIstringcount\fR, const char *\fIstringname\fR,  .B const char **\fIstringptr\fP);
77    .PP
78    .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
79  .ti +5n  .ti +5n
80  .B const char **\fIstringptr\fR);  .B const char *\fIname\fP);
81  .PP  .PP
82  .br  .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
 .B int pcre_get_stringnumber(const pcre *\fIcode\fR,  
83  .ti +5n  .ti +5n
84  .B const char *\fIname\fR);  .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
85  .PP  .PP
86  .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,  
87  .ti +5n  .ti +5n
88  .B int \fIstringcount\fR, int \fIstringnumber\fR,  .B int \fIstringcount\fP, int \fIstringnumber\fP,
89  .ti +5n  .ti +5n
90  .B const char **\fIstringptr\fR);  .B const char **\fIstringptr\fP);
91  .PP  .PP
92  .br  .B int pcre_get_substring_list(const char *\fIsubject\fP,
 .B int pcre_get_substring_list(const char *\fIsubject\fR,  
93  .ti +5n  .ti +5n
94  .B int *\fIovector\fR, int \fIstringcount\fR, "const char ***\fIlistptr\fR);"  .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
95  .PP  .PP
96  .br  .B void pcre_free_substring(const char *\fIstringptr\fP);
 .B void pcre_free_substring(const char *\fIstringptr\fR);  
97  .PP  .PP
98  .br  .B void pcre_free_substring_list(const char **\fIstringptr\fP);
 .B void pcre_free_substring_list(const char **\fIstringptr\fR);  
99  .PP  .PP
 .br  
100  .B const unsigned char *pcre_maketables(void);  .B const unsigned char *pcre_maketables(void);
101  .PP  .PP
102  .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,"  
103  .ti +5n  .ti +5n
104  .B int \fIwhat\fR, void *\fIwhere\fR);  .B int \fIwhat\fP, void *\fIwhere\fP);
105  .PP  .PP
106  .br  .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
107  .B int pcre_info(const pcre *\fIcode\fR, int *\fIoptptr\fR, int  .B *\fIfirstcharptr\fP);
 .B *\fIfirstcharptr\fR);  
108  .PP  .PP
109  .br  .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
 .B int pcre_config(int \fIwhat\fR, void *\fIwhere\fR);  
110  .PP  .PP
111  .br  .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
 .B char *pcre_version(void);  
112  .PP  .PP
113  .br  .B char *pcre_version(void);
114    .
115    .
116    .SH "PCRE NATIVE API INDIRECTED FUNCTIONS"
117    .rs
118    .sp
119  .B void *(*pcre_malloc)(size_t);  .B void *(*pcre_malloc)(size_t);
120  .PP  .PP
 .br  
121  .B void (*pcre_free)(void *);  .B void (*pcre_free)(void *);
122  .PP  .PP
 .br  
123  .B void *(*pcre_stack_malloc)(size_t);  .B void *(*pcre_stack_malloc)(size_t);
124  .PP  .PP
 .br  
125  .B void (*pcre_stack_free)(void *);  .B void (*pcre_stack_free)(void *);
126  .PP  .PP
 .br  
127  .B int (*pcre_callout)(pcre_callout_block *);  .B int (*pcre_callout)(pcre_callout_block *);
128    .
129  .SH PCRE API  .
130    .SH "PCRE API OVERVIEW"
131  .rs  .rs
132  .sp  .sp
133  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
134  a set of wrapper functions that correspond to the POSIX regular expression API.  also some wrapper functions that correspond to the POSIX regular expression
135  These are described in the \fBpcreposix\fR documentation.  API, but they do not give access to all the functionality. They are described
136    in the
137  The native API function prototypes are defined in the header file \fBpcre.h\fR,  .\" HREF
138  and on Unix systems the library itself is called \fBlibpcre.a\fR, so can be  \fBpcreposix\fP
139  accessed by adding \fB-lpcre\fR to the command for linking an application which  .\"
140  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++
141  contain the major and minor release numbers for the library. Applications can  wrapper is also distributed with PCRE. It is documented in the
142  use these to include support for different releases.  .\" HREF
143    \fBpcrecpp\fP
144  The functions \fBpcre_compile()\fR, \fBpcre_study()\fR, and \fBpcre_exec()\fR  .\"
145  are used for compiling and matching regular expressions. A sample program that  page.
146  demonstrates the simplest way of using them is given in the file  .P
147  \fIpcredemo.c\fR. The \fBpcresample\fR documentation describes how to run it.  The native API C function prototypes are defined in the header file
148    \fBpcre.h\fP, and on Unix systems the library itself is called \fBlibpcre\fP.
149  There are convenience functions for extracting captured substrings from a  It can normally be accessed by adding \fB-lpcre\fP to the command for linking
150  matched subject string. They are:  an application that uses PCRE. The header file defines the macros PCRE_MAJOR
151    and PCRE_MINOR to contain the major and minor release numbers for the library.
152    \fBpcre_copy_substring()\fR  Applications can use these to include support for different releases of PCRE.
153    \fBpcre_copy_named_substring()\fR  .P
154    \fBpcre_get_substring()\fR  In a Windows environment, if you want to statically link an application program
155    \fBpcre_get_named_substring()\fR  against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
156    \fBpcre_get_substring_list()\fR  including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
157    \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
158  \fBpcre_free_substring()\fR and \fBpcre_free_substring_list()\fR are also  \fB__declspec(dllimport)\fP, with unwanted results.
159    .P
160    The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
161    and \fBpcre_exec()\fP are used for compiling and matching regular expressions
162    in a Perl-compatible manner. A sample program that demonstrates the simplest
163    way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
164    source distribution. A listing of this program is given in the
165    .\" HREF
166    \fBpcredemo\fP
167    .\"
168    documentation, and the
169    .\" HREF
170    \fBpcresample\fP
171    .\"
172    documentation describes how to compile and run it.
173    .P
174    Just-in-time compiler support is an optional feature of PCRE that can be built
175    in appropriate hardware environments. It greatly speeds up the matching
176    performance of many patterns. Simple programs can easily request that it be
177    used if available, by setting an option that is ignored when it is not
178    relevant. More complicated programs might need to make use of the functions
179    \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
180    \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
181    These functions are discussed in the
182    .\" HREF
183    \fBpcrejit\fP
184    .\"
185    documentation.
186    .P
187    A second matching function, \fBpcre_dfa_exec()\fP, which is not
188    Perl-compatible, is also provided. This uses a different algorithm for the
189    matching. The alternative algorithm finds all possible matches (at a given
190    point in the subject), and scans the subject just once (unless there are
191    lookbehind assertions). However, this algorithm does not return captured
192    substrings. A description of the two matching algorithms and their advantages
193    and disadvantages is given in the
194    .\" HREF
195    \fBpcrematching\fP
196    .\"
197    documentation.
198    .P
199    In addition to the main compiling and matching functions, there are convenience
200    functions for extracting captured substrings from a subject string that is
201    matched by \fBpcre_exec()\fP. They are:
202    .sp
203      \fBpcre_copy_substring()\fP
204      \fBpcre_copy_named_substring()\fP
205      \fBpcre_get_substring()\fP
206      \fBpcre_get_named_substring()\fP
207      \fBpcre_get_substring_list()\fP
208      \fBpcre_get_stringnumber()\fP
209      \fBpcre_get_stringtable_entries()\fP
210    .sp
211    \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
212  provided, to free the memory used for extracted strings.  provided, to free the memory used for extracted strings.
213    .P
214  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
215  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,
216    or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
217  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
218  compiled pattern; \fBpcre_info()\fR is an obsolete version which returns only  internal tables that are generated when PCRE is built are used.
219    .P
220    The function \fBpcre_fullinfo()\fP is used to find out information about a
221    compiled pattern; \fBpcre_info()\fP is an obsolete version that returns only
222  some of the available information, but is retained for backwards compatibility.  some of the available information, but is retained for backwards compatibility.
223  The function \fBpcre_version()\fR returns a pointer to a string containing the  The function \fBpcre_version()\fP returns a pointer to a string containing the
224  version of PCRE and its date of release.  version of PCRE and its date of release.
225    .P
226  The global variables \fBpcre_malloc\fR and \fBpcre_free\fR initially contain  The function \fBpcre_refcount()\fP maintains a reference count in a data block
227  the entry points of the standard \fBmalloc()\fR and \fBfree()\fR functions  containing a compiled pattern. This is provided for the benefit of
228    object-oriented applications.
229    .P
230    The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
231    the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
232  respectively. PCRE calls the memory management functions via these variables,  respectively. PCRE calls the memory management functions via these variables,
233  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
234  should be done before calling any PCRE functions.  should be done before calling any PCRE functions.
235    .P
236  The global variables \fBpcre_stack_malloc\fR and \fBpcre_stack_free\fR are also  The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
237  indirections to memory management functions. These special functions are used  indirections to memory management functions. These special functions are used
238  only when PCRE is compiled to use the heap for remembering data, instead of  only when PCRE is compiled to use the heap for remembering data, instead of
239  recursive function calls. This is a non-standard way of building PCRE, for use  recursive function calls, when running the \fBpcre_exec()\fP function. See the
240  in environments that have limited stacks. Because of the greater use of memory  .\" HREF
241  management, it runs more slowly. Separate functions are provided so that  \fBpcrebuild\fP
242  special-purpose external code can be used for this case. When used, these  .\"
243  functions are always called in a stack-like manner (last obtained, first  documentation for details of how to do this. It is a non-standard way of
244  freed), and always for memory blocks of the same size.  building PCRE, for use in environments that have limited stacks. Because of the
245    greater use of memory management, it runs more slowly. Separate functions are
246  The global variable \fBpcre_callout\fR initially contains NULL. It can be set  provided so that special-purpose external code can be used for this case. When
247    used, these functions are always called in a stack-like manner (last obtained,
248    first freed), and always for memory blocks of the same size. There is a
249    discussion about PCRE's stack usage in the
250    .\" HREF
251    \fBpcrestack\fP
252    .\"
253    documentation.
254    .P
255    The global variable \fBpcre_callout\fP initially contains NULL. It can be set
256  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
257  points during a matching operation. Details are given in the \fBpcrecallout\fR  points during a matching operation. Details are given in the
258    .\" HREF
259    \fBpcrecallout\fP
260    .\"
261  documentation.  documentation.
262    .
263    .
264    .\" HTML <a name="newlines"></a>
265    .SH NEWLINES
266    .rs
267    .sp
268    PCRE supports five different conventions for indicating line breaks in
269    strings: a single CR (carriage return) character, a single LF (linefeed)
270    character, the two-character sequence CRLF, any of the three preceding, or any
271    Unicode newline sequence. The Unicode newline sequences are the three just
272    mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
273    U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
274    (paragraph separator, U+2029).
275    .P
276    Each of the first three conventions is used by at least one operating system as
277    its standard newline sequence. When PCRE is built, a default can be specified.
278    The default default is LF, which is the Unix standard. When PCRE is run, the
279    default can be overridden, either when a pattern is compiled, or when it is
280    matched.
281    .P
282    At compile time, the newline convention can be specified by the \fIoptions\fP
283    argument of \fBpcre_compile()\fP, or it can be specified by special text at the
284    start of the pattern itself; this overrides any other settings. See the
285    .\" HREF
286    \fBpcrepattern\fP
287    .\"
288    page for details of the special character sequences.
289    .P
290    In the PCRE documentation the word "newline" is used to mean "the character or
291    pair of characters that indicate a line break". The choice of newline
292    convention affects the handling of the dot, circumflex, and dollar
293    metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
294    recognized line ending sequence, the match position advancement for a
295    non-anchored pattern. There is more detail about this in the
296    .\" HTML <a href="#execoptions">
297    .\" </a>
298    section on \fBpcre_exec()\fP options
299    .\"
300    below.
301    .P
302    The choice of newline convention does not affect the interpretation of
303    the \en or \er escape sequences, nor does it affect what \eR matches, which is
304    controlled in a similar way, but by separate options.
305    .
306    .
307  .SH MULTITHREADING  .SH MULTITHREADING
308  .rs  .rs
309  .sp  .sp
310  The PCRE functions can be used in multi-threading applications, with the  The PCRE functions can be used in multi-threading applications, with the
311  proviso that the memory management functions pointed to by \fBpcre_malloc\fR,  proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
312  \fBpcre_free\fR, \fBpcre_stack_malloc\fR, and \fBpcre_stack_free\fR, and the  \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
313  callout function pointed to by \fBpcre_callout\fR, are shared by all threads.  callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
314    .P
315  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
316  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.
317    .P
318  .SH CHECKING BUILD-TIME OPTIONS  If the just-in-time optimization feature is being used, it needs separate
319    memory stack areas for each thread. See the
320    .\" HREF
321    \fBpcrejit\fP
322    .\"
323    documentation for more details.
324    .
325    .
326    .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
327  .rs  .rs
328  .sp  .sp
329  .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
330    time, possibly by a different program, and even on a host other than the one on
331    which it was compiled. Details are given in the
332    .\" HREF
333    \fBpcreprecompile\fP
334    .\"
335    documentation. However, compiling a regular expression with one version of PCRE
336    for use with a different version is not guaranteed to work and may cause
337    crashes.
338    .
339    .
340    .SH "CHECKING BUILD-TIME OPTIONS"
341    .rs
342    .sp
343    .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
344  .PP  .PP
345  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
346  discover which optional features have been compiled into the PCRE library. The  discover which optional features have been compiled into the PCRE library. The
347  .\" HREF  .\" HREF
348  \fBpcrebuild\fR  \fBpcrebuild\fP
349  .\"  .\"
350  documentation has more details about these optional features.  documentation has more details about these optional features.
351    .P
352  The first argument for \fBpcre_config()\fR is an integer, specifying which  The first argument for \fBpcre_config()\fP is an integer, specifying which
353  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
354  which the information is placed. The following information is available:  which the information is placed. The following information is available:
355    .sp
356    PCRE_CONFIG_UTF8    PCRE_CONFIG_UTF8
357    .sp
358  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;
359  otherwise it is set to zero.  otherwise it is set to zero.
360    .sp
361      PCRE_CONFIG_UNICODE_PROPERTIES
362    .sp
363    The output is an integer that is set to one if support for Unicode character
364    properties is available; otherwise it is set to zero.
365    .sp
366      PCRE_CONFIG_JIT
367    .sp
368    The output is an integer that is set to one if support for just-in-time
369    compiling is available; otherwise it is set to zero.
370    .sp
371    PCRE_CONFIG_NEWLINE    PCRE_CONFIG_NEWLINE
372    .sp
373  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
374  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
375  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.
376    Though they are derived from ASCII, the same values are returned in EBCDIC
377    environments. The default should normally correspond to the standard sequence
378    for your operating system.
379    .sp
380      PCRE_CONFIG_BSR
381    .sp
382    The output is an integer whose value indicates what character sequences the \eR
383    escape sequence matches by default. A value of 0 means that \eR matches any
384    Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
385    or CRLF. The default can be overridden when a pattern is compiled or matched.
386    .sp
387    PCRE_CONFIG_LINK_SIZE    PCRE_CONFIG_LINK_SIZE
388    .sp
389  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
390  linkage in compiled regular expressions. The value is 2, 3, or 4. Larger values  linkage in compiled regular expressions. The value is 2, 3, or 4. Larger values
391  allow larger regular expressions to be compiled, at the expense of slower  allow larger regular expressions to be compiled, at the expense of slower
392  matching. The default value of 2 is sufficient for all but the most massive  matching. The default value of 2 is sufficient for all but the most massive
393  patterns, since it allows the compiled pattern to be up to 64K in size.  patterns, since it allows the compiled pattern to be up to 64K in size.
394    .sp
395    PCRE_CONFIG_POSIX_MALLOC_THRESHOLD    PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
396    .sp
397  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
398  interface uses \fBmalloc()\fR for output vectors. Further details are given in  interface uses \fBmalloc()\fP for output vectors. Further details are given in
399  the \fBpcreposix\fR documentation.  the
400    .\" HREF
401    \fBpcreposix\fP
402    .\"
403    documentation.
404    .sp
405    PCRE_CONFIG_MATCH_LIMIT    PCRE_CONFIG_MATCH_LIMIT
406    .sp
407  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
408  internal matching function calls in a \fBpcre_exec()\fR execution. Further  internal matching function calls in a \fBpcre_exec()\fP execution. Further
409  details are given with \fBpcre_exec()\fR below.  details are given with \fBpcre_exec()\fP below.
410    .sp
411      PCRE_CONFIG_MATCH_LIMIT_RECURSION
412    .sp
413    The output is a long integer that gives the default limit for the depth of
414    recursion when calling the internal matching function in a \fBpcre_exec()\fP
415    execution. Further details are given with \fBpcre_exec()\fP below.
416    .sp
417    PCRE_CONFIG_STACKRECURSE    PCRE_CONFIG_STACKRECURSE
418    .sp
419  The output is an integer that is set to one if internal recursion is  The output is an integer that is set to one if internal recursion when running
420  implemented by recursive function calls that use the stack to remember their  \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
421  state. This is the usual way that PCRE is compiled. The output is zero if PCRE  to remember their state. This is the usual way that PCRE is compiled. The
422  was compiled to use blocks of data on the heap instead of recursive function  output is zero if PCRE was compiled to use blocks of data on the heap instead
423  calls. In this case, \fBpcre_stack_malloc\fR and \fBpcre_stack_free\fR are  of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
424  called to manage memory blocks on the heap, thus avoiding the use of the stack.  \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
425    avoiding the use of the stack.
426  .SH COMPILING A PATTERN  .
427  .rs  .
428  .sp  .SH "COMPILING A PATTERN"
429  .B pcre *pcre_compile(const char *\fIpattern\fR, int \fIoptions\fR,  .rs
430  .ti +5n  .sp
431  .B const char **\fIerrptr\fR, int *\fIerroffset\fR,  .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
432  .ti +5n  .ti +5n
433  .B const unsigned char *\fItableptr\fR);  .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
434  .PP  .ti +5n
435    .B const unsigned char *\fItableptr\fP);
436  The function \fBpcre_compile()\fR is called to compile a pattern into an  .sp
437  internal form. The pattern is a C string terminated by a binary zero, and  .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
438  is passed in the argument \fIpattern\fR. A pointer to a single block of memory  .ti +5n
439  that is obtained via \fBpcre_malloc\fR is returned. This contains the compiled  .B int *\fIerrorcodeptr\fP,
440  code and related data. The \fBpcre\fR type is defined for the returned block;  .ti +5n
441  this is a typedef for a structure whose contents are not externally defined. It  .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
442  is up to the caller to free the memory when it is no longer required.  .ti +5n
443    .B const unsigned char *\fItableptr\fP);
444    .P
445    Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
446    called to compile a pattern into an internal form. The only difference between
447    the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
448    \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
449    too much repetition, we refer just to \fBpcre_compile()\fP below, but the
450    information applies equally to \fBpcre_compile2()\fP.
451    .P
452    The pattern is a C string terminated by a binary zero, and is passed in the
453    \fIpattern\fP argument. A pointer to a single block of memory that is obtained
454    via \fBpcre_malloc\fP is returned. This contains the compiled code and related
455    data. The \fBpcre\fP type is defined for the returned block; this is a typedef
456    for a structure whose contents are not externally defined. It is up to the
457    caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
458    .P
459  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
460  depend on memory location, the complete \fBpcre\fR data block is not  depend on memory location, the complete \fBpcre\fP data block is not
461  fully relocatable, because it contains a copy of the \fItableptr\fR argument,  fully relocatable, because it may contain a copy of the \fItableptr\fP
462  which is an address (see below).  argument, which is an address (see below).
463    .P
464  The \fIoptions\fR argument contains independent bits that affect the  The \fIoptions\fP argument contains various bit settings that affect the
465  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
466  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
467  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
468  in the \fBpcrepattern\fR documentation). For these options, the contents of the  within the pattern (see the detailed description in the
469  \fIoptions\fR argument specifies their initial settings at the start of  .\" HREF
470  compilation and execution. The PCRE_ANCHORED option can be set at the time of  \fBpcrepattern\fP
471  matching as well as at compile time.  .\"
472    documentation). For those options that can be different in different parts of
473  If \fIerrptr\fR is NULL, \fBpcre_compile()\fR returns NULL immediately.  the pattern, the contents of the \fIoptions\fP argument specifies their
474  Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fR returns  settings at the start of compilation and execution. The PCRE_ANCHORED,
475  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
476  error message. The offset from the start of the pattern to the character where  PCRE_NO_START_OPT options can be set at the time of matching as well as at
477  the error was discovered is placed in the variable pointed to by  compile time.
478  \fIerroffset\fR, which must not be NULL. If it is, an immediate error is given.  .P
479    If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
480  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
481  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
482  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
483  \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
484    byte that was being processed when the error was discovered is placed in the
485  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
486    immediate error is given). However, for an invalid UTF-8 string, the offset is
487    that of the first byte of the failing character. Also, some errors are not
488    detected until checks are carried out when the whole pattern has been scanned;
489    in these cases the offset passed back is the length of the pattern.
490    .P
491    Note that the offset is in bytes, not characters, even in UTF-8 mode. It may
492    sometimes point into the middle of a UTF-8 character.
493    .P
494    If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
495    \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
496    returned via this argument in the event of an error. This is in addition to the
497    textual error message. Error codes and messages are listed below.
498    .P
499    If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
500    character tables that are built when PCRE is compiled, using the default C
501    locale. Otherwise, \fItableptr\fP must be an address that is the result of a
502    call to \fBpcre_maketables()\fP. This value is stored with the compiled
503    pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
504    passed to it. For more discussion, see the section on locale support below.
505    .P
506    This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
507    .sp
508    pcre *re;    pcre *re;
509    const char *error;    const char *error;
510    int erroffset;    int erroffset;
# Line 290  This code fragment shows a typical strai Line 514  This code fragment shows a typical strai
514      &error,           /* for error message */      &error,           /* for error message */
515      &erroffset,       /* for error offset */      &erroffset,       /* for error offset */
516      NULL);            /* use default character tables */      NULL);            /* use default character tables */
517    .sp
518  The following option bits are defined:  The following names for option bits are defined in the \fBpcre.h\fP header
519    file:
520    .sp
521    PCRE_ANCHORED    PCRE_ANCHORED
522    .sp
523  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
524  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
525  being searched (the "subject string"). This effect can also be achieved by  being searched (the "subject string"). This effect can also be achieved by
526  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
527  Perl.  Perl.
528    .sp
529      PCRE_AUTO_CALLOUT
530    .sp
531    If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
532    all with number 255, before each pattern item. For discussion of the callout
533    facility, see the
534    .\" HREF
535    \fBpcrecallout\fP
536    .\"
537    documentation.
538    .sp
539      PCRE_BSR_ANYCRLF
540      PCRE_BSR_UNICODE
541    .sp
542    These options (which are mutually exclusive) control what the \eR escape
543    sequence matches. The choice is either to match only CR, LF, or CRLF, or to
544    match any Unicode newline sequence. The default is specified when PCRE is
545    built. It can be overridden from within the pattern, or by setting an option
546    when a compiled pattern is matched.
547    .sp
548    PCRE_CASELESS    PCRE_CASELESS
549    .sp
550  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
551  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
552  pattern by a (?i) option setting.  pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
553    concept of case for characters whose values are less than 128, so caseless
554    matching is always possible. For characters with higher values, the concept of
555    case is supported if PCRE is compiled with Unicode property support, but not
556    otherwise. If you want to use caseless matching for characters 128 and above,
557    you must ensure that PCRE is compiled with Unicode property support as well as
558    with UTF-8 support.
559    .sp
560    PCRE_DOLLAR_ENDONLY    PCRE_DOLLAR_ENDONLY
561    .sp
562  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
563  end of the subject string. Without this option, a dollar also matches  end of the subject string. Without this option, a dollar also matches
564  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
565  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.
566  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
567  a pattern.  pattern.
568    .sp
569    PCRE_DOTALL    PCRE_DOTALL
570    .sp
571  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
572  including newlines. Without it, newlines are excluded. This option is  any value, including one that indicates a newline. However, it only ever
573    matches one character, even if newlines are coded as CRLF. Without this option,
574    a dot does not match when the current position is at a newline. This option is
575  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
576  (?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
577  character, independent of the setting of this option.  characters, independent of the setting of this option.
578    .sp
579      PCRE_DUPNAMES
580    .sp
581    If this bit is set, names used to identify capturing subpatterns need not be
582    unique. This can be helpful for certain types of pattern when it is known that
583    only one instance of the named subpattern can ever be matched. There are more
584    details of named subpatterns below; see also the
585    .\" HREF
586    \fBpcrepattern\fP
587    .\"
588    documentation.
589    .sp
590    PCRE_EXTENDED    PCRE_EXTENDED
591    .sp
592  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
593  ignored except when escaped or inside a character class. Whitespace does not  ignored except when escaped or inside a character class. Whitespace does not
594  include the VT character (code 11). In addition, characters between an  include the VT character (code 11). In addition, characters between an
595  unescaped # outside a character class and the next newline character,  unescaped # outside a character class and the next newline, inclusive, are also
596  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
597  be changed within a pattern by a (?x) option setting.  pattern by a (?x) option setting.
598    .P
599    Which characters are interpreted as newlines is controlled by the options
600    passed to \fBpcre_compile()\fP or by a special sequence at the start of the
601    pattern, as described in the section entitled
602    .\" HTML <a href="pcrepattern.html#newlines">
603    .\" </a>
604    "Newline conventions"
605    .\"
606    in the \fBpcrepattern\fP documentation. Note that the end of this type of
607    comment is a literal newline sequence in the pattern; escape sequences that
608    happen to represent a newline do not count.
609    .P
610  This option makes it possible to include comments inside complicated patterns.  This option makes it possible to include comments inside complicated patterns.
611  Note, however, that this applies only to data characters. Whitespace characters  Note, however, that this applies only to data characters. Whitespace characters
612  may never appear within special character sequences in a pattern, for example  may never appear within special character sequences in a pattern, for example
613  within the sequence (?( which introduces a conditional subpattern.  within the sequence (?( that introduces a conditional subpattern.
614    .sp
615    PCRE_EXTRA    PCRE_EXTRA
616    .sp
617  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
618  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
619  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
620  special meaning causes an error, thus reserving these combinations for future  special meaning causes an error, thus reserving these combinations for future
621  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
622  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
623  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
624  pattern.  no other features controlled by this option. It can also be set by a (?X)
625    option setting within a pattern.
626    .sp
627      PCRE_FIRSTLINE
628    .sp
629    If this option is set, an unanchored pattern is required to match before or at
630    the first newline in the subject string, though the matched text may continue
631    over the newline.
632    .sp
633      PCRE_JAVASCRIPT_COMPAT
634    .sp
635    If this option is set, PCRE's behaviour is changed in some ways so that it is
636    compatible with JavaScript rather than Perl. The changes are as follows:
637    .P
638    (1) A lone closing square bracket in a pattern causes a compile-time error,
639    because this is illegal in JavaScript (by default it is treated as a data
640    character). Thus, the pattern AB]CD becomes illegal when this option is set.
641    .P
642    (2) At run time, a back reference to an unset subpattern group matches an empty
643    string (by default this causes the current matching alternative to fail). A
644    pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
645    an "a" in the subject), whereas it fails by default, for Perl compatibility.
646    .P
647    (3) \eU matches an upper case "U" character; by default \eU causes a compile
648    time error (Perl uses \eU to upper case subsequent characters).
649    .P
650    (4) \eu matches a lower case "u" character unless it is followed by four
651    hexadecimal digits, in which case the hexadecimal number defines the code point
652    to match. By default, \eu causes a compile time error (Perl uses it to upper
653    case the following character).
654    .P
655    (5) \ex matches a lower case "x" character unless it is followed by two
656    hexadecimal digits, in which case the hexadecimal number defines the code point
657    to match. By default, as in Perl, a hexadecimal number is always expected after
658    \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
659    binary zero character followed by z).
660    .sp
661    PCRE_MULTILINE    PCRE_MULTILINE
662    .sp
663  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
664  characters (even if it actually contains several newlines). The "start of line"  characters (even if it actually contains newlines). The "start of line"
665  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
666  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
667  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
668  Perl.  Perl.
669    .P
670  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
671  match immediately following or immediately before any newline in the subject  match immediately following or immediately before internal newlines in the
672  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
673  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
674  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
675  occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.  occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
676    .sp
677      PCRE_NEWLINE_CR
678      PCRE_NEWLINE_LF
679      PCRE_NEWLINE_CRLF
680      PCRE_NEWLINE_ANYCRLF
681      PCRE_NEWLINE_ANY
682    .sp
683    These options override the default newline definition that was chosen when PCRE
684    was built. Setting the first or the second specifies that a newline is
685    indicated by a single character (CR or LF, respectively). Setting
686    PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
687    CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
688    preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
689    that any Unicode newline sequence should be recognized. The Unicode newline
690    sequences are the three just mentioned, plus the single characters VT (vertical
691    tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
692    separator, U+2028), and PS (paragraph separator, U+2029). The last two are
693    recognized only in UTF-8 mode.
694    .P
695    The newline setting in the options word uses three bits that are treated
696    as a number, giving eight possibilities. Currently only six are used (default
697    plus the five values above). This means that if you set more than one newline
698    option, the combination may or may not be sensible. For example,
699    PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
700    other combinations may yield unused numbers and cause an error.
701    .P
702    The only time that a line break in a pattern is specially recognized when
703    compiling is when PCRE_EXTENDED is set. CR and LF are whitespace characters,
704    and so are ignored in this mode. Also, an unescaped # outside a character class
705    indicates a comment that lasts until after the next line break sequence. In
706    other circumstances, line break sequences in patterns are treated as literal
707    data.
708    .P
709    The newline option that is set at compile time becomes the default that is used
710    for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
711    .sp
712    PCRE_NO_AUTO_CAPTURE    PCRE_NO_AUTO_CAPTURE
713    .sp
714  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
715  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
716  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
717  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
718  in Perl.  in Perl.
719    .sp
720      NO_START_OPTIMIZE
721    .sp
722    This is an option that acts at matching time; that is, it is really an option
723    for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
724    it is remembered with the compiled pattern and assumed at matching time. For
725    details see the discussion of PCRE_NO_START_OPTIMIZE
726    .\" HTML <a href="#execoptions">
727    .\" </a>
728    below.
729    .\"
730    .sp
731      PCRE_UCP
732    .sp
733    This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
734    \ew, and some of the POSIX character classes. By default, only ASCII characters
735    are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
736    classify characters. More details are given in the section on
737    .\" HTML <a href="pcre.html#genericchartypes">
738    .\" </a>
739    generic character types
740    .\"
741    in the
742    .\" HREF
743    \fBpcrepattern\fP
744    .\"
745    page. If you set PCRE_UCP, matching one of the items it affects takes much
746    longer. The option is available only if PCRE has been compiled with Unicode
747    property support.
748    .sp
749    PCRE_UNGREEDY    PCRE_UNGREEDY
750    .sp
751  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
752  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
753  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.
754    .sp
755    PCRE_UTF8    PCRE_UTF8
756    .sp
757  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
758  of UTF-8 characters instead of single-byte character strings. However, it is  of UTF-8 characters instead of single-byte character strings. However, it is
759  available only if PCRE has been built to include UTF-8 support. If not, the use  available only when PCRE is built to include UTF-8 support. If not, the use
760  of this option provokes an error. Details of how this option changes the  of this option provokes an error. Details of how this option changes the
761  behaviour of PCRE are given in the  behaviour of PCRE are given in the
 .\" HTML <a href="pcre.html#utf8support">  
 .\" </a>  
 section on UTF-8 support  
 .\"  
 in the main  
762  .\" HREF  .\" HREF
763  \fBpcre\fR  \fBpcreunicode\fP
764  .\"  .\"
765  page.  page.
766    .sp
767    PCRE_NO_UTF8_CHECK    PCRE_NO_UTF8_CHECK
768    .sp
769  When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is  When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
770  automatically checked. If an invalid UTF-8 sequence of bytes is found,  automatically checked. There is a discussion about the
771  \fBpcre_compile()\fR returns an error. If you already know that your pattern is  .\" HTML <a href="pcre.html#utf8strings">
772  valid, and you want to skip this check for performance reasons, you can set the  .\" </a>
773  PCRE_NO_UTF8_CHECK option. When it is set, the effect of passing an invalid  validity of UTF-8 strings
774  UTF-8 string as a pattern is undefined. It may cause your program to crash.  .\"
775  Note that there is a similar option for suppressing the checking of subject  in the main
776  strings passed to \fBpcre_exec()\fR.  .\" HREF
777    \fBpcre\fP
778    .\"
779  .SH STUDYING A PATTERN  page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_compile()\fP
780  .rs  returns an error. If you already know that your pattern is valid, and you want
781  .sp  to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
782  .B pcre_extra *pcre_study(const pcre *\fIcode\fR, int \fIoptions\fR,  option. When it is set, the effect of passing an invalid UTF-8 string as a
783  .ti +5n  pattern is undefined. It may cause your program to crash. Note that this option
784  .B const char **\fIerrptr\fR);  can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
785  .PP  the UTF-8 validity checking of subject strings.
786  When a pattern is going to be used several times, it is worth spending more  .
787  time analyzing it in order to speed up the time taken for matching. The  .
788  function \fBpcre_study()\fR takes a pointer to a compiled pattern as its first  .SH "COMPILATION ERROR CODES"
789  argument. If studing the pattern produces additional information that will help  .rs
790  speed up matching, \fBpcre_study()\fR returns a pointer to a \fBpcre_extra\fR  .sp
791  block, in which the \fIstudy_data\fR field points to the results of the study.  The following table lists the error codes than may be returned by
792    \fBpcre_compile2()\fP, along with the error messages that may be returned by
793  The returned value from a \fBpcre_study()\fR can be passed directly to  both compiling functions. As PCRE has developed, some error codes have fallen
794  \fBpcre_exec()\fR. However, the \fBpcre_extra\fR block also contains other  out of use. To avoid confusion, they have not been re-used.
795  fields that can be set by the caller before the block is passed; these are  .sp
796  described below. If studying the pattern does not produce any additional     0  no error
797  information, \fBpcre_study()\fR returns NULL. In that circumstance, if the     1  \e at end of pattern
798  calling program wants to pass some of the other fields to \fBpcre_exec()\fR, it     2  \ec at end of pattern
799  must set up its own \fBpcre_extra\fR block.     3  unrecognized character follows \e
800       4  numbers out of order in {} quantifier
801  The second argument contains option bits. At present, no options are defined     5  number too big in {} quantifier
802  for \fBpcre_study()\fR, and this argument should always be zero.     6  missing terminating ] for character class
803       7  invalid escape sequence in character class
804  The third argument for \fBpcre_study()\fR is a pointer for an error message. If     8  range out of order in character class
805       9  nothing to repeat
806      10  [this code is not in use]
807      11  internal error: unexpected repeat
808      12  unrecognized character after (? or (?-
809      13  POSIX named classes are supported only within a class
810      14  missing )
811      15  reference to non-existent subpattern
812      16  erroffset passed as NULL
813      17  unknown option bit(s) set
814      18  missing ) after comment
815      19  [this code is not in use]
816      20  regular expression is too large
817      21  failed to get memory
818      22  unmatched parentheses
819      23  internal error: code overflow
820      24  unrecognized character after (?<
821      25  lookbehind assertion is not fixed length
822      26  malformed number or name after (?(
823      27  conditional group contains more than two branches
824      28  assertion expected after (?(
825      29  (?R or (?[+-]digits must be followed by )
826      30  unknown POSIX class name
827      31  POSIX collating elements are not supported
828      32  this version of PCRE is not compiled with PCRE_UTF8 support
829      33  [this code is not in use]
830      34  character value in \ex{...} sequence is too large
831      35  invalid condition (?(0)
832      36  \eC not allowed in lookbehind assertion
833      37  PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
834      38  number after (?C is > 255
835      39  closing ) for (?C expected
836      40  recursive call could loop indefinitely
837      41  unrecognized character after (?P
838      42  syntax error in subpattern name (missing terminator)
839      43  two named subpatterns have the same name
840      44  invalid UTF-8 string
841      45  support for \eP, \ep, and \eX has not been compiled
842      46  malformed \eP or \ep sequence
843      47  unknown property name after \eP or \ep
844      48  subpattern name is too long (maximum 32 characters)
845      49  too many named subpatterns (maximum 10000)
846      50  [this code is not in use]
847      51  octal value is greater than \e377 (not in UTF-8 mode)
848      52  internal error: overran compiling workspace
849      53  internal error: previously-checked referenced subpattern
850            not found
851      54  DEFINE group contains more than one branch
852      55  repeating a DEFINE group is not allowed
853      56  inconsistent NEWLINE options
854      57  \eg is not followed by a braced, angle-bracketed, or quoted
855            name/number or by a plain number
856      58  a numbered reference must not be zero
857      59  an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
858      60  (*VERB) not recognized
859      61  number is too big
860      62  subpattern name expected
861      63  digit expected after (?+
862      64  ] is an invalid data character in JavaScript compatibility mode
863      65  different names for subpatterns of the same number are
864            not allowed
865      66  (*MARK) must have an argument
866      67  this version of PCRE is not compiled with PCRE_UCP support
867      68  \ec must be followed by an ASCII character
868      69  \ek is not followed by a braced, angle-bracketed, or quoted name
869    .sp
870    The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
871    be used if the limits were changed when PCRE was built.
872    .
873    .
874    .\" HTML <a name="studyingapattern"></a>
875    .SH "STUDYING A PATTERN"
876    .rs
877    .sp
878    .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
879    .ti +5n
880    .B const char **\fIerrptr\fP);
881    .PP
882    If a compiled pattern is going to be used several times, it is worth spending
883    more time analyzing it in order to speed up the time taken for matching. The
884    function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
885    argument. If studying the pattern produces additional information that will
886    help speed up matching, \fBpcre_study()\fP returns a pointer to a
887    \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
888    results of the study.
889    .P
890    The returned value from \fBpcre_study()\fP can be passed directly to
891    \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
892    also contains other fields that can be set by the caller before the block is
893    passed; these are described
894    .\" HTML <a href="#extradata">
895    .\" </a>
896    below
897    .\"
898    in the section on matching a pattern.
899    .P
900    If studying the pattern does not produce any useful information,
901    \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program
902    wants to pass any of the other fields to \fBpcre_exec()\fP or
903    \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block.
904    .P
905    The second argument of \fBpcre_study()\fP contains option bits. There is only
906    one option: PCRE_STUDY_JIT_COMPILE. If this is set, and the just-in-time
907    compiler is available, the pattern is further compiled into machine code that
908    executes much faster than the \fBpcre_exec()\fP matching function. If
909    the just-in-time compiler is not available, this option is ignored. All other
910    bits in the \fIoptions\fP argument must be zero.
911    .P
912    JIT compilation is a heavyweight optimization. It can take some time for
913    patterns to be analyzed, and for one-off matches and simple patterns the
914    benefit of faster execution might be offset by a much slower study time.
915    Not all patterns can be optimized by the JIT compiler. For those that cannot be
916    handled, matching automatically falls back to the \fBpcre_exec()\fP
917    interpreter. For more details, see the
918    .\" HREF
919    \fBpcrejit\fP
920    .\"
921    documentation.
922    .P
923    The third argument for \fBpcre_study()\fP is a pointer for an error message. If
924  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
925  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
926  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
927  be sure that it has run successfully.  should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
928    sure that it has run successfully.
929  This is a typical call to \fBpcre_study\fR():  .P
930    When you are finished with a pattern, you can free the memory used for the
931    pcre_extra *pe;  study data by calling \fBpcre_free_study()\fP. This function was added to the
932    pe = pcre_study(  API for release 8.20. For earlier versions, the memory could be freed with
933    \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
934    where PCRE_STUDY_JIT_COMPILE is not used, but it is advisable to change to the
935    new function when convenient.
936    .P
937    This is a typical way in which \fBpcre_study\fP() is used (except that in a
938    real application there should be tests for errors):
939    .sp
940      int rc;
941      pcre *re;
942      pcre_extra *sd;
943      re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
944      sd = pcre_study(
945      re,             /* result of pcre_compile() */      re,             /* result of pcre_compile() */
946      0,              /* no options exist */      0,              /* no options */
947      &error);        /* set to NULL or points to a message */      &error);        /* set to NULL or points to a message */
948      rc = pcre_exec(   /* see below for details of pcre_exec() options */
949  At present, studying a pattern is useful only for non-anchored patterns that do      re, sd, "subject", 7, 0, 0, ovector, 30);
950  not have a single fixed starting character. A bitmap of possible starting    ...
951  characters is created.    pcre_free_study(sd);
952      pcre_free(re);
953    .sp
954    Studying a pattern does two things: first, a lower bound for the length of
955    subject string that is needed to match the pattern is computed. This does not
956    mean that there are any strings of that length that match, but it does
957    guarantee that no shorter strings match. The value is used by
958    \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP to avoid wasting time by trying to
959    match strings that are shorter than the lower bound. You can find out the value
960    in a calling program via the \fBpcre_fullinfo()\fP function.
961    .P
962    Studying a pattern is also useful for non-anchored patterns that do not have a
963    single fixed starting character. A bitmap of possible starting bytes is
964    created. This speeds up finding a position in the subject at which to start
965    matching.
966    .P
967    These two optimizations apply to both \fBpcre_exec()\fP and
968    \fBpcre_dfa_exec()\fP. However, they are not used by \fBpcre_exec()\fP if
969    \fBpcre_study()\fP is called with the PCRE_STUDY_JIT_COMPILE option, and
970    just-in-time compiling is successful. The optimizations can be disabled by
971    setting the PCRE_NO_START_OPTIMIZE option when calling \fBpcre_exec()\fP or
972    \fBpcre_dfa_exec()\fP. You might want to do this if your pattern contains
973    callouts or (*MARK) (which cannot be handled by the JIT compiler), and you want
974    to make use of these facilities in cases where matching fails. See the
975    discussion of PCRE_NO_START_OPTIMIZE
976    .\" HTML <a href="#execoptions">
977    .\" </a>
978    below.
979    .\"
980    .
981    .
982  .\" HTML <a name="localesupport"></a>  .\" HTML <a name="localesupport"></a>
983  .SH LOCALE SUPPORT  .SH "LOCALE SUPPORT"
984  .rs  .rs
985  .sp  .sp
986  PCRE handles caseless matching, and determines whether characters are letters,  PCRE handles caseless matching, and determines whether characters are letters,
987  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
988  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 with codes
989  contains a default set of tables that is created in the default C locale when  less than 128. By default, higher-valued codes never match escapes such as \ew
990  PCRE is compiled. This is used when the final argument of \fBpcre_compile()\fR  or \ed, but they can be tested with \ep if PCRE is built with Unicode character
991  is NULL, and is sufficient for many applications.  property support. Alternatively, the PCRE_UCP option can be set at compile
992    time; this causes \ew and friends to use Unicode property support instead of
993  An alternative set of tables can, however, be supplied. Such tables are built  built-in tables. The use of locales with Unicode is discouraged. If you are
994  by calling the \fBpcre_maketables()\fR function, which has no arguments, in the  handling characters with codes greater than 128, you should either use UTF-8
995  relevant locale. The result can then be passed to \fBpcre_compile()\fR as often  and Unicode, or use locales, but not try to mix the two.
996  as necessary. For example, to build and use tables that are appropriate for the  .P
997  French locale (where accented characters with codes greater than 128 are  PCRE contains an internal set of tables that are used when the final argument
998  treated as letters), the following code could be used:  of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
999    Normally, the internal tables recognize only ASCII characters. However, when
1000    setlocale(LC_CTYPE, "fr");  PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1001    default "C" locale of the local system, which may cause them to be different.
1002    .P
1003    The internal tables can always be overridden by tables supplied by the
1004    application that calls PCRE. These may be created in a different locale from
1005    the default. As more and more applications change to using Unicode, the need
1006    for this locale support is expected to die away.
1007    .P
1008    External tables are built by calling the \fBpcre_maketables()\fP function,
1009    which has no arguments, in the relevant locale. The result can then be passed
1010    to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
1011    example, to build and use tables that are appropriate for the French locale
1012    (where accented characters with values greater than 128 are treated as letters),
1013    the following code could be used:
1014    .sp
1015      setlocale(LC_CTYPE, "fr_FR");
1016    tables = pcre_maketables();    tables = pcre_maketables();
1017    re = pcre_compile(..., tables);    re = pcre_compile(..., tables);
1018    .sp
1019  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
1020  pointer that is passed to \fBpcre_compile\fR is saved with the compiled  are using Windows, the name for the French locale is "french".
1021  pattern, and the same tables are used via this pointer by \fBpcre_study()\fR  .P
1022  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
1023  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
1024  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
1025  memory containing the tables remains available for as long as it is needed.  needed.
1026    .P
1027  .SH INFORMATION ABOUT A PATTERN  The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1028    pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1029    and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
1030    pattern, compilation, studying and matching all happen in the same locale, but
1031    different patterns can be compiled in different locales.
1032    .P
1033    It is possible to pass a table pointer or NULL (indicating the use of the
1034    internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
1035    this facility could be used to match a pattern in a different locale from the
1036    one in which it was compiled. Passing table pointers at run time is discussed
1037    below in the section on matching a pattern.
1038    .
1039    .
1040    .\" HTML <a name="infoaboutpattern"></a>
1041    .SH "INFORMATION ABOUT A PATTERN"
1042  .rs  .rs
1043  .sp  .sp
1044  .B int pcre_fullinfo(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"  .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1045  .ti +5n  .ti +5n
1046  .B int \fIwhat\fR, void *\fIwhere\fR);  .B int \fIwhat\fP, void *\fIwhere\fP);
1047  .PP  .PP
1048  The \fBpcre_fullinfo()\fR function returns information about a compiled  The \fBpcre_fullinfo()\fP function returns information about a compiled
1049  pattern. It replaces the obsolete \fBpcre_info()\fR function, which is  pattern. It replaces the obsolete \fBpcre_info()\fP function, which is
1050  nevertheless retained for backwards compability (and is documented below).  nevertheless retained for backwards compability (and is documented below).
1051    .P
1052  The first argument for \fBpcre_fullinfo()\fR is a pointer to the compiled  The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1053  pattern. The second argument is the result of \fBpcre_study()\fR, or NULL if  pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1054  the pattern was not studied. The third argument specifies which piece of  the pattern was not studied. The third argument specifies which piece of
1055  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
1056  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
1057  the following negative numbers:  the following negative numbers:
1058    .sp
1059    PCRE_ERROR_NULL       the argument \fIcode\fR was NULL    PCRE_ERROR_NULL       the argument \fIcode\fP was NULL
1060                          the argument \fIwhere\fR was NULL                          the argument \fIwhere\fP was NULL
1061    PCRE_ERROR_BADMAGIC   the "magic number" was not found    PCRE_ERROR_BADMAGIC   the "magic number" was not found
1062    PCRE_ERROR_BADOPTION  the value of \fIwhat\fR was invalid    PCRE_ERROR_BADOPTION  the value of \fIwhat\fP was invalid
1063    .sp
1064  Here is a typical call of \fBpcre_fullinfo()\fR, to obtain the length of the  The "magic number" is placed at the start of each compiled pattern as an simple
1065  compiled pattern:  check against passing an arbitrary memory pointer. Here is a typical call of
1066    \fBpcre_fullinfo()\fP, to obtain the length of the compiled pattern:
1067    .sp
1068    int rc;    int rc;
1069    unsigned long int length;    size_t length;
1070    rc = pcre_fullinfo(    rc = pcre_fullinfo(
1071      re,               /* result of pcre_compile() */      re,               /* result of pcre_compile() */
1072      pe,               /* result of pcre_study(), or NULL */      sd,               /* result of pcre_study(), or NULL */
1073      PCRE_INFO_SIZE,   /* what is required */      PCRE_INFO_SIZE,   /* what is required */
1074      &length);         /* where to put the data */      &length);         /* where to put the data */
1075    .sp
1076  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
1077  as follows:  as follows:
1078    .sp
1079    PCRE_INFO_BACKREFMAX    PCRE_INFO_BACKREFMAX
1080    .sp
1081  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
1082  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
1083  no back references.  no back references.
1084    .sp
1085    PCRE_INFO_CAPTURECOUNT    PCRE_INFO_CAPTURECOUNT
1086    .sp
1087  Return the number of capturing subpatterns in the pattern. The fourth argument  Return the number of capturing subpatterns in the pattern. The fourth argument
1088  should point to an \fbint\fR variable.  should point to an \fBint\fP variable.
1089    .sp
1090      PCRE_INFO_DEFAULT_TABLES
1091    .sp
1092    Return a pointer to the internal default character tables within PCRE. The
1093    fourth argument should point to an \fBunsigned char *\fP variable. This
1094    information call is provided for internal use by the \fBpcre_study()\fP
1095    function. External callers can cause PCRE to use its internal tables by passing
1096    a NULL table pointer.
1097    .sp
1098    PCRE_INFO_FIRSTBYTE    PCRE_INFO_FIRSTBYTE
1099    .sp
1100  Return information about the first byte of any matched string, for a  Return information about the first byte of any matched string, for a
1101  non-anchored pattern. (This option used to be called PCRE_INFO_FIRSTCHAR; the  non-anchored pattern. The fourth argument should point to an \fBint\fP
1102  old name is still recognized for backwards compatibility.)  variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
1103    still recognized for backwards compatibility.)
1104  If there is a fixed first byte, e.g. from a pattern such as (cat|cow|coyote),  .P
1105  it is returned in the integer pointed to by \fIwhere\fR. Otherwise, if either  If there is a fixed first byte, for example, from a pattern such as
1106    (cat|cow|coyote), its value is returned. Otherwise, if either
1107    .sp
1108  (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
1109  starts with "^", or  starts with "^", or
1110    .sp
1111  (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
1112  (if it were set, the pattern would be anchored),  (if it were set, the pattern would be anchored),
1113    .sp
1114  -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
1115  subject string or after any newline within the string. Otherwise -2 is  subject string or after any newline within the string. Otherwise -2 is
1116  returned. For anchored patterns, -2 is returned.  returned. For anchored patterns, -2 is returned.
1117    .sp
1118    PCRE_INFO_FIRSTTABLE    PCRE_INFO_FIRSTTABLE
1119    .sp
1120  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
1121  table indicating a fixed set of bytes for the first byte in any matching  table indicating a fixed set of bytes for the first byte in any matching
1122  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
1123  fourth argument should point to an \fBunsigned char *\fR variable.  fourth argument should point to an \fBunsigned char *\fP variable.
1124    .sp
1125      PCRE_INFO_HASCRORLF
1126    .sp
1127    Return 1 if the pattern contains any explicit matches for CR or LF characters,
1128    otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1129    explicit match is either a literal CR or LF character, or \er or \en.
1130    .sp
1131      PCRE_INFO_JCHANGED
1132    .sp
1133    Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1134    0. The fourth argument should point to an \fBint\fP variable. (?J) and
1135    (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1136    .sp
1137      PCRE_INFO_JIT
1138    .sp
1139    Return 1 if the pattern was studied with the PCRE_STUDY_JIT_COMPILE option, and
1140    just-in-time compiling was successful. The fourth argument should point to an
1141    \fBint\fP variable. A return value of 0 means that JIT support is not available
1142    in this version of PCRE, or that the pattern was not studied with the
1143    PCRE_STUDY_JIT_COMPILE option, or that the JIT compiler could not handle this
1144    particular pattern. See the
1145    .\" HREF
1146    \fBpcrejit\fP
1147    .\"
1148    documentation for details of what can and cannot be handled.
1149    .sp
1150    PCRE_INFO_LASTLITERAL    PCRE_INFO_LASTLITERAL
1151    .sp
1152  Return the value of the rightmost literal byte that must exist in any matched  Return the value of the rightmost literal byte that must exist in any matched
1153  string, other than at its start, if such a byte has been recorded. The fourth  string, other than at its start, if such a byte has been recorded. The fourth
1154  argument should point to an \fBint\fR variable. If there is no such byte, -1 is  argument should point to an \fBint\fP variable. If there is no such byte, -1 is
1155  returned. For anchored patterns, a last literal byte is recorded only if it  returned. For anchored patterns, a last literal byte is recorded only if it
1156  follows something of variable length. For example, for the pattern  follows something of variable length. For example, for the pattern
1157  /^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
1158  is -1.  is -1.
1159    .sp
1160      PCRE_INFO_MINLENGTH
1161    .sp
1162    If the pattern was studied and a minimum length for matching subject strings
1163    was computed, its value is returned. Otherwise the returned value is -1. The
1164    value is a number of characters, not bytes (this may be relevant in UTF-8
1165    mode). The fourth argument should point to an \fBint\fP variable. A
1166    non-negative value is a lower bound to the length of any matching string. There
1167    may not be any strings of that length that do actually match, but every string
1168    that does match is at least that long.
1169    .sp
1170    PCRE_INFO_NAMECOUNT    PCRE_INFO_NAMECOUNT
1171    PCRE_INFO_NAMEENTRYSIZE    PCRE_INFO_NAMEENTRYSIZE
1172    PCRE_INFO_NAMETABLE    PCRE_INFO_NAMETABLE
1173    .sp
1174  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
1175  names are just an additional way of identifying the parentheses, which still  names are just an additional way of identifying the parentheses, which still
1176  acquire a number. A caller that wants to extract data from a named subpattern  acquire numbers. Several convenience functions such as
1177  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
1178  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
1179  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
1180  table for the pattern.  output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1181    you need to use the name-to-number map, which is described by these three
1182    values.
1183    .P
1184  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
1185  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
1186  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
1187  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
1188  entry of the table (a pointer to \fBchar\fR). The first two bytes of each entry  entry of the table (a pointer to \fBchar\fP). The first two bytes of each entry
1189  are the number of the capturing parenthesis, most significant byte first. The  are the number of the capturing parenthesis, most significant byte first. The
1190  rest of the entry is the corresponding name, zero terminated. The names are in  rest of the entry is the corresponding name, zero terminated.
1191  alphabetical order. For example, consider the following pattern (assume  .P
1192  PCRE_EXTENDED is set, so white space - including newlines - is ignored):  The names are in alphabetical order. Duplicate names may appear if (?| is used
1193    to create multiple groups with the same number, as described in the
1194    (?P<date> (?P<year>(\\d\\d)?\\d\\d) -  .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1195    (?P<month>\\d\\d) - (?P<day>\\d\\d) )  .\" </a>
1196    section on duplicate subpattern numbers
1197    .\"
1198    in the
1199    .\" HREF
1200    \fBpcrepattern\fP
1201    .\"
1202    page. Duplicate names for subpatterns with different numbers are permitted only
1203    if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1204    table in the order in which they were found in the pattern. In the absence of
1205    (?| this is the order of increasing number; when (?| is used this is not
1206    necessarily the case because later subpatterns may have lower numbers.
1207    .P
1208    As a simple example of the name/number table, consider the following pattern
1209    (assume PCRE_EXTENDED is set, so white space - including newlines - is
1210    ignored):
1211    .sp
1212    .\" JOIN
1213      (?<date> (?<year>(\ed\ed)?\ed\ed) -
1214      (?<month>\ed\ed) - (?<day>\ed\ed) )
1215    .sp
1216  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
1217  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
1218  bytes shows in hex, and undefined bytes shown as ??:  bytes shows in hexadecimal, and undefined bytes shown as ??:
1219    .sp
1220    00 01 d  a  t  e  00 ??    00 01 d  a  t  e  00 ??
1221    00 05 d  a  y  00 ?? ??    00 05 d  a  y  00 ?? ??
1222    00 04 m  o  n  t  h  00    00 04 m  o  n  t  h  00
1223    00 02 y  e  a  r  00 ??    00 02 y  e  a  r  00 ??
1224    .sp
1225  When writing code to extract data from named subpatterns, remember that the  When writing code to extract data from named subpatterns using the
1226  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
1227    different for each compiled pattern.
1228    .sp
1229      PCRE_INFO_OKPARTIAL
1230    .sp
1231    Return 1 if the pattern can be used for partial matching with
1232    \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1233    \fBint\fP variable. From release 8.00, this always returns 1, because the
1234    restrictions that previously applied to partial matching have been lifted. The
1235    .\" HREF
1236    \fBpcrepartial\fP
1237    .\"
1238    documentation gives details of partial matching.
1239    .sp
1240    PCRE_INFO_OPTIONS    PCRE_INFO_OPTIONS
1241    .sp
1242  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
1243  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
1244  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
1245  top-level option settings within the pattern itself.  top-level option settings at the start of the pattern itself. In other words,
1246    they are the options that will be in force when matching starts. For example,
1247    if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1248    result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1249    .P
1250  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
1251  alternatives begin with one of the following:  alternatives begin with one of the following:
1252    .sp
1253    ^     unless PCRE_MULTILINE is set    ^     unless PCRE_MULTILINE is set
1254    \\A    always    \eA    always
1255    \\G    always    \eG    always
1256    .\" JOIN
1257    .*    if PCRE_DOTALL is set and there are no back    .*    if PCRE_DOTALL is set and there are no back
1258            references to the subpattern in which .* appears            references to the subpattern in which .* appears
1259    .sp
1260  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
1261  \fBpcre_fullinfo()\fR.  \fBpcre_fullinfo()\fP.
1262    .sp
1263    PCRE_INFO_SIZE    PCRE_INFO_SIZE
1264    .sp
1265  Return the size of the compiled pattern, that is, the value that was passed as  Return the size of the compiled pattern. The fourth argument should point to a
1266  the argument to \fBpcre_malloc()\fR when PCRE was getting memory in which to  \fBsize_t\fP variable. This value does not include the size of the \fBpcre\fP
1267  place the compiled data. The fourth argument should point to a \fBsize_t\fR  structure that is returned by \fBpcre_compile()\fP. The value that is passed as
1268  variable.  the argument to \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory
1269    in which to place the compiled data is the value returned by this option plus
1270    the size of the \fBpcre\fP structure. Studying a compiled pattern, with or
1271    without JIT, does not alter the value returned by this option.
1272    .sp
1273    PCRE_INFO_STUDYSIZE    PCRE_INFO_STUDYSIZE
1274    .sp
1275  Returns the size of the data block pointed to by the \fIstudy_data\fR field in  Return the size of the data block pointed to by the \fIstudy_data\fP field in a
1276  a \fBpcre_extra\fR block. That is, it is the value that was passed to  \fBpcre_extra\fP block. If \fBpcre_extra\fP is NULL, or there is no study data,
1277  \fBpcre_malloc()\fR when PCRE was getting memory into which to place the data  zero is returned. The fourth argument should point to a \fBsize_t\fP variable.
1278  created by \fBpcre_study()\fR. The fourth argument should point to a  The \fIstudy_data\fP field is set by \fBpcre_study()\fP to record information
1279  \fBsize_t\fR variable.  that will speed up matching (see the section entitled
1280    .\" HTML <a href="#studyingapattern">
1281  .SH OBSOLETE INFO FUNCTION  .\" </a>
1282    "Studying a pattern"
1283    .\"
1284    above). The format of the \fIstudy_data\fP block is private, but its length
1285    is made available via this option so that it can be saved and restored (see the
1286    .\" HREF
1287    \fBpcreprecompile\fP
1288    .\"
1289    documentation for details).
1290    .
1291    .
1292    .SH "OBSOLETE INFO FUNCTION"
1293  .rs  .rs
1294  .sp  .sp
1295  .B int pcre_info(const pcre *\fIcode\fR, int *\fIoptptr\fR, int  .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
1296  .B *\fIfirstcharptr\fR);  .B *\fIfirstcharptr\fP);
1297  .PP  .PP
1298  The \fBpcre_info()\fR function is now obsolete because its interface is too  The \fBpcre_info()\fP function is now obsolete because its interface is too
1299  restrictive to return all the available data about a compiled pattern. New  restrictive to return all the available data about a compiled pattern. New
1300  programs should use \fBpcre_fullinfo()\fR instead. The yield of  programs should use \fBpcre_fullinfo()\fP instead. The yield of
1301  \fBpcre_info()\fR is the number of capturing subpatterns, or one of the  \fBpcre_info()\fP is the number of capturing subpatterns, or one of the
1302  following negative numbers:  following negative numbers:
1303    .sp
1304    PCRE_ERROR_NULL       the argument \fIcode\fR was NULL    PCRE_ERROR_NULL       the argument \fIcode\fP was NULL
1305    PCRE_ERROR_BADMAGIC   the "magic number" was not found    PCRE_ERROR_BADMAGIC   the "magic number" was not found
1306    .sp
1307  If the \fIoptptr\fR argument is not NULL, a copy of the options with which the  If the \fIoptptr\fP argument is not NULL, a copy of the options with which the
1308  pattern was compiled is placed in the integer it points to (see  pattern was compiled is placed in the integer it points to (see
1309  PCRE_INFO_OPTIONS above).  PCRE_INFO_OPTIONS above).
1310    .P
1311  If the pattern is not anchored and the \fIfirstcharptr\fR argument is not NULL,  If the pattern is not anchored and the \fIfirstcharptr\fP argument is not NULL,
1312  it is used to pass back information about the first character of any matched  it is used to pass back information about the first character of any matched
1313  string (see PCRE_INFO_FIRSTBYTE above).  string (see PCRE_INFO_FIRSTBYTE above).
1314    .
1315  .SH MATCHING A PATTERN  .
1316    .SH "REFERENCE COUNTS"
1317  .rs  .rs
1318  .sp  .sp
1319  .B int pcre_exec(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"  .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1320    .PP
1321    The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1322    data block that contains a compiled pattern. It is provided for the benefit of
1323    applications that operate in an object-oriented manner, where different parts
1324    of the application may be using the same compiled pattern, but you want to free
1325    the block when they are all done.
1326    .P
1327    When a pattern is compiled, the reference count field is initialized to zero.
1328    It is changed only by calling this function, whose action is to add the
1329    \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1330    function is the new value. However, the value of the count is constrained to
1331    lie between 0 and 65535, inclusive. If the new value is outside these limits,
1332    it is forced to the appropriate limit value.
1333    .P
1334    Except when it is zero, the reference count is not correctly preserved if a
1335    pattern is compiled on one host and then transferred to a host whose byte-order
1336    is different. (This seems a highly unlikely scenario.)
1337    .
1338    .
1339    .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1340    .rs
1341    .sp
1342    .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1343  .ti +5n  .ti +5n
1344  .B "const char *\fIsubject\fR," int \fIlength\fR, int \fIstartoffset\fR,  .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1345  .ti +5n  .ti +5n
1346  .B int \fIoptions\fR, int *\fIovector\fR, int \fIovecsize\fR);  .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1347  .PP  .P
1348  The function \fBpcre_exec()\fR is called to match a subject string against a  The function \fBpcre_exec()\fP is called to match a subject string against a
1349  pre-compiled pattern, which is passed in the \fIcode\fR argument. If the  compiled pattern, which is passed in the \fIcode\fP argument. If the
1350  pattern has been studied, the result of the study should be passed in the  pattern was studied, the result of the study should be passed in the
1351  \fIextra\fR argument.  \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1352    and \fIextra\fP arguments as many times as you like, in order to match
1353  Here is an example of a simple call to \fBpcre_exec()\fR:  different subject strings with the same pattern.
1354    .P
1355    This function is the main matching facility of the library, and it operates in
1356    a Perl-like manner. For specialist use there is also an alternative matching
1357    function, which is described
1358    .\" HTML <a href="#dfamatch">
1359    .\" </a>
1360    below
1361    .\"
1362    in the section about the \fBpcre_dfa_exec()\fP function.
1363    .P
1364    In most applications, the pattern will have been compiled (and optionally
1365    studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1366    possible to save compiled patterns and study data, and then use them later
1367    in different processes, possibly even on different hosts. For a discussion
1368    about this, see the
1369    .\" HREF
1370    \fBpcreprecompile\fP
1371    .\"
1372    documentation.
1373    .P
1374    Here is an example of a simple call to \fBpcre_exec()\fP:
1375    .sp
1376    int rc;    int rc;
1377    int ovector[30];    int ovector[30];
1378    rc = pcre_exec(    rc = pcre_exec(
# Line 684  Here is an example of a simple call to \ Line 1382  Here is an example of a simple call to \
1382      11,             /* the length of the subject string */      11,             /* the length of the subject string */
1383      0,              /* start at offset 0 in the subject */      0,              /* start at offset 0 in the subject */
1384      0,              /* default options */      0,              /* default options */
1385      ovector,        /* vector for substring information */      ovector,        /* vector of integers for substring information */
1386      30);            /* number of elements in the vector */      30);            /* number of elements (NOT size in bytes) */
1387    .
1388  If the \fIextra\fR argument is not NULL, it must point to a \fBpcre_extra\fR  .
1389  data block. The \fBpcre_study()\fR function returns such a block (when it  .\" HTML <a name="extradata"></a>
1390    .SS "Extra data for \fBpcre_exec()\fR"
1391    .rs
1392    .sp
1393    If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1394    data block. The \fBpcre_study()\fP function returns such a block (when it
1395  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
1396  additional information in it. The fields in the block are as follows:  additional information in it. The \fBpcre_extra\fP block contains the following
1397    fields (not necessarily in this order):
1398    unsigned long int \fIflags\fR;  .sp
1399    void *\fIstudy_data\fR;    unsigned long int \fIflags\fP;
1400    unsigned long int \fImatch_limit\fR;    void *\fIstudy_data\fP;
1401    void *\fIcallout_data\fR;    void *\fIexecutable_jit\fP;
1402      unsigned long int \fImatch_limit\fP;
1403  The \fIflags\fR field is a bitmap that specifies which of the other fields    unsigned long int \fImatch_limit_recursion\fP;
1404      void *\fIcallout_data\fP;
1405      const unsigned char *\fItables\fP;
1406      unsigned char **\fImark\fP;
1407    .sp
1408    The \fIflags\fP field is a bitmap that specifies which of the other fields
1409  are set. The flag bits are:  are set. The flag bits are:
1410    .sp
1411    PCRE_EXTRA_STUDY_DATA    PCRE_EXTRA_STUDY_DATA
1412      PCRE_EXTRA_EXECUTABLE_JIT
1413    PCRE_EXTRA_MATCH_LIMIT    PCRE_EXTRA_MATCH_LIMIT
1414      PCRE_EXTRA_MATCH_LIMIT_RECURSION
1415    PCRE_EXTRA_CALLOUT_DATA    PCRE_EXTRA_CALLOUT_DATA
1416      PCRE_EXTRA_TABLES
1417  Other flag bits should be set to zero. The \fIstudy_data\fR field is set in the    PCRE_EXTRA_MARK
1418  \fBpcre_extra\fR block that is returned by \fBpcre_study()\fR, together with  .sp
1419  the appropriate flag bit. You should not set this yourself, but you can add to  Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1420  the block by setting the other fields.  the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1421    returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1422  The \fImatch_limit\fR field provides a means of preventing PCRE from using up a  should not set these yourself, but you may add to the block by setting the
1423    other fields and their corresponding flag bits.
1424    .P
1425    The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1426  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,
1427  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
1428  classic example is the use of nested unlimited repeats. Internally, PCRE uses a  classic example is a pattern that uses nested unlimited repeats.
1429  function called \fBmatch()\fR which it calls repeatedly (sometimes  .P
1430  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
1431  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
1432  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
1433  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
1434    patterns that are not anchored, the count restarts from zero for each position
1435  The default limit for the library can be set when PCRE is built; the default  in the subject string.
1436    .P
1437    When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1438    with the PCRE_STUDY_JIT_COMPILE option, the way that the matching is executed
1439    is entirely different. However, there is still the possibility of runaway
1440    matching that goes on for a very long time, and so the \fImatch_limit\fP value
1441    is also used in this case (but in a different way) to limit how long the
1442    matching can continue.
1443    .P
1444    The default value for the limit can be set when PCRE is built; the default
1445  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
1446  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
1447  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
1448  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
1449  exceeded, \fBpcre_exec()\fR returns PCRE_ERROR_MATCHLIMIT.  PCRE_ERROR_MATCHLIMIT.
1450    .P
1451  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
1452  which is described in the \fBpcrecallout\fR documentation.  instead of limiting the total number of times that \fBmatch()\fP is called, it
1453    limits the depth of recursion. The recursion depth is a smaller number than the
1454  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.
1455  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.
1456  first matching position. However, if a pattern was compiled with PCRE_ANCHORED,  .P
1457  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
1458  unachored at matching time.  used, or, when PCRE has been compiled to use memory on the heap instead of the
1459    stack, the amount of heap memory that can be used. This limit is not relevant,
1460  When PCRE_UTF8 was set at compile time, the validity of the subject as a UTF-8  and is ignored, if the pattern was successfully studied with
1461  string is automatically checked, and the value of \fIstartoffset\fR is also  PCRE_STUDY_JIT_COMPILE.
1462  checked to ensure that it points to the start of a UTF-8 character. If an  .P
1463  invalid UTF-8 sequence of bytes is found, \fBpcre_exec()\fR returns the error  The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1464  PCRE_ERROR_BADUTF8. If \fIstartoffset\fR contains an invalid value,  built; the default default is the same value as the default for
1465  PCRE_ERROR_BADUTF8_OFFSET is returned.  \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1466    with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1467  If you already know that your subject is valid, and you want to skip these  PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1468  checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when  is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1469  calling \fBpcre_exec()\fR. You might want to do this for the second and  .P
1470  subsequent calls to \fBpcre_exec()\fR if you are making repeated calls to find  The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1471  all the matches in a single subject string. However, you should be sure that  and is described in the
1472  the value of \fIstartoffset\fR points to the start of a UTF-8 character. When  .\" HREF
1473  PCRE_NO_UTF8_CHECK is set, the effect of passing an invalid UTF-8 string as a  \fBpcrecallout\fP
1474  subject, or a value of \fIstartoffset\fR that does not point to the start of a  .\"
1475  UTF-8 character, is undefined. Your program may crash.  documentation.
1476    .P
1477  There are also three further options that can be set only at matching time:  The \fItables\fP field is used to pass a character tables pointer to
1478    \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1479    pattern. A non-NULL value is stored with the compiled pattern only if custom
1480    tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1481    If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1482    internal tables to be used. This facility is helpful when re-using patterns
1483    that have been saved after compiling with an external set of tables, because
1484    the external tables might be at a different address when \fBpcre_exec()\fP is
1485    called. See the
1486    .\" HREF
1487    \fBpcreprecompile\fP
1488    .\"
1489    documentation for a discussion of saving compiled patterns for later use.
1490    .P
1491    If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1492    be set to point to a \fBchar *\fP variable. If the pattern contains any
1493    backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1494    a name to pass back, a pointer to the name string (zero terminated) is placed
1495    in the variable pointed to by the \fImark\fP field. The names are within the
1496    compiled pattern; if you wish to retain such a name you must copy it before
1497    freeing the memory of a compiled pattern. If there is no name to pass back, the
1498    variable pointed to by the \fImark\fP field set to NULL. For details of the
1499    backtracking control verbs, see the section entitled
1500    .\" HTML <a href="pcrepattern#backtrackcontrol">
1501    .\" </a>
1502    "Backtracking control"
1503    .\"
1504    in the
1505    .\" HREF
1506    \fBpcrepattern\fP
1507    .\"
1508    documentation.
1509    .
1510    .
1511    .\" HTML <a name="execoptions"></a>
1512    .SS "Option bits for \fBpcre_exec()\fP"
1513    .rs
1514    .sp
1515    The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1516    zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1517    PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1518    PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_SOFT, and
1519    PCRE_PARTIAL_HARD.
1520    .P
1521    If the pattern was successfully studied with the PCRE_STUDY_JIT_COMPILE option,
1522    the only supported options for JIT execution are PCRE_NO_UTF8_CHECK,
1523    PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, and PCRE_NOTEMPTY_ATSTART. Note in
1524    particular that partial matching is not supported. If an unsupported option is
1525    used, JIT execution is disabled and the normal interpretive code in
1526    \fBpcre_exec()\fP is run.
1527    .sp
1528      PCRE_ANCHORED
1529    .sp
1530    The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1531    matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1532    to be anchored by virtue of its contents, it cannot be made unachored at
1533    matching time.
1534    .sp
1535      PCRE_BSR_ANYCRLF
1536      PCRE_BSR_UNICODE
1537    .sp
1538    These options (which are mutually exclusive) control what the \eR escape
1539    sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1540    match any Unicode newline sequence. These options override the choice that was
1541    made or defaulted when the pattern was compiled.
1542    .sp
1543      PCRE_NEWLINE_CR
1544      PCRE_NEWLINE_LF
1545      PCRE_NEWLINE_CRLF
1546      PCRE_NEWLINE_ANYCRLF
1547      PCRE_NEWLINE_ANY
1548    .sp
1549    These options override the newline definition that was chosen or defaulted when
1550    the pattern was compiled. For details, see the description of
1551    \fBpcre_compile()\fP above. During matching, the newline choice affects the
1552    behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1553    the way the match position is advanced after a match failure for an unanchored
1554    pattern.
1555    .P
1556    When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1557    match attempt for an unanchored pattern fails when the current position is at a
1558    CRLF sequence, and the pattern contains no explicit matches for CR or LF
1559    characters, the match position is advanced by two characters instead of one, in
1560    other words, to after the CRLF.
1561    .P
1562    The above rule is a compromise that makes the most common cases work as
1563    expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1564    set), it does not match the string "\er\enA" because, after failing at the
1565    start, it skips both the CR and the LF before retrying. However, the pattern
1566    [\er\en]A does match that string, because it contains an explicit CR or LF
1567    reference, and so advances only by one character after the first failure.
1568    .P
1569    An explicit match for CR of LF is either a literal appearance of one of those
1570    characters, or one of the \er or \en escape sequences. Implicit matches such as
1571    [^X] do not count, nor does \es (which includes CR and LF in the characters
1572    that it matches).
1573    .P
1574    Notwithstanding the above, anomalous effects may still occur when CRLF is a
1575    valid newline sequence and explicit \er or \en escapes appear in the pattern.
1576    .sp
1577    PCRE_NOTBOL    PCRE_NOTBOL
1578    .sp
1579  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
1580  circumflex metacharacter should not match before it. Setting this without  beginning of a line, so the circumflex metacharacter should not match before
1581  PCRE_MULTILINE (at compile time) causes circumflex never to match.  it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1582    never to match. This option affects only the behaviour of the circumflex
1583    metacharacter. It does not affect \eA.
1584    .sp
1585    PCRE_NOTEOL    PCRE_NOTEOL
1586    .sp
1587  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
1588  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
1589  it. Setting this without PCRE_MULTILINE (at compile time) causes dollar never  mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1590  to match.  compile time) causes dollar never to match. This option affects only the
1591    behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1592    .sp
1593    PCRE_NOTEMPTY    PCRE_NOTEMPTY
1594    .sp
1595  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
1596  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
1597  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
1598    .sp
1599    a?b?    a?b?
1600    .sp
1601  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
1602  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
1603  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".
1604    .sp
1605  Perl has no direct equivalent of PCRE_NOTEMPTY, but it does make a special case    PCRE_NOTEMPTY_ATSTART
1606  of a pattern match of the empty string within its \fBsplit()\fR function, and  .sp
1607  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
1608  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
1609  PCRE_NOTEMPTY set, and then if that fails by advancing the starting offset (see  can occur only if the pattern contains \eK.
1610  below) and trying an ordinary match again.  .P
1611    Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1612  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
1613  \fIsubject\fR, a length in \fIlength\fR, and a starting byte offset in  \fBsplit()\fP function, and when using the /g modifier. It is possible to
1614  \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
1615  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
1616  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
1617    ordinary match again. There is some code that demonstrates how to do this in
1618  If the pattern was compiled with the PCRE_UTF8 option, the subject must be a  the
1619  sequence of bytes that is a valid UTF-8 string, and the starting offset must  .\" HREF
1620  point to the beginning of a UTF-8 character. If an invalid UTF-8 string or  \fBpcredemo\fP
1621  offset is passed, an error (either PCRE_ERROR_BADUTF8 or  .\"
1622  PCRE_ERROR_BADUTF8_OFFSET) is returned, unless the option PCRE_NO_UTF8_CHECK is  sample program. In the most general case, you have to check to see if the
1623  set, in which case PCRE's behaviour is not defined.  newline convention recognizes CRLF as a newline, and if so, and the current
1624    character is CR followed by LF, advance the starting offset by two characters
1625    instead of one.
1626    .sp
1627      PCRE_NO_START_OPTIMIZE
1628    .sp
1629    There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1630    a match, in order to speed up the process. For example, if it is known that an
1631    unanchored match must start with a specific character, it searches the subject
1632    for that character, and fails immediately if it cannot find it, without
1633    actually running the main matching function. This means that a special item
1634    such as (*COMMIT) at the start of a pattern is not considered until after a
1635    suitable starting point for the match has been found. When callouts or (*MARK)
1636    items are in use, these "start-up" optimizations can cause them to be skipped
1637    if the pattern is never actually used. The start-up optimizations are in effect
1638    a pre-scan of the subject that takes place before the pattern is run.
1639    .P
1640    The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1641    causing performance to suffer, but ensuring that in cases where the result is
1642    "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1643    are considered at every possible starting position in the subject string. If
1644    PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1645    time.
1646    .P
1647    Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1648    Consider the pattern
1649    .sp
1650      (*COMMIT)ABC
1651    .sp
1652    When this is compiled, PCRE records the fact that a match must start with the
1653    character "A". Suppose the subject string is "DEFABC". The start-up
1654    optimization scans along the subject, finds "A" and runs the first match
1655    attempt from there. The (*COMMIT) item means that the pattern must match the
1656    current starting position, which in this case, it does. However, if the same
1657    match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1658    subject string does not happen. The first match attempt is run starting from
1659    "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1660    the overall result is "no match". If the pattern is studied, more start-up
1661    optimizations may be used. For example, a minimum length for the subject may be
1662    recorded. Consider the pattern
1663    .sp
1664      (*MARK:A)(X|Y)
1665    .sp
1666    The minimum length for a match is one character. If the subject is "ABC", there
1667    will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1668    If the pattern is studied, the final attempt does not take place, because PCRE
1669    knows that the subject is too short, and so the (*MARK) is never encountered.
1670    In this case, studying the pattern does not affect the overall match result,
1671    which is still "no match", but it does affect the auxiliary information that is
1672    returned.
1673    .sp
1674      PCRE_NO_UTF8_CHECK
1675    .sp
1676    When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1677    string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1678    The value of \fIstartoffset\fP is also checked to ensure that it points to the
1679    start of a UTF-8 character. There is a discussion about the validity of UTF-8
1680    strings in the
1681    .\" HTML <a href="pcre.html#utf8strings">
1682    .\" </a>
1683    section on UTF-8 support
1684    .\"
1685    in the main
1686    .\" HREF
1687    \fBpcre\fP
1688    .\"
1689    page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_exec()\fP returns
1690    the error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is
1691    a truncated UTF-8 character at the end of the subject, PCRE_ERROR_SHORTUTF8. In
1692    both cases, information about the precise nature of the error may also be
1693    returned (see the descriptions of these errors in the section entitled \fIError
1694    return values from\fP \fBpcre_exec()\fP
1695    .\" HTML <a href="#errorlist">
1696    .\" </a>
1697    below).
1698    .\"
1699    If \fIstartoffset\fP contains a value that does not point to the start of a
1700    UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1701    returned.
1702    .P
1703    If you already know that your subject is valid, and you want to skip these
1704    checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1705    calling \fBpcre_exec()\fP. You might want to do this for the second and
1706    subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1707    all the matches in a single subject string. However, you should be sure that
1708    the value of \fIstartoffset\fP points to the start of a UTF-8 character (or the
1709    end of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1710    invalid UTF-8 string as a subject or an invalid value of \fIstartoffset\fP is
1711    undefined. Your program may crash.
1712    .sp
1713      PCRE_PARTIAL_HARD
1714      PCRE_PARTIAL_SOFT
1715    .sp
1716    These options turn on the partial matching feature. For backwards
1717    compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1718    occurs if the end of the subject string is reached successfully, but there are
1719    not enough subject characters to complete the match. If this happens when
1720    PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1721    testing any remaining alternatives. Only if no complete match can be found is
1722    PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1723    PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1724    but only if no complete match can be found.
1725    .P
1726    If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1727    partial match is found, \fBpcre_exec()\fP immediately returns
1728    PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1729    when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1730    important that an alternative complete match.
1731    .P
1732    In both cases, the portion of the string that was inspected when the partial
1733    match was found is set as the first matching string. There is a more detailed
1734    discussion of partial and multi-segment matching, with examples, in the
1735    .\" HREF
1736    \fBpcrepartial\fP
1737    .\"
1738    documentation.
1739    .
1740    .
1741    .SS "The string to be matched by \fBpcre_exec()\fP"
1742    .rs
1743    .sp
1744    The subject string is passed to \fBpcre_exec()\fP as a pointer in
1745    \fIsubject\fP, a length (in bytes) in \fIlength\fP, and a starting byte offset
1746    in \fIstartoffset\fP. If this is negative or greater than the length of the
1747    subject, \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting
1748    offset is zero, the search for a match starts at the beginning of the subject,
1749    and this is by far the most common case. In UTF-8 mode, the byte offset must
1750    point to the start of a UTF-8 character (or the end of the subject). Unlike the
1751    pattern string, the subject may contain binary zero bytes.
1752    .P
1753  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
1754  same subject by calling \fBpcre_exec()\fR again after a previous success.  same subject by calling \fBpcre_exec()\fP again after a previous success.
1755  Setting \fIstartoffset\fR differs from just passing over a shortened string and  Setting \fIstartoffset\fP differs from just passing over a shortened string and
1756  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
1757  lookbehind. For example, consider the pattern  lookbehind. For example, consider the pattern
1758    .sp
1759    \\Biss\\B    \eBiss\eB
1760    .sp
1761  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
1762  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
1763  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
1764  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
1765  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
1766  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
1767  \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
1768  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
1769  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.
1770    .P
1771    Finding all the matches in a subject is tricky when the pattern can match an
1772    empty string. It is possible to emulate Perl's /g behaviour by first trying the
1773    match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1774    PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1775    and trying an ordinary match again. There is some code that demonstrates how to
1776    do this in the
1777    .\" HREF
1778    \fBpcredemo\fP
1779    .\"
1780    sample program. In the most general case, you have to check to see if the
1781    newline convention recognizes CRLF as a newline, and if so, and the current
1782    character is CR followed by LF, advance the starting offset by two characters
1783    instead of one.
1784    .P
1785  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
1786  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
1787  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.
1788    .
1789    .
1790    .SS "How \fBpcre_exec()\fP returns captured substrings"
1791    .rs
1792    .sp
1793  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
1794  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
1795  pattern. Following the usage in Jeffrey Friedl's book, this is called  pattern. Following the usage in Jeffrey Friedl's book, this is called
1796  "capturing" in what follows, and the phrase "capturing subpattern" is used for  "capturing" in what follows, and the phrase "capturing subpattern" is used for
1797  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
1798  kinds of parenthesized subpattern that do not cause substrings to be captured.  kinds of parenthesized subpattern that do not cause substrings to be captured.
1799    .P
1800  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
1801  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
1802  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
1803  back captured substrings, each substring using a pair of integers. The  argument is NOT the size of \fIovector\fP in bytes.
1804  remaining third of the vector is used as workspace by \fBpcre_exec()\fR while  .P
1805  matching capturing subpatterns, and is not available for passing back  The first two-thirds of the vector is used to pass back captured substrings,
1806  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
1807  three. If it is not, it is rounded down.  used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1808    and is not available for passing back information. The number passed in
1809  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
1810  returned in pairs of integers, starting at the beginning of \fIovector\fR, and  rounded down.
1811  continuing up to two-thirds of its length at the most. The first element of a  .P
1812  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
1813  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
1814  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
1815  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
1816  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
1817  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
1818  subpatterns, the return value from a successful match is 1, indicating that  mode. They are not character counts.
1819  just the first pair of offsets has been set.  .P
1820    The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
1821  Some convenience functions are provided for extracting the captured substrings  portion of the subject string matched by the entire pattern. The next pair is
1822  as separate strings. These are described in the following section.  used for the first capturing subpattern, and so on. The value returned by
1823    \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
1824  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
1825  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
1826  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.
1827  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.  
   
1828  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
1829  string that it matched that gets returned.  string that it matched that is returned.
1830    .P
1831  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
1832  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
1833  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
1834  \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
1835  \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
1836  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
1837  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
1838  to supply an \fIovector\fR.  is usually advisable to supply an \fIovector\fP of reasonable size.
1839    .P
1840  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
1841    in fact the vector is exactly the right size for the final match. For example,
1842    consider the pattern
1843    .sp
1844      (a)(?:(b)c|bd)
1845    .sp
1846    If a vector of 6 elements (allowing for only 1 captured substring) is given
1847    with subject string "abd", \fBpcre_exec()\fP will try to set the second
1848    captured string, thereby recording a vector overflow, before failing to match
1849    "c" and backing up to try the second alternative. The zero return, however,
1850    does correctly indicate that the maximum number of slots (namely 2) have been
1851    filled. In similar cases where there is temporary overflow, but the final
1852    number of used slots is actually less than the maximum, a non-zero value is
1853    returned.
1854    .P
1855    The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
1856  subpatterns there are in a compiled pattern. The smallest size for  subpatterns there are in a compiled pattern. The smallest size for
1857  \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
1858  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.
1859    .P
1860  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
1861    the subject when subpattern \fIn\fP has not been used at all. For example, if
1862    the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1863    function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1864    happens, both values in the offset pairs corresponding to unused subpatterns
1865    are set to -1.
1866    .P
1867    Offset values that correspond to unused subpatterns at the end of the
1868    expression are also set to -1. For example, if the string "abc" is matched
1869    against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1870    return from the function is 2, because the highest used capturing subpattern
1871    number is 1, and the offsets for for the second and third capturing subpatterns
1872    (assuming the vector is large enough, of course) are set to -1.
1873    .P
1874    \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
1875    correspond to capturing parentheses in the pattern are never changed. That is,
1876    if a pattern contains \fIn\fP capturing parentheses, no more than
1877    \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
1878    elements (in the first two-thirds) retain whatever values they previously had.
1879    .P
1880    Some convenience functions are provided for extracting the captured substrings
1881    as separate strings. These are described below.
1882    .
1883    .
1884    .\" HTML <a name="errorlist"></a>
1885    .SS "Error return values from \fBpcre_exec()\fP"
1886    .rs
1887    .sp
1888    If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1889  defined in the header file:  defined in the header file:
1890    .sp
1891    PCRE_ERROR_NOMATCH        (-1)    PCRE_ERROR_NOMATCH        (-1)
1892    .sp
1893  The subject string did not match the pattern.  The subject string did not match the pattern.
1894    .sp
1895    PCRE_ERROR_NULL           (-2)    PCRE_ERROR_NULL           (-2)
1896    .sp
1897  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
1898  NULL and \fIovecsize\fR was not zero.  NULL and \fIovecsize\fP was not zero.
1899    .sp
1900    PCRE_ERROR_BADOPTION      (-3)    PCRE_ERROR_BADOPTION      (-3)
1901    .sp
1902  An unrecognized bit was set in the \fIoptions\fR argument.  An unrecognized bit was set in the \fIoptions\fP argument.
1903    .sp
1904    PCRE_ERROR_BADMAGIC       (-4)    PCRE_ERROR_BADMAGIC       (-4)
1905    .sp
1906  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
1907  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
1908  magic number isn't present.  compiled in an environment of one endianness is run in an environment with the
1909    other endianness. This is the error that PCRE gives when the magic number is
1910    PCRE_ERROR_UNKNOWN_NODE   (-5)  not present.
1911    .sp
1912      PCRE_ERROR_UNKNOWN_OPCODE (-5)
1913    .sp
1914  While running the pattern match, an unknown item was encountered in the  While running the pattern match, an unknown item was encountered in the
1915  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
1916  of the compiled pattern.  of the compiled pattern.
1917    .sp
1918    PCRE_ERROR_NOMEMORY       (-6)    PCRE_ERROR_NOMEMORY       (-6)
1919    .sp
1920  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
1921  \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
1922  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
1923  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
1924  the end of matching.  automatically freed at the end of matching.
1925    .P
1926    This error is also given if \fBpcre_stack_malloc()\fP fails in
1927    \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
1928    \fB--disable-stack-for-recursion\fP.
1929    .sp
1930    PCRE_ERROR_NOSUBSTRING    (-7)    PCRE_ERROR_NOSUBSTRING    (-7)
1931    .sp
1932  This error is used by the \fBpcre_copy_substring()\fR,  This error is used by the \fBpcre_copy_substring()\fP,
1933  \fBpcre_get_substring()\fR, and \fBpcre_get_substring_list()\fR functions (see  \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
1934  below). It is never returned by \fBpcre_exec()\fR.  below). It is never returned by \fBpcre_exec()\fP.
1935    .sp
1936    PCRE_ERROR_MATCHLIMIT     (-8)    PCRE_ERROR_MATCHLIMIT     (-8)
1937    .sp
1938  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
1939  field in a \fBpcre_extra\fR structure (or defaulted) was reached. See the  \fBpcre_extra\fP structure (or defaulted) was reached. See the description
1940  description above.  above.
1941    .sp
1942    PCRE_ERROR_CALLOUT        (-9)    PCRE_ERROR_CALLOUT        (-9)
1943    .sp
1944  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
1945  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
1946  \fBpcrecallout\fR documentation for details.  .\" HREF
1947    \fBpcrecallout\fP
1948    .\"
1949    documentation for details.
1950    .sp
1951    PCRE_ERROR_BADUTF8        (-10)    PCRE_ERROR_BADUTF8        (-10)
1952    .sp
1953  A string that contains an invalid UTF-8 byte sequence was passed as a subject.  A string that contains an invalid UTF-8 byte sequence was passed as a subject,
1954    and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
1955    (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
1956    UTF-8 character is placed in the first element, and a reason code is placed in
1957    the second element. The reason codes are listed in the
1958    .\" HTML <a href="#badutf8reasons">
1959    .\" </a>
1960    following section.
1961    .\"
1962    For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
1963    truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
1964    PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
1965    .sp
1966    PCRE_ERROR_BADUTF8_OFFSET (-11)    PCRE_ERROR_BADUTF8_OFFSET (-11)
1967    .sp
1968  The UTF-8 byte sequence that was passed as a subject was valid, but the value  The UTF-8 byte sequence that was passed as a subject was checked and found to
1969  of \fIstartoffset\fR did not point to the beginning of a UTF-8 character.  be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
1970    \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
1971  .SH EXTRACTING CAPTURED SUBSTRINGS BY NUMBER  end of the subject.
1972    .sp
1973      PCRE_ERROR_PARTIAL        (-12)
1974    .sp
1975    The subject string did not match, but it did match partially. See the
1976    .\" HREF
1977    \fBpcrepartial\fP
1978    .\"
1979    documentation for details of partial matching.
1980    .sp
1981      PCRE_ERROR_BADPARTIAL     (-13)
1982    .sp
1983    This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1984    option was used with a compiled pattern containing items that were not
1985    supported for partial matching. From release 8.00 onwards, there are no
1986    restrictions on partial matching.
1987    .sp
1988      PCRE_ERROR_INTERNAL       (-14)
1989    .sp
1990    An unexpected internal error has occurred. This error could be caused by a bug
1991    in PCRE or by overwriting of the compiled pattern.
1992    .sp
1993      PCRE_ERROR_BADCOUNT       (-15)
1994    .sp
1995    This error is given if the value of the \fIovecsize\fP argument is negative.
1996    .sp
1997      PCRE_ERROR_RECURSIONLIMIT (-21)
1998    .sp
1999    The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2000    field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2001    description above.
2002    .sp
2003      PCRE_ERROR_BADNEWLINE     (-23)
2004    .sp
2005    An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2006    .sp
2007      PCRE_ERROR_BADOFFSET      (-24)
2008    .sp
2009    The value of \fIstartoffset\fP was negative or greater than the length of the
2010    subject, that is, the value in \fIlength\fP.
2011    .sp
2012      PCRE_ERROR_SHORTUTF8      (-25)
2013    .sp
2014    This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2015    ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2016    Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2017    fact sufficient to detect this case, but this special error code for
2018    PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2019    retained for backwards compatibility.
2020    .sp
2021      PCRE_ERROR_RECURSELOOP    (-26)
2022    .sp
2023    This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2024    the pattern. Specifically, it means that either the whole pattern or a
2025    subpattern has been called recursively for the second time at the same position
2026    in the subject string. Some simple patterns that might do this are detected and
2027    faulted at compile time, but more complicated cases, in particular mutual
2028    recursions between two different subpatterns, cannot be detected until run
2029    time.
2030    .sp
2031      PCRE_ERROR_JIT_STACKLIMIT (-27)
2032    .sp
2033    This error is returned when a pattern that was successfully studied using the
2034    PCRE_STUDY_JIT_COMPILE option is being matched, but the memory available for
2035    the just-in-time processing stack is not large enough. See the
2036    .\" HREF
2037    \fBpcrejit\fP
2038    .\"
2039    documentation for more details.
2040    .P
2041    Error numbers -16 to -20 and -22 are not used by \fBpcre_exec()\fP.
2042    .
2043    .
2044    .\" HTML <a name="badutf8reasons"></a>
2045    .SS "Reason codes for invalid UTF-8 strings"
2046    .rs
2047    .sp
2048    When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2049    PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2050    least 2, the offset of the start of the invalid UTF-8 character is placed in
2051    the first output vector element (\fIovector[0]\fP) and a reason code is placed
2052    in the second element (\fIovector[1]\fP). The reason codes are given names in
2053    the \fBpcre.h\fP header file:
2054    .sp
2055      PCRE_UTF8_ERR1
2056      PCRE_UTF8_ERR2
2057      PCRE_UTF8_ERR3
2058      PCRE_UTF8_ERR4
2059      PCRE_UTF8_ERR5
2060    .sp
2061    The string ends with a truncated UTF-8 character; the code specifies how many
2062    bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2063    no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2064    allows for up to 6 bytes, and this is checked first; hence the possibility of
2065    4 or 5 missing bytes.
2066    .sp
2067      PCRE_UTF8_ERR6
2068      PCRE_UTF8_ERR7
2069      PCRE_UTF8_ERR8
2070      PCRE_UTF8_ERR9
2071      PCRE_UTF8_ERR10
2072    .sp
2073    The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2074    character do not have the binary value 0b10 (that is, either the most
2075    significant bit is 0, or the next bit is 1).
2076    .sp
2077      PCRE_UTF8_ERR11
2078      PCRE_UTF8_ERR12
2079    .sp
2080    A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2081    these code points are excluded by RFC 3629.
2082    .sp
2083      PCRE_UTF8_ERR13
2084    .sp
2085    A 4-byte character has a value greater than 0x10fff; these code points are
2086    excluded by RFC 3629.
2087    .sp
2088      PCRE_UTF8_ERR14
2089    .sp
2090    A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2091    code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2092    from UTF-8.
2093    .sp
2094      PCRE_UTF8_ERR15
2095      PCRE_UTF8_ERR16
2096      PCRE_UTF8_ERR17
2097      PCRE_UTF8_ERR18
2098      PCRE_UTF8_ERR19
2099    .sp
2100    A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2101    value that can be represented by fewer bytes, which is invalid. For example,
2102    the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2103    one byte.
2104    .sp
2105      PCRE_UTF8_ERR20
2106    .sp
2107    The two most significant bits of the first byte of a character have the binary
2108    value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2109    byte can only validly occur as the second or subsequent byte of a multi-byte
2110    character.
2111    .sp
2112      PCRE_UTF8_ERR21
2113    .sp
2114    The first byte of a character has the value 0xfe or 0xff. These values can
2115    never occur in a valid UTF-8 string.
2116    .
2117    .
2118    .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2119  .rs  .rs
2120  .sp  .sp
2121  .B int pcre_copy_substring(const char *\fIsubject\fR, int *\fIovector\fR,  .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2122  .ti +5n  .ti +5n
2123  .B int \fIstringcount\fR, int \fIstringnumber\fR, char *\fIbuffer\fR,  .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
2124  .ti +5n  .ti +5n
2125  .B int \fIbuffersize\fR);  .B int \fIbuffersize\fP);
2126  .PP  .PP
2127  .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,  
2128  .ti +5n  .ti +5n
2129  .B int \fIstringcount\fR, int \fIstringnumber\fR,  .B int \fIstringcount\fP, int \fIstringnumber\fP,
2130  .ti +5n  .ti +5n
2131  .B const char **\fIstringptr\fR);  .B const char **\fIstringptr\fP);
2132  .PP  .PP
2133  .br  .B int pcre_get_substring_list(const char *\fIsubject\fP,
 .B int pcre_get_substring_list(const char *\fIsubject\fR,  
2134  .ti +5n  .ti +5n
2135  .B int *\fIovector\fR, int \fIstringcount\fR, "const char ***\fIlistptr\fR);"  .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
2136  .PP  .PP
2137  Captured substrings can be accessed directly by using the offsets returned by  Captured substrings can be accessed directly by using the offsets returned by
2138  \fBpcre_exec()\fR in \fIovector\fR. For convenience, the functions  \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2139  \fBpcre_copy_substring()\fR, \fBpcre_get_substring()\fR, and  \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2140  \fBpcre_get_substring_list()\fR are provided for extracting captured substrings  \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2141  as new, separate, zero-terminated strings. These functions identify substrings  as new, separate, zero-terminated strings. These functions identify substrings
2142  by number. The next section describes functions for extracting named  by number. The next section describes functions for extracting named
2143  substrings. A substring that contains a binary zero is correctly extracted and  substrings.
2144  has a further zero added on the end, but the result is not, of course,  .P
2145  a C string.  A substring that contains a binary zero is correctly extracted and has a
2146    further zero added on the end, but the result is not, of course, a C string.
2147    However, you can process such a string by referring to the length that is
2148    returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2149    Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2150    for handling strings containing binary zeros, because the end of the final
2151    string is not independently indicated.
2152    .P
2153  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:
2154  \fIsubject\fR is the subject string which has just been successfully matched,  \fIsubject\fP is the subject string that has just been successfully matched,
2155  \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
2156  \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
2157  captured by the match, including the substring that matched the entire regular  captured by the match, including the substring that matched the entire regular
2158  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
2159  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
2160  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
2161  the vector divided by three.  number of elements in the vector divided by three.
2162    .P
2163  The functions \fBpcre_copy_substring()\fR and \fBpcre_get_substring()\fR  The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2164  extract a single substring, whose number is given as \fIstringnumber\fR. A  extract a single substring, whose number is given as \fIstringnumber\fP. A
2165  value of zero extracts the substring that matched the entire pattern, while  value of zero extracts the substring that matched the entire pattern, whereas
2166  higher values extract the captured substrings. For \fBpcre_copy_substring()\fR,  higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2167  the string is placed in \fIbuffer\fR, whose length is given by  the string is placed in \fIbuffer\fP, whose length is given by
2168  \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
2169  obtained via \fBpcre_malloc\fR, and its address is returned via  obtained via \fBpcre_malloc\fP, and its address is returned via
2170  \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
2171  including the terminating zero, or one of  including the terminating zero, or one of these error codes:
2172    .sp
2173    PCRE_ERROR_NOMEMORY       (-6)    PCRE_ERROR_NOMEMORY       (-6)
2174    .sp
2175  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
2176  memory failed for \fBpcre_get_substring()\fR.  memory failed for \fBpcre_get_substring()\fP.
2177    .sp
2178    PCRE_ERROR_NOSUBSTRING    (-7)    PCRE_ERROR_NOSUBSTRING    (-7)
2179    .sp
2180  There is no substring whose number is \fIstringnumber\fR.  There is no substring whose number is \fIstringnumber\fP.
2181    .P
2182  The \fBpcre_get_substring_list()\fR function extracts all available substrings  The \fBpcre_get_substring_list()\fP function extracts all available substrings
2183  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
2184  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
2185  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
2186  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
2187  function is zero if all went well, or  function is zero if all went well, or the error code
2188    .sp
2189    PCRE_ERROR_NOMEMORY       (-6)    PCRE_ERROR_NOMEMORY       (-6)
2190    .sp
2191  if the attempt to get the memory block failed.  if the attempt to get the memory block failed.
2192    .P
2193  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
2194  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
2195  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
2196  string. This can be distinguished from a genuine zero-length substring by  string. This can be distinguished from a genuine zero-length substring by
2197  inspecting the appropriate offset in \fIovector\fR, which is negative for unset  inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2198  substrings.  substrings.
2199    .P
2200  The two convenience functions \fBpcre_free_substring()\fR and  The two convenience functions \fBpcre_free_substring()\fP and
2201  \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
2202  a previous call of \fBpcre_get_substring()\fR or  a previous call of \fBpcre_get_substring()\fP or
2203  \fBpcre_get_substring_list()\fR, respectively. They do nothing more than call  \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2204  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
2205  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
2206  linked via a special interface to another programming language which cannot use  linked via a special interface to another programming language that cannot use
2207  \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
2208  provided.  provided.
2209    .
2210  .SH EXTRACTING CAPTURED SUBSTRINGS BY NAME  .
2211    .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2212  .rs  .rs
2213  .sp  .sp
2214  .B int pcre_copy_named_substring(const pcre *\fIcode\fR,  .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2215  .ti +5n  .ti +5n
2216  .B const char *\fIsubject\fR, int *\fIovector\fR,  .B const char *\fIname\fP);
2217    .PP
2218    .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2219  .ti +5n  .ti +5n
2220  .B int \fIstringcount\fR, const char *\fIstringname\fR,  .B const char *\fIsubject\fP, int *\fIovector\fP,
2221  .ti +5n  .ti +5n
2222  .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,  
2223  .ti +5n  .ti +5n
2224  .B const char *\fIname\fR);  .B char *\fIbuffer\fP, int \fIbuffersize\fP);
2225  .PP  .PP
2226  .br  .B int pcre_get_named_substring(const pcre *\fIcode\fP,
 .B int pcre_get_named_substring(const pcre *\fIcode\fR,  
2227  .ti +5n  .ti +5n
2228  .B const char *\fIsubject\fR, int *\fIovector\fR,  .B const char *\fIsubject\fP, int *\fIovector\fP,
2229  .ti +5n  .ti +5n
2230  .B int \fIstringcount\fR, const char *\fIstringname\fR,  .B int \fIstringcount\fP, const char *\fIstringname\fP,
2231  .ti +5n  .ti +5n
2232  .B const char **\fIstringptr\fR);  .B const char **\fIstringptr\fP);
2233  .PP  .PP
2234  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.
2235  can be done by calling \fBpcre_get_stringnumber()\fR. The first argument is the  For example, for this pattern
2236  compiled pattern, and the second is the name. For example, for this pattern  .sp
2237      (a+)b(?<xxx>\ed+)...
2238    ab(?<xxx>\\d+)...  .sp
2239    the number of the subpattern called "xxx" is 2. If the name is known to be
2240  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
2241  extract the substring directly, or use one of the functions described in the  calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2242  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
2243  whole job.  subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2244    that name.
2245  Most of the arguments of \fIpcre_copy_named_substring()\fR and  .P
2246  \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
2247  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
2248  differences.  two functions that do the whole job.
2249    .P
2250    Most of the arguments of \fBpcre_copy_named_substring()\fP and
2251    \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2252    functions that extract by number. As these are described in the previous
2253    section, they are not re-described here. There are just two differences:
2254    .P
2255  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
2256  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
2257  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
2258  translation table.  translation table.
2259    .P
2260  These functions call \fBpcre_get_stringnumber()\fR, and if it succeeds, they  These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2261  then call \fIpcre_copy_substring()\fR or \fIpcre_get_substring()\fR, as  then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2262  appropriate.  appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2263    the behaviour may not be what you want (see the next section).
2264  .in 0  .P
2265  Last updated: 09 December 2003  \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2266  .br  subpatterns with the same number, as described in the
2267  Copyright (c) 1997-2003 University of Cambridge.  .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2268    .\" </a>
2269    section on duplicate subpattern numbers
2270    .\"
2271    in the
2272    .\" HREF
2273    \fBpcrepattern\fP
2274    .\"
2275    page, you cannot use names to distinguish the different subpatterns, because
2276    names are not included in the compiled code. The matching process uses only
2277    numbers. For this reason, the use of different names for subpatterns of the
2278    same number causes an error at compile time.
2279    .
2280    .
2281    .SH "DUPLICATE SUBPATTERN NAMES"
2282    .rs
2283    .sp
2284    .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2285    .ti +5n
2286    .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
2287    .PP
2288    When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2289    are not required to be unique. (Duplicate names are always allowed for
2290    subpatterns with the same number, created by using the (?| feature. Indeed, if
2291    such subpatterns are named, they are required to use the same names.)
2292    .P
2293    Normally, patterns with duplicate names are such that in any one match, only
2294    one of the named subpatterns participates. An example is shown in the
2295    .\" HREF
2296    \fBpcrepattern\fP
2297    .\"
2298    documentation.
2299    .P
2300    When duplicates are present, \fBpcre_copy_named_substring()\fP and
2301    \fBpcre_get_named_substring()\fP return the first substring corresponding to
2302    the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2303    returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2304    returns one of the numbers that are associated with the name, but it is not
2305    defined which it is.
2306    .P
2307    If you want to get full details of all captured substrings for a given name,
2308    you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2309    argument is the compiled pattern, and the second is the name. The third and
2310    fourth are pointers to variables which are updated by the function. After it
2311    has run, they point to the first and last entries in the name-to-number table
2312    for the given name. The function itself returns the length of each entry, or
2313    PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2314    described above in the section entitled \fIInformation about a pattern\fP
2315    .\" HTML <a href="#infoaboutpattern">
2316    .\" </a>
2317    above.
2318    .\"
2319    Given all the relevant entries for the name, you can extract each of their
2320    numbers, and hence the captured data, if any.
2321    .
2322    .
2323    .SH "FINDING ALL POSSIBLE MATCHES"
2324    .rs
2325    .sp
2326    The traditional matching function uses a similar algorithm to Perl, which stops
2327    when it finds the first match, starting at a given point in the subject. If you
2328    want to find all possible matches, or the longest possible match, consider
2329    using the alternative matching function (see below) instead. If you cannot use
2330    the alternative function, but still need to find all possible matches, you
2331    can kludge it up by making use of the callout facility, which is described in
2332    the
2333    .\" HREF
2334    \fBpcrecallout\fP
2335    .\"
2336    documentation.
2337    .P
2338    What you have to do is to insert a callout right at the end of the pattern.
2339    When your callout function is called, extract and save the current matched
2340    substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2341    other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2342    will yield PCRE_ERROR_NOMATCH.
2343    .
2344    .
2345    .\" HTML <a name="dfamatch"></a>
2346    .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2347    .rs
2348    .sp
2349    .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2350    .ti +5n
2351    .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2352    .ti +5n
2353    .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2354    .ti +5n
2355    .B int *\fIworkspace\fP, int \fIwscount\fP);
2356    .P
2357    The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2358    a compiled pattern, using a matching algorithm that scans the subject string
2359    just once, and does not backtrack. This has different characteristics to the
2360    normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2361    patterns are not supported. Nevertheless, there are times when this kind of
2362    matching can be useful. For a discussion of the two matching algorithms, and a
2363    list of features that \fBpcre_dfa_exec()\fP does not support, see the
2364    .\" HREF
2365    \fBpcrematching\fP
2366    .\"
2367    documentation.
2368    .P
2369    The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2370    \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2371    different way, and this is described below. The other common arguments are used
2372    in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2373    here.
2374    .P
2375    The two additional arguments provide workspace for the function. The workspace
2376    vector should contain at least 20 elements. It is used for keeping track of
2377    multiple paths through the pattern tree. More workspace will be needed for
2378    patterns and subjects where there are a lot of potential matches.
2379    .P
2380    Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2381    .sp
2382      int rc;
2383      int ovector[10];
2384      int wspace[20];
2385      rc = pcre_dfa_exec(
2386        re,             /* result of pcre_compile() */
2387        NULL,           /* we didn't study the pattern */
2388        "some string",  /* the subject string */
2389        11,             /* the length of the subject string */
2390        0,              /* start at offset 0 in the subject */
2391        0,              /* default options */
2392        ovector,        /* vector of integers for substring information */
2393        10,             /* number of elements (NOT size in bytes) */
2394        wspace,         /* working space vector */
2395        20);            /* number of elements (NOT size in bytes) */
2396    .
2397    .SS "Option bits for \fBpcre_dfa_exec()\fP"
2398    .rs
2399    .sp
2400    The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2401    zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2402    PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2403    PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2404    PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2405    All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2406    so their description is not repeated here.
2407    .sp
2408      PCRE_PARTIAL_HARD
2409      PCRE_PARTIAL_SOFT
2410    .sp
2411    These have the same general effect as they do for \fBpcre_exec()\fP, but the
2412    details are slightly different. When PCRE_PARTIAL_HARD is set for
2413    \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2414    is reached and there is still at least one matching possibility that requires
2415    additional characters. This happens even if some complete matches have also
2416    been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2417    is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2418    there have been no complete matches, but there is still at least one matching
2419    possibility. The portion of the string that was inspected when the longest
2420    partial match was found is set as the first matching string in both cases.
2421    There is a more detailed discussion of partial and multi-segment matching, with
2422    examples, in the
2423    .\" HREF
2424    \fBpcrepartial\fP
2425    .\"
2426    documentation.
2427    .sp
2428      PCRE_DFA_SHORTEST
2429    .sp
2430    Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2431    soon as it has found one match. Because of the way the alternative algorithm
2432    works, this is necessarily the shortest possible match at the first possible
2433    matching point in the subject string.
2434    .sp
2435      PCRE_DFA_RESTART
2436    .sp
2437    When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2438    again, with additional subject characters, and have it continue with the same
2439    match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2440    \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2441    before because data about the match so far is left in them after a partial
2442    match. There is more discussion of this facility in the
2443    .\" HREF
2444    \fBpcrepartial\fP
2445    .\"
2446    documentation.
2447    .
2448    .
2449    .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2450    .rs
2451    .sp
2452    When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2453    substring in the subject. Note, however, that all the matches from one run of
2454    the function start at the same point in the subject. The shorter matches are
2455    all initial substrings of the longer matches. For example, if the pattern
2456    .sp
2457      <.*>
2458    .sp
2459    is matched against the string
2460    .sp
2461      This is <something> <something else> <something further> no more
2462    .sp
2463    the three matched strings are
2464    .sp
2465      <something>
2466      <something> <something else>
2467      <something> <something else> <something further>
2468    .sp
2469    On success, the yield of the function is a number greater than zero, which is
2470    the number of matched substrings. The substrings themselves are returned in
2471    \fIovector\fP. Each string uses two elements; the first is the offset to the
2472    start, and the second is the offset to the end. In fact, all the strings have
2473    the same start offset. (Space could have been saved by giving this only once,
2474    but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2475    returns data, even though the meaning of the strings is different.)
2476    .P
2477    The strings are returned in reverse order of length; that is, the longest
2478    matching string is given first. If there were too many matches to fit into
2479    \fIovector\fP, the yield of the function is zero, and the vector is filled with
2480    the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2481    the entire \fIovector\fP for returning matched strings.
2482    .
2483    .
2484    .SS "Error returns from \fBpcre_dfa_exec()\fP"
2485    .rs
2486    .sp
2487    The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2488    Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2489    described
2490    .\" HTML <a href="#errorlist">
2491    .\" </a>
2492    above.
2493    .\"
2494    There are in addition the following errors that are specific to
2495    \fBpcre_dfa_exec()\fP:
2496    .sp
2497      PCRE_ERROR_DFA_UITEM      (-16)
2498    .sp
2499    This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2500    that it does not support, for instance, the use of \eC or a back reference.
2501    .sp
2502      PCRE_ERROR_DFA_UCOND      (-17)
2503    .sp
2504    This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2505    uses a back reference for the condition, or a test for recursion in a specific
2506    group. These are not supported.
2507    .sp
2508      PCRE_ERROR_DFA_UMLIMIT    (-18)
2509    .sp
2510    This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2511    block that contains a setting of the \fImatch_limit\fP or
2512    \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2513    meaningless for DFA matching).
2514    .sp
2515      PCRE_ERROR_DFA_WSSIZE     (-19)
2516    .sp
2517    This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2518    \fIworkspace\fP vector.
2519    .sp
2520      PCRE_ERROR_DFA_RECURSE    (-20)
2521    .sp
2522    When a recursive subpattern is processed, the matching function calls itself
2523    recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2524    error is given if the output vector is not large enough. This should be
2525    extremely rare, as a vector of size 1000 is used.
2526    .
2527    .
2528    .SH "SEE ALSO"
2529    .rs
2530    .sp
2531    \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
2532    \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
2533    \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
2534    .
2535    .
2536    .SH AUTHOR
2537    .rs
2538    .sp
2539    .nf
2540    Philip Hazel
2541    University Computing Service
2542    Cambridge CB2 3QH, England.
2543    .fi
2544    .
2545    .
2546    .SH REVISION
2547    .rs
2548    .sp
2549    .nf
2550    Last updated: 01 December 2011
2551    Copyright (c) 1997-2011 University of Cambridge.
2552    .fi

Legend:
Removed from v.73  
changed lines
  Added in v.776

  ViewVC Help
Powered by ViewVC 1.1.5