/[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 71 by nigel, Sat Feb 24 21:40:24 2007 UTC revision 518 by ph10, Tue May 18 15:47:01 2010 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"
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 int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
 .B int pcre_exec(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"  
29  .ti +5n  .ti +5n
30  .B "const char *\fIsubject\fR," int \fIlength\fR, int \fIstartoffset\fR,  .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
31  .ti +5n  .ti +5n
32  .B int \fIoptions\fR, int *\fIovector\fR, int \fIovecsize\fR);  .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
33  .PP  .PP
34  .br  .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
 .B int pcre_copy_named_substring(const pcre *\fIcode\fR,  
35  .ti +5n  .ti +5n
36  .B const char *\fIsubject\fR, int *\fIovector\fR,  .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
37  .ti +5n  .ti +5n
38  .B int \fIstringcount\fR, const char *\fIstringname\fR,  .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
39  .ti +5n  .ti +5n
40  .B char *\fIbuffer\fR, int \fIbuffersize\fR);  .B int *\fIworkspace\fP, int \fIwscount\fP);
41  .PP  .PP
42  .br  .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
 .B int pcre_copy_substring(const char *\fIsubject\fR, int *\fIovector\fR,  
43  .ti +5n  .ti +5n
44  .B int \fIstringcount\fR, int \fIstringnumber\fR, char *\fIbuffer\fR,  .B const char *\fIsubject\fP, int *\fIovector\fP,
45  .ti +5n  .ti +5n
46  .B int \fIbuffersize\fR);  .B int \fIstringcount\fP, const char *\fIstringname\fP,
47    .ti +5n
48    .B char *\fIbuffer\fP, int \fIbuffersize\fP);
49  .PP  .PP
50  .br  .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
51  .B int pcre_get_named_substring(const pcre *\fIcode\fR,  .ti +5n
52    .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
53  .ti +5n  .ti +5n
54  .B const char *\fIsubject\fR, int *\fIovector\fR,  .B int \fIbuffersize\fP);
55    .PP
56    .B int pcre_get_named_substring(const pcre *\fIcode\fP,
57  .ti +5n  .ti +5n
58  .B int \fIstringcount\fR, const char *\fIstringname\fR,  .B const char *\fIsubject\fP, int *\fIovector\fP,
59  .ti +5n  .ti +5n
60  .B const char **\fIstringptr\fR);  .B int \fIstringcount\fP, const char *\fIstringname\fP,
61    .ti +5n
62    .B const char **\fIstringptr\fP);
63    .PP
64    .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
65    .ti +5n
66    .B const char *\fIname\fP);
67  .PP  .PP
68  .br  .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
 .B int pcre_get_stringnumber(const pcre *\fIcode\fR,  
69  .ti +5n  .ti +5n
70  .B const char *\fIname\fR);  .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
71  .PP  .PP
72  .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,  
73  .ti +5n  .ti +5n
74  .B int \fIstringcount\fR, int \fIstringnumber\fR,  .B int \fIstringcount\fP, int \fIstringnumber\fP,
75  .ti +5n  .ti +5n
76  .B const char **\fIstringptr\fR);  .B const char **\fIstringptr\fP);
77  .PP  .PP
78  .br  .B int pcre_get_substring_list(const char *\fIsubject\fP,
 .B int pcre_get_substring_list(const char *\fIsubject\fR,  
79  .ti +5n  .ti +5n
80  .B int *\fIovector\fR, int \fIstringcount\fR, "const char ***\fIlistptr\fR);"  .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
81  .PP  .PP
82  .br  .B void pcre_free_substring(const char *\fIstringptr\fP);
 .B void pcre_free_substring(const char *\fIstringptr\fR);  
83  .PP  .PP
84  .br  .B void pcre_free_substring_list(const char **\fIstringptr\fP);
 .B void pcre_free_substring_list(const char **\fIstringptr\fR);  
85  .PP  .PP
 .br  
86  .B const unsigned char *pcre_maketables(void);  .B const unsigned char *pcre_maketables(void);
87  .PP  .PP
88  .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,"  
89  .ti +5n  .ti +5n
90  .B int \fIwhat\fR, void *\fIwhere\fR);  .B int \fIwhat\fP, void *\fIwhere\fP);
91  .PP  .PP
92  .br  .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
93  .B int pcre_info(const pcre *\fIcode\fR, int *\fIoptptr\fR, int  .B *\fIfirstcharptr\fP);
 .B *\fIfirstcharptr\fR);  
94  .PP  .PP
95  .br  .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
96  .B int pcre_config(int \fIwhat\fR, void *\fIwhere\fR);  .PP
97    .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
98  .PP  .PP
 .br  
99  .B char *pcre_version(void);  .B char *pcre_version(void);
100  .PP  .PP
 .br  
101  .B void *(*pcre_malloc)(size_t);  .B void *(*pcre_malloc)(size_t);
102  .PP  .PP
 .br  
103  .B void (*pcre_free)(void *);  .B void (*pcre_free)(void *);
104  .PP  .PP
105  .br  .B void *(*pcre_stack_malloc)(size_t);
106    .PP
107    .B void (*pcre_stack_free)(void *);
108    .PP
109  .B int (*pcre_callout)(pcre_callout_block *);  .B int (*pcre_callout)(pcre_callout_block *);
110    .
111  .SH PCRE API  .
112    .SH "PCRE API OVERVIEW"
113  .rs  .rs
114  .sp  .sp
115  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
116  a set of wrapper functions that correspond to the POSIX regular expression API.  also some wrapper functions that correspond to the POSIX regular expression
117  These are described in the \fBpcreposix\fR documentation.  API. These are described in the
118    .\" HREF
119  The native API function prototypes are defined in the header file \fBpcre.h\fR,  \fBpcreposix\fP
120  and on Unix systems the library itself is called \fBlibpcre.a\fR, so can be  .\"
121  accessed by adding \fB-lpcre\fR to the command for linking an application which  documentation. Both of these APIs define a set of C function calls. A C++
122  calls it. The header file defines the macros PCRE_MAJOR and PCRE_MINOR to  wrapper is distributed with PCRE. It is documented in the
123  contain the major and minor release numbers for the library. Applications can  .\" HREF
124  use these to include support for different releases.  \fBpcrecpp\fP
125    .\"
126  The functions \fBpcre_compile()\fR, \fBpcre_study()\fR, and \fBpcre_exec()\fR  page.
127  are used for compiling and matching regular expressions. A sample program that  .P
128  demonstrates the simplest way of using them is given in the file  The native API C function prototypes are defined in the header file
129  \fIpcredemo.c\fR. The \fBpcresample\fR documentation describes how to run it.  \fBpcre.h\fP, and on Unix systems the library itself is called \fBlibpcre\fP.
130    It can normally be accessed by adding \fB-lpcre\fP to the command for linking
131  There are convenience functions for extracting captured substrings from a  an application that uses PCRE. The header file defines the macros PCRE_MAJOR
132  matched subject string. They are:  and PCRE_MINOR to contain the major and minor release numbers for the library.
133    Applications can use these to include support for different releases of PCRE.
134    \fBpcre_copy_substring()\fR  .P
135    \fBpcre_copy_named_substring()\fR  The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
136    \fBpcre_get_substring()\fR  and \fBpcre_exec()\fP are used for compiling and matching regular expressions
137    \fBpcre_get_named_substring()\fR  in a Perl-compatible manner. A sample program that demonstrates the simplest
138    \fBpcre_get_substring_list()\fR  way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
139    source distribution. A listing of this program is given in the
140  \fBpcre_free_substring()\fR and \fBpcre_free_substring_list()\fR are also  .\" HREF
141    \fBpcredemo\fP
142    .\"
143    documentation, and the
144    .\" HREF
145    \fBpcresample\fP
146    .\"
147    documentation describes how to compile and run it.
148    .P
149    A second matching function, \fBpcre_dfa_exec()\fP, which is not
150    Perl-compatible, is also provided. This uses a different algorithm for the
151    matching. The alternative algorithm finds all possible matches (at a given
152    point in the subject), and scans the subject just once (unless there are
153    lookbehind assertions). However, this algorithm does not return captured
154    substrings. A description of the two matching algorithms and their advantages
155    and disadvantages is given in the
156    .\" HREF
157    \fBpcrematching\fP
158    .\"
159    documentation.
160    .P
161    In addition to the main compiling and matching functions, there are convenience
162    functions for extracting captured substrings from a subject string that is
163    matched by \fBpcre_exec()\fP. They are:
164    .sp
165      \fBpcre_copy_substring()\fP
166      \fBpcre_copy_named_substring()\fP
167      \fBpcre_get_substring()\fP
168      \fBpcre_get_named_substring()\fP
169      \fBpcre_get_substring_list()\fP
170      \fBpcre_get_stringnumber()\fP
171      \fBpcre_get_stringtable_entries()\fP
172    .sp
173    \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
174  provided, to free the memory used for extracted strings.  provided, to free the memory used for extracted strings.
175    .P
176  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
177  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,
178    or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
179  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
180  compiled pattern; \fBpcre_info()\fR is an obsolete version which returns only  internal tables that are generated when PCRE is built are used.
181    .P
182    The function \fBpcre_fullinfo()\fP is used to find out information about a
183    compiled pattern; \fBpcre_info()\fP is an obsolete version that returns only
184  some of the available information, but is retained for backwards compatibility.  some of the available information, but is retained for backwards compatibility.
185  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
186  version of PCRE and its date of release.  version of PCRE and its date of release.
187    .P
188  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
189  the entry points of the standard \fBmalloc()\fR and \fBfree()\fR functions  containing a compiled pattern. This is provided for the benefit of
190    object-oriented applications.
191    .P
192    The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
193    the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
194  respectively. PCRE calls the memory management functions via these variables,  respectively. PCRE calls the memory management functions via these variables,
195  so a calling program can replace them if it wishes to intercept the calls. This  so a calling program can replace them if it wishes to intercept the calls. This
196  should be done before calling any PCRE functions.  should be done before calling any PCRE functions.
197    .P
198  The global variable \fBpcre_callout\fR initially contains NULL. It can be set  The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
199    indirections to memory management functions. These special functions are used
200    only when PCRE is compiled to use the heap for remembering data, instead of
201    recursive function calls, when running the \fBpcre_exec()\fP function. See the
202    .\" HREF
203    \fBpcrebuild\fP
204    .\"
205    documentation for details of how to do this. It is a non-standard way of
206    building PCRE, for use in environments that have limited stacks. Because of the
207    greater use of memory management, it runs more slowly. Separate functions are
208    provided so that special-purpose external code can be used for this case. When
209    used, these functions are always called in a stack-like manner (last obtained,
210    first freed), and always for memory blocks of the same size. There is a
211    discussion about PCRE's stack usage in the
212    .\" HREF
213    \fBpcrestack\fP
214    .\"
215    documentation.
216    .P
217    The global variable \fBpcre_callout\fP initially contains NULL. It can be set
218  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
219  points during a matching operation. Details are given in the \fBpcrecallout\fR  points during a matching operation. Details are given in the
220    .\" HREF
221    \fBpcrecallout\fP
222    .\"
223  documentation.  documentation.
224    .
225    .
226    .\" HTML <a name="newlines"></a>
227    .SH NEWLINES
228    .rs
229    .sp
230    PCRE supports five different conventions for indicating line breaks in
231    strings: a single CR (carriage return) character, a single LF (linefeed)
232    character, the two-character sequence CRLF, any of the three preceding, or any
233    Unicode newline sequence. The Unicode newline sequences are the three just
234    mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
235    U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
236    (paragraph separator, U+2029).
237    .P
238    Each of the first three conventions is used by at least one operating system as
239    its standard newline sequence. When PCRE is built, a default can be specified.
240    The default default is LF, which is the Unix standard. When PCRE is run, the
241    default can be overridden, either when a pattern is compiled, or when it is
242    matched.
243    .P
244    At compile time, the newline convention can be specified by the \fIoptions\fP
245    argument of \fBpcre_compile()\fP, or it can be specified by special text at the
246    start of the pattern itself; this overrides any other settings. See the
247    .\" HREF
248    \fBpcrepattern\fP
249    .\"
250    page for details of the special character sequences.
251    .P
252    In the PCRE documentation the word "newline" is used to mean "the character or
253    pair of characters that indicate a line break". The choice of newline
254    convention affects the handling of the dot, circumflex, and dollar
255    metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
256    recognized line ending sequence, the match position advancement for a
257    non-anchored pattern. There is more detail about this in the
258    .\" HTML <a href="#execoptions">
259    .\" </a>
260    section on \fBpcre_exec()\fP options
261    .\"
262    below.
263    .P
264    The choice of newline convention does not affect the interpretation of
265    the \en or \er escape sequences, nor does it affect what \eR matches, which is
266    controlled in a similar way, but by separate options.
267    .
268    .
269  .SH MULTITHREADING  .SH MULTITHREADING
270  .rs  .rs
271  .sp  .sp
272  The PCRE functions can be used in multi-threading applications, with the  The PCRE functions can be used in multi-threading applications, with the
273  proviso that the memory management functions pointed to by \fBpcre_malloc\fR  proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
274  and \fBpcre_free\fR, and the callout function pointed to by \fBpcre_callout\fR,  \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
275  are shared by all threads.  callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
276    .P
277  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
278  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.
279    .
280  .SH CHECKING BUILD-TIME OPTIONS  .
281    .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
282    .rs
283    .sp
284    The compiled form of a regular expression can be saved and re-used at a later
285    time, possibly by a different program, and even on a host other than the one on
286    which it was compiled. Details are given in the
287    .\" HREF
288    \fBpcreprecompile\fP
289    .\"
290    documentation. However, compiling a regular expression with one version of PCRE
291    for use with a different version is not guaranteed to work and may cause
292    crashes.
293    .
294    .
295    .SH "CHECKING BUILD-TIME OPTIONS"
296  .rs  .rs
297  .sp  .sp
298  .B int pcre_config(int \fIwhat\fR, void *\fIwhere\fR);  .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
299  .PP  .PP
300  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
301  discover which optional features have been compiled into the PCRE library. The  discover which optional features have been compiled into the PCRE library. The
302  .\" HREF  .\" HREF
303  \fBpcrebuild\fR  \fBpcrebuild\fP
304  .\"  .\"
305  documentation has more details about these optional features.  documentation has more details about these optional features.
306    .P
307  The first argument for \fBpcre_config()\fR is an integer, specifying which  The first argument for \fBpcre_config()\fP is an integer, specifying which
308  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
309  which the information is placed. The following information is available:  which the information is placed. The following information is available:
310    .sp
311    PCRE_CONFIG_UTF8    PCRE_CONFIG_UTF8
312    .sp
313  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;
314  otherwise it is set to zero.  otherwise it is set to zero.
315    .sp
316      PCRE_CONFIG_UNICODE_PROPERTIES
317    .sp
318    The output is an integer that is set to one if support for Unicode character
319    properties is available; otherwise it is set to zero.
320    .sp
321    PCRE_CONFIG_NEWLINE    PCRE_CONFIG_NEWLINE
322    .sp
323  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
324  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
325  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.
326    Though they are derived from ASCII, the same values are returned in EBCDIC
327    environments. The default should normally correspond to the standard sequence
328    for your operating system.
329    .sp
330      PCRE_CONFIG_BSR
331    .sp
332    The output is an integer whose value indicates what character sequences the \eR
333    escape sequence matches by default. A value of 0 means that \eR matches any
334    Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
335    or CRLF. The default can be overridden when a pattern is compiled or matched.
336    .sp
337    PCRE_CONFIG_LINK_SIZE    PCRE_CONFIG_LINK_SIZE
338    .sp
339  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
340  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
341  allow larger regular expressions to be compiled, at the expense of slower  allow larger regular expressions to be compiled, at the expense of slower
342  matching. The default value of 2 is sufficient for all but the most massive  matching. The default value of 2 is sufficient for all but the most massive
343  patterns, since it allows the compiled pattern to be up to 64K in size.  patterns, since it allows the compiled pattern to be up to 64K in size.
344    .sp
345    PCRE_CONFIG_POSIX_MALLOC_THRESHOLD    PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
346    .sp
347  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
348  interface uses \fBmalloc()\fR for output vectors. Further details are given in  interface uses \fBmalloc()\fP for output vectors. Further details are given in
349  the \fBpcreposix\fR documentation.  the
350    .\" HREF
351    \fBpcreposix\fP
352    .\"
353    documentation.
354    .sp
355    PCRE_CONFIG_MATCH_LIMIT    PCRE_CONFIG_MATCH_LIMIT
356    .sp
357  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
358  internal matching function calls in a \fBpcre_exec()\fR execution. Further  internal matching function calls in a \fBpcre_exec()\fP execution. Further
359  details are given with \fBpcre_exec()\fR below.  details are given with \fBpcre_exec()\fP below.
360    .sp
361  .SH COMPILING A PATTERN    PCRE_CONFIG_MATCH_LIMIT_RECURSION
362  .rs  .sp
363  .sp  The output is a long integer that gives the default limit for the depth of
364  .B pcre *pcre_compile(const char *\fIpattern\fR, int \fIoptions\fR,  recursion when calling the internal matching function in a \fBpcre_exec()\fP
365  .ti +5n  execution. Further details are given with \fBpcre_exec()\fP below.
366  .B const char **\fIerrptr\fR, int *\fIerroffset\fR,  .sp
367  .ti +5n    PCRE_CONFIG_STACKRECURSE
368  .B const unsigned char *\fItableptr\fR);  .sp
369  .PP  The output is an integer that is set to one if internal recursion when running
370    \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
371  The function \fBpcre_compile()\fR is called to compile a pattern into an  to remember their state. This is the usual way that PCRE is compiled. The
372  internal form. The pattern is a C string terminated by a binary zero, and  output is zero if PCRE was compiled to use blocks of data on the heap instead
373  is passed in the argument \fIpattern\fR. A pointer to a single block of memory  of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
374  that is obtained via \fBpcre_malloc\fR is returned. This contains the compiled  \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
375  code and related data. The \fBpcre\fR type is defined for the returned block;  avoiding the use of the stack.
376  this is a typedef for a structure whose contents are not externally defined. It  .
377  is up to the caller to free the memory when it is no longer required.  .
378    .SH "COMPILING A PATTERN"
379    .rs
380    .sp
381    .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
382    .ti +5n
383    .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
384    .ti +5n
385    .B const unsigned char *\fItableptr\fP);
386    .sp
387    .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
388    .ti +5n
389    .B int *\fIerrorcodeptr\fP,
390    .ti +5n
391    .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
392    .ti +5n
393    .B const unsigned char *\fItableptr\fP);
394    .P
395    Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
396    called to compile a pattern into an internal form. The only difference between
397    the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
398    \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
399    too much repetition, we refer just to \fBpcre_compile()\fP below, but the
400    information applies equally to \fBpcre_compile2()\fP.
401    .P
402    The pattern is a C string terminated by a binary zero, and is passed in the
403    \fIpattern\fP argument. A pointer to a single block of memory that is obtained
404    via \fBpcre_malloc\fP is returned. This contains the compiled code and related
405    data. The \fBpcre\fP type is defined for the returned block; this is a typedef
406    for a structure whose contents are not externally defined. It is up to the
407    caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
408    .P
409  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
410  depend on memory location, the complete \fBpcre\fR data block is not  depend on memory location, the complete \fBpcre\fP data block is not
411  fully relocatable, because it contains a copy of the \fItableptr\fR argument,  fully relocatable, because it may contain a copy of the \fItableptr\fP
412  which is an address (see below).  argument, which is an address (see below).
413    .P
414  The \fIoptions\fR argument contains independent bits that affect the  The \fIoptions\fP argument contains various bit settings that affect the
415  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
416  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
417  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
418  in the \fBpcrepattern\fR documentation). For these options, the contents of the  within the pattern (see the detailed description in the
419  \fIoptions\fR argument specifies their initial settings at the start of  .\" HREF
420  compilation and execution. The PCRE_ANCHORED option can be set at the time of  \fBpcrepattern\fP
421  matching as well as at compile time.  .\"
422    documentation). For those options that can be different in different parts of
423  If \fIerrptr\fR is NULL, \fBpcre_compile()\fR returns NULL immediately.  the pattern, the contents of the \fIoptions\fP argument specifies their
424  Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fR returns  settings at the start of compilation and execution. The PCRE_ANCHORED,
425  NULL, and sets the variable pointed to by \fIerrptr\fR to point to a textual  PCRE_BSR_\fIxxx\fP, and PCRE_NEWLINE_\fIxxx\fP options can be set at the time
426  error message. The offset from the start of the pattern to the character where  of matching as well as at compile time.
427  the error was discovered is placed in the variable pointed to by  .P
428  \fIerroffset\fR, which must not be NULL. If it is, an immediate error is given.  If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
429    Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
430  If the final argument, \fItableptr\fR, is NULL, PCRE uses a default set of  NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
431  character tables which are built when it is compiled, using the default C  error message. This is a static string that is part of the library. You must
432  locale. Otherwise, \fItableptr\fR must be the result of a call to  not try to free it. The byte offset from the start of the pattern to the
433  \fBpcre_maketables()\fR. See the section on locale support below.  character that was being processed when the error was discovered is placed in
434    the variable pointed to by \fIerroffset\fP, which must not be NULL. If it is,
435  This code fragment shows a typical straightforward call to \fBpcre_compile()\fR:  an immediate error is given. Some errors are not detected until checks are
436    carried out when the whole pattern has been scanned; in this case the offset is
437    set to the end of the pattern.
438    .P
439    If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
440    \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
441    returned via this argument in the event of an error. This is in addition to the
442    textual error message. Error codes and messages are listed below.
443    .P
444    If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
445    character tables that are built when PCRE is compiled, using the default C
446    locale. Otherwise, \fItableptr\fP must be an address that is the result of a
447    call to \fBpcre_maketables()\fP. This value is stored with the compiled
448    pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
449    passed to it. For more discussion, see the section on locale support below.
450    .P
451    This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
452    .sp
453    pcre *re;    pcre *re;
454    const char *error;    const char *error;
455    int erroffset;    int erroffset;
# Line 265  This code fragment shows a typical strai Line 459  This code fragment shows a typical strai
459      &error,           /* for error message */      &error,           /* for error message */
460      &erroffset,       /* for error offset */      &erroffset,       /* for error offset */
461      NULL);            /* use default character tables */      NULL);            /* use default character tables */
462    .sp
463  The following option bits are defined:  The following names for option bits are defined in the \fBpcre.h\fP header
464    file:
465    .sp
466    PCRE_ANCHORED    PCRE_ANCHORED
467    .sp
468  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
469  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
470  being searched (the "subject string"). This effect can also be achieved by  being searched (the "subject string"). This effect can also be achieved by
471  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
472  Perl.  Perl.
473    .sp
474      PCRE_AUTO_CALLOUT
475    .sp
476    If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
477    all with number 255, before each pattern item. For discussion of the callout
478    facility, see the
479    .\" HREF
480    \fBpcrecallout\fP
481    .\"
482    documentation.
483    .sp
484      PCRE_BSR_ANYCRLF
485      PCRE_BSR_UNICODE
486    .sp
487    These options (which are mutually exclusive) control what the \eR escape
488    sequence matches. The choice is either to match only CR, LF, or CRLF, or to
489    match any Unicode newline sequence. The default is specified when PCRE is
490    built. It can be overridden from within the pattern, or by setting an option
491    when a compiled pattern is matched.
492    .sp
493    PCRE_CASELESS    PCRE_CASELESS
494    .sp
495  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
496  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
497  pattern by a (?i) option setting.  pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
498    concept of case for characters whose values are less than 128, so caseless
499    matching is always possible. For characters with higher values, the concept of
500    case is supported if PCRE is compiled with Unicode property support, but not
501    otherwise. If you want to use caseless matching for characters 128 and above,
502    you must ensure that PCRE is compiled with Unicode property support as well as
503    with UTF-8 support.
504    .sp
505    PCRE_DOLLAR_ENDONLY    PCRE_DOLLAR_ENDONLY
506    .sp
507  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
508  end of the subject string. Without this option, a dollar also matches  end of the subject string. Without this option, a dollar also matches
509  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
510  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.
511  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
512  a pattern.  pattern.
513    .sp
514    PCRE_DOTALL    PCRE_DOTALL
515    .sp
516  If this bit is set, a dot metacharater in the pattern matches all characters,  If this bit is set, a dot metacharater in the pattern matches all characters,
517  including newlines. Without it, newlines are excluded. This option is  including those that indicate newline. Without it, a dot does not match when
518  equivalent to Perl's /s option, and it can be changed within a pattern by a  the current position is at a newline. This option is equivalent to Perl's /s
519  (?s) option setting. A negative class such as [^a] always matches a newline  option, and it can be changed within a pattern by a (?s) option setting. A
520  character, independent of the setting of this option.  negative class such as [^a] always matches newline characters, independent of
521    the setting of this option.
522    .sp
523      PCRE_DUPNAMES
524    .sp
525    If this bit is set, names used to identify capturing subpatterns need not be
526    unique. This can be helpful for certain types of pattern when it is known that
527    only one instance of the named subpattern can ever be matched. There are more
528    details of named subpatterns below; see also the
529    .\" HREF
530    \fBpcrepattern\fP
531    .\"
532    documentation.
533    .sp
534    PCRE_EXTENDED    PCRE_EXTENDED
535    .sp
536  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
537  ignored except when escaped or inside a character class. Whitespace does not  ignored except when escaped or inside a character class. Whitespace does not
538  include the VT character (code 11). In addition, characters between an  include the VT character (code 11). In addition, characters between an
539  unescaped # outside a character class and the next newline character,  unescaped # outside a character class and the next newline, inclusive, are also
540  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
541  be changed within a pattern by a (?x) option setting.  pattern by a (?x) option setting.
542    .P
543  This option makes it possible to include comments inside complicated patterns.  This option makes it possible to include comments inside complicated patterns.
544  Note, however, that this applies only to data characters. Whitespace characters  Note, however, that this applies only to data characters. Whitespace characters
545  may never appear within special character sequences in a pattern, for example  may never appear within special character sequences in a pattern, for example
546  within the sequence (?( which introduces a conditional subpattern.  within the sequence (?( which introduces a conditional subpattern.
547    .sp
548    PCRE_EXTRA    PCRE_EXTRA
549    .sp
550  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
551  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
552  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
553  special meaning causes an error, thus reserving these combinations for future  special meaning causes an error, thus reserving these combinations for future
554  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
555  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
556  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
557  pattern.  no other features controlled by this option. It can also be set by a (?X)
558    option setting within a pattern.
559    .sp
560      PCRE_FIRSTLINE
561    .sp
562    If this option is set, an unanchored pattern is required to match before or at
563    the first newline in the subject string, though the matched text may continue
564    over the newline.
565    .sp
566      PCRE_JAVASCRIPT_COMPAT
567    .sp
568    If this option is set, PCRE's behaviour is changed in some ways so that it is
569    compatible with JavaScript rather than Perl. The changes are as follows:
570    .P
571    (1) A lone closing square bracket in a pattern causes a compile-time error,
572    because this is illegal in JavaScript (by default it is treated as a data
573    character). Thus, the pattern AB]CD becomes illegal when this option is set.
574    .P
575    (2) At run time, a back reference to an unset subpattern group matches an empty
576    string (by default this causes the current matching alternative to fail). A
577    pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
578    an "a" in the subject), whereas it fails by default, for Perl compatibility.
579    .sp
580    PCRE_MULTILINE    PCRE_MULTILINE
581    .sp
582  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
583  characters (even if it actually contains several newlines). The "start of line"  characters (even if it actually contains newlines). The "start of line"
584  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
585  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
586  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
587  Perl.  Perl.
588    .P
589  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
590  match immediately following or immediately before any newline in the subject  match immediately following or immediately before internal newlines in the
591  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
592  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
593  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
594  occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.  occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
595    .sp
596      PCRE_NEWLINE_CR
597      PCRE_NEWLINE_LF
598      PCRE_NEWLINE_CRLF
599      PCRE_NEWLINE_ANYCRLF
600      PCRE_NEWLINE_ANY
601    .sp
602    These options override the default newline definition that was chosen when PCRE
603    was built. Setting the first or the second specifies that a newline is
604    indicated by a single character (CR or LF, respectively). Setting
605    PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
606    CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
607    preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
608    that any Unicode newline sequence should be recognized. The Unicode newline
609    sequences are the three just mentioned, plus the single characters VT (vertical
610    tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
611    separator, U+2028), and PS (paragraph separator, U+2029). The last two are
612    recognized only in UTF-8 mode.
613    .P
614    The newline setting in the options word uses three bits that are treated
615    as a number, giving eight possibilities. Currently only six are used (default
616    plus the five values above). This means that if you set more than one newline
617    option, the combination may or may not be sensible. For example,
618    PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
619    other combinations may yield unused numbers and cause an error.
620    .P
621    The only time that a line break is specially recognized when compiling a
622    pattern is if PCRE_EXTENDED is set, and an unescaped # outside a character
623    class is encountered. This indicates a comment that lasts until after the next
624    line break sequence. In other circumstances, line break sequences are treated
625    as literal data, except that in PCRE_EXTENDED mode, both CR and LF are treated
626    as whitespace characters and are therefore ignored.
627    .P
628    The newline option that is set at compile time becomes the default that is used
629    for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
630    .sp
631    PCRE_NO_AUTO_CAPTURE    PCRE_NO_AUTO_CAPTURE
632    .sp
633  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
634  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
635  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
636  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
637  in Perl.  in Perl.
638    .sp
639      PCRE_UCP
640    .sp
641    This option changes the way PCRE processes \eb, \ed, \es, \ew, and some of the
642    POSIX character classes. By default, only ASCII characters are recognized, but
643    if PCRE_UCP is set, Unicode properties are used instead to classify characters.
644    More details are given in the section on
645    .\" HTML <a href="pcre.html#genericchartypes">
646    .\" </a>
647    generic character types
648    .\"
649    in the
650    .\" HREF
651    \fBpcrepattern\fP
652    .\"
653    page. If you set PCRE_UCP, matching one of the items it affects takes much
654    longer. The option is available only if PCRE has been compiled with Unicode
655    property support.
656    .sp
657    PCRE_UNGREEDY    PCRE_UNGREEDY
658    .sp
659  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
660  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
661  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.
662    .sp
663    PCRE_UTF8    PCRE_UTF8
664    .sp
665  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
666  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
667  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
668  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
669  behaviour of PCRE are given in the  behaviour of PCRE are given in the
670  .\" HTML <a href="pcre.html#utf8support">  .\" HTML <a href="pcre.html#utf8support">
# Line 367  section on UTF-8 support Line 673  section on UTF-8 support
673  .\"  .\"
674  in the main  in the main
675  .\" HREF  .\" HREF
676  \fBpcre\fR  \fBpcre\fP
677  .\"  .\"
678  page.  page.
679    .sp
680    PCRE_NO_UTF8_CHECK    PCRE_NO_UTF8_CHECK
681    .sp
682  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
683  automatically checked. If an invalid UTF-8 sequence of bytes is found,  automatically checked. There is a discussion about the
684  \fBpcre_compile()\fR returns an error. If you already know that your pattern is  .\" HTML <a href="pcre.html#utf8strings">
685  valid, and you want to skip this check for performance reasons, you can set the  .\" </a>
686  PCRE_NO_UTF8_CHECK option. When it is set, the effect of passing an invalid  validity of UTF-8 strings
687  UTF-8 string as a pattern is undefined. It may cause your program to crash.  .\"
688  Note that there is a similar option for suppressing the checking of subject  in the main
689  strings passed to \fBpcre_exec()\fR.  .\" HREF
690    \fBpcre\fP
691    .\"
692  .SH STUDYING A PATTERN  page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_compile()\fP
693  .rs  returns an error. If you already know that your pattern is valid, and you want
694  .sp  to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
695  .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
696  .ti +5n  pattern is undefined. It may cause your program to crash. Note that this option
697  .B const char **\fIerrptr\fR);  can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
698  .PP  the UTF-8 validity checking of subject strings.
699  When a pattern is going to be used several times, it is worth spending more  .
700  time analyzing it in order to speed up the time taken for matching. The  .
701  function \fBpcre_study()\fR takes a pointer to a compiled pattern as its first  .SH "COMPILATION ERROR CODES"
702  argument. If studing the pattern produces additional information that will help  .rs
703  speed up matching, \fBpcre_study()\fR returns a pointer to a \fBpcre_extra\fR  .sp
704  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
705    \fBpcre_compile2()\fP, along with the error messages that may be returned by
706  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
707  \fBpcre_exec()\fR. However, the \fBpcre_extra\fR block also contains other  out of use. To avoid confusion, they have not been re-used.
708  fields that can be set by the caller before the block is passed; these are  .sp
709  described below. If studying the pattern does not produce any additional     0  no error
710  information, \fBpcre_study()\fR returns NULL. In that circumstance, if the     1  \e at end of pattern
711  calling program wants to pass some of the other fields to \fBpcre_exec()\fR, it     2  \ec at end of pattern
712  must set up its own \fBpcre_extra\fR block.     3  unrecognized character follows \e
713       4  numbers out of order in {} quantifier
714  The second argument contains option bits. At present, no options are defined     5  number too big in {} quantifier
715  for \fBpcre_study()\fR, and this argument should always be zero.     6  missing terminating ] for character class
716       7  invalid escape sequence in character class
717  The third argument for \fBpcre_study()\fR is a pointer for an error message. If     8  range out of order in character class
718       9  nothing to repeat
719      10  [this code is not in use]
720      11  internal error: unexpected repeat
721      12  unrecognized character after (? or (?-
722      13  POSIX named classes are supported only within a class
723      14  missing )
724      15  reference to non-existent subpattern
725      16  erroffset passed as NULL
726      17  unknown option bit(s) set
727      18  missing ) after comment
728      19  [this code is not in use]
729      20  regular expression is too large
730      21  failed to get memory
731      22  unmatched parentheses
732      23  internal error: code overflow
733      24  unrecognized character after (?<
734      25  lookbehind assertion is not fixed length
735      26  malformed number or name after (?(
736      27  conditional group contains more than two branches
737      28  assertion expected after (?(
738      29  (?R or (?[+-]digits must be followed by )
739      30  unknown POSIX class name
740      31  POSIX collating elements are not supported
741      32  this version of PCRE is not compiled with PCRE_UTF8 support
742      33  [this code is not in use]
743      34  character value in \ex{...} sequence is too large
744      35  invalid condition (?(0)
745      36  \eC not allowed in lookbehind assertion
746      37  PCRE does not support \eL, \el, \eN, \eU, or \eu
747      38  number after (?C is > 255
748      39  closing ) for (?C expected
749      40  recursive call could loop indefinitely
750      41  unrecognized character after (?P
751      42  syntax error in subpattern name (missing terminator)
752      43  two named subpatterns have the same name
753      44  invalid UTF-8 string
754      45  support for \eP, \ep, and \eX has not been compiled
755      46  malformed \eP or \ep sequence
756      47  unknown property name after \eP or \ep
757      48  subpattern name is too long (maximum 32 characters)
758      49  too many named subpatterns (maximum 10000)
759      50  [this code is not in use]
760      51  octal value is greater than \e377 (not in UTF-8 mode)
761      52  internal error: overran compiling workspace
762      53  internal error: previously-checked referenced subpattern not found
763      54  DEFINE group contains more than one branch
764      55  repeating a DEFINE group is not allowed
765      56  inconsistent NEWLINE options
766      57  \eg is not followed by a braced, angle-bracketed, or quoted
767            name/number or by a plain number
768      58  a numbered reference must not be zero
769      59  an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
770      60  (*VERB) not recognized
771      61  number is too big
772      62  subpattern name expected
773      63  digit expected after (?+
774      64  ] is an invalid data character in JavaScript compatibility mode
775      65  different names for subpatterns of the same number are not allowed
776      66  (*MARK) must have an argument
777      67  this version of PCRE is not compiled with PCRE_UCP support
778    .sp
779    The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
780    be used if the limits were changed when PCRE was built.
781    .
782    .
783    .SH "STUDYING A PATTERN"
784    .rs
785    .sp
786    .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
787    .ti +5n
788    .B const char **\fIerrptr\fP);
789    .PP
790    If a compiled pattern is going to be used several times, it is worth spending
791    more time analyzing it in order to speed up the time taken for matching. The
792    function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
793    argument. If studying the pattern produces additional information that will
794    help speed up matching, \fBpcre_study()\fP returns a pointer to a
795    \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
796    results of the study.
797    .P
798    The returned value from \fBpcre_study()\fP can be passed directly to
799    \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
800    also contains other fields that can be set by the caller before the block is
801    passed; these are described
802    .\" HTML <a href="#extradata">
803    .\" </a>
804    below
805    .\"
806    in the section on matching a pattern.
807    .P
808    If studying the pattern does not produce any useful information,
809    \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program
810    wants to pass any of the other fields to \fBpcre_exec()\fP or
811    \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block.
812    .P
813    The second argument of \fBpcre_study()\fP contains option bits. At present, no
814    options are defined, and this argument should always be zero.
815    .P
816    The third argument for \fBpcre_study()\fP is a pointer for an error message. If
817  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
818  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
819  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
820  be sure that it has run successfully.  should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
821    sure that it has run successfully.
822  This is a typical call to \fBpcre_study\fR():  .P
823    This is a typical call to \fBpcre_study\fP():
824    .sp
825    pcre_extra *pe;    pcre_extra *pe;
826    pe = pcre_study(    pe = pcre_study(
827      re,             /* result of pcre_compile() */      re,             /* result of pcre_compile() */
828      0,              /* no options exist */      0,              /* no options exist */
829      &error);        /* set to NULL or points to a message */      &error);        /* set to NULL or points to a message */
830    .sp
831  At present, studying a pattern is useful only for non-anchored patterns that do  Studying a pattern does two things: first, a lower bound for the length of
832  not have a single fixed starting character. A bitmap of possible starting  subject string that is needed to match the pattern is computed. This does not
833  characters is created.  mean that there are any strings of that length that match, but it does
834    guarantee that no shorter strings match. The value is used by
835    \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP to avoid wasting time by trying to
836    match strings that are shorter than the lower bound. You can find out the value
837    in a calling program via the \fBpcre_fullinfo()\fP function.
838    .P
839    Studying a pattern is also useful for non-anchored patterns that do not have a
840    single fixed starting character. A bitmap of possible starting bytes is
841    created. This speeds up finding a position in the subject at which to start
842    matching.
843    .
844    .
845  .\" HTML <a name="localesupport"></a>  .\" HTML <a name="localesupport"></a>
846  .SH LOCALE SUPPORT  .SH "LOCALE SUPPORT"
847  .rs  .rs
848  .sp  .sp
849  PCRE handles caseless matching, and determines whether characters are letters,  PCRE handles caseless matching, and determines whether characters are letters,
850  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
851  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
852  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
853  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
854  is NULL, and is sufficient for many applications.  property support. Alternatively, the PCRE_UCP option can be set at compile
855    time; this causes \ew and friends to use Unicode property support instead of
856  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
857  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
858  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.
859  as necessary. For example, to build and use tables that are appropriate for the  .P
860  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
861  treated as letters), the following code could be used:  of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
862    Normally, the internal tables recognize only ASCII characters. However, when
863    setlocale(LC_CTYPE, "fr");  PCRE is built, it is possible to cause the internal tables to be rebuilt in the
864    default "C" locale of the local system, which may cause them to be different.
865    .P
866    The internal tables can always be overridden by tables supplied by the
867    application that calls PCRE. These may be created in a different locale from
868    the default. As more and more applications change to using Unicode, the need
869    for this locale support is expected to die away.
870    .P
871    External tables are built by calling the \fBpcre_maketables()\fP function,
872    which has no arguments, in the relevant locale. The result can then be passed
873    to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
874    example, to build and use tables that are appropriate for the French locale
875    (where accented characters with values greater than 128 are treated as letters),
876    the following code could be used:
877    .sp
878      setlocale(LC_CTYPE, "fr_FR");
879    tables = pcre_maketables();    tables = pcre_maketables();
880    re = pcre_compile(..., tables);    re = pcre_compile(..., tables);
881    .sp
882  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
883  pointer that is passed to \fBpcre_compile\fR is saved with the compiled  are using Windows, the name for the French locale is "french".
884  pattern, and the same tables are used via this pointer by \fBpcre_study()\fR  .P
885  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
886  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
887  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
888  memory containing the tables remains available for as long as it is needed.  needed.
889    .P
890  .SH INFORMATION ABOUT A PATTERN  The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
891    pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
892    and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
893    pattern, compilation, studying and matching all happen in the same locale, but
894    different patterns can be compiled in different locales.
895    .P
896    It is possible to pass a table pointer or NULL (indicating the use of the
897    internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
898    this facility could be used to match a pattern in a different locale from the
899    one in which it was compiled. Passing table pointers at run time is discussed
900    below in the section on matching a pattern.
901    .
902    .
903    .SH "INFORMATION ABOUT A PATTERN"
904  .rs  .rs
905  .sp  .sp
906  .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,"
907  .ti +5n  .ti +5n
908  .B int \fIwhat\fR, void *\fIwhere\fR);  .B int \fIwhat\fP, void *\fIwhere\fP);
909  .PP  .PP
910  The \fBpcre_fullinfo()\fR function returns information about a compiled  The \fBpcre_fullinfo()\fP function returns information about a compiled
911  pattern. It replaces the obsolete \fBpcre_info()\fR function, which is  pattern. It replaces the obsolete \fBpcre_info()\fP function, which is
912  nevertheless retained for backwards compability (and is documented below).  nevertheless retained for backwards compability (and is documented below).
913    .P
914  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
915  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
916  the pattern was not studied. The third argument specifies which piece of  the pattern was not studied. The third argument specifies which piece of
917  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
918  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
919  the following negative numbers:  the following negative numbers:
920    .sp
921    PCRE_ERROR_NULL       the argument \fIcode\fR was NULL    PCRE_ERROR_NULL       the argument \fIcode\fP was NULL
922                          the argument \fIwhere\fR was NULL                          the argument \fIwhere\fP was NULL
923    PCRE_ERROR_BADMAGIC   the "magic number" was not found    PCRE_ERROR_BADMAGIC   the "magic number" was not found
924    PCRE_ERROR_BADOPTION  the value of \fIwhat\fR was invalid    PCRE_ERROR_BADOPTION  the value of \fIwhat\fP was invalid
925    .sp
926  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
927  compiled pattern:  check against passing an arbitrary memory pointer. Here is a typical call of
928    \fBpcre_fullinfo()\fP, to obtain the length of the compiled pattern:
929    .sp
930    int rc;    int rc;
931    unsigned long int length;    size_t length;
932    rc = pcre_fullinfo(    rc = pcre_fullinfo(
933      re,               /* result of pcre_compile() */      re,               /* result of pcre_compile() */
934      pe,               /* result of pcre_study(), or NULL */      pe,               /* result of pcre_study(), or NULL */
935      PCRE_INFO_SIZE,   /* what is required */      PCRE_INFO_SIZE,   /* what is required */
936      &length);         /* where to put the data */      &length);         /* where to put the data */
937    .sp
938  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
939  as follows:  as follows:
940    .sp
941    PCRE_INFO_BACKREFMAX    PCRE_INFO_BACKREFMAX
942    .sp
943  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
944  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
945  no back references.  no back references.
946    .sp
947    PCRE_INFO_CAPTURECOUNT    PCRE_INFO_CAPTURECOUNT
948    .sp
949  Return the number of capturing subpatterns in the pattern. The fourth argument  Return the number of capturing subpatterns in the pattern. The fourth argument
950  should point to an \fbint\fR variable.  should point to an \fBint\fP variable.
951    .sp
952      PCRE_INFO_DEFAULT_TABLES
953    .sp
954    Return a pointer to the internal default character tables within PCRE. The
955    fourth argument should point to an \fBunsigned char *\fP variable. This
956    information call is provided for internal use by the \fBpcre_study()\fP
957    function. External callers can cause PCRE to use its internal tables by passing
958    a NULL table pointer.
959    .sp
960    PCRE_INFO_FIRSTBYTE    PCRE_INFO_FIRSTBYTE
961    .sp
962  Return information about the first byte of any matched string, for a  Return information about the first byte of any matched string, for a
963  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
964  old name is still recognized for backwards compatibility.)  variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
965    still recognized for backwards compatibility.)
966  If there is a fixed first byte, e.g. from a pattern such as (cat|cow|coyote),  .P
967  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
968    (cat|cow|coyote), its value is returned. Otherwise, if either
969    .sp
970  (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
971  starts with "^", or  starts with "^", or
972    .sp
973  (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
974  (if it were set, the pattern would be anchored),  (if it were set, the pattern would be anchored),
975    .sp
976  -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
977  subject string or after any newline within the string. Otherwise -2 is  subject string or after any newline within the string. Otherwise -2 is
978  returned. For anchored patterns, -2 is returned.  returned. For anchored patterns, -2 is returned.
979    .sp
980    PCRE_INFO_FIRSTTABLE    PCRE_INFO_FIRSTTABLE
981    .sp
982  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
983  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
984  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
985  fourth argument should point to an \fBunsigned char *\fR variable.  fourth argument should point to an \fBunsigned char *\fP variable.
986    .sp
987      PCRE_INFO_HASCRORLF
988    .sp
989    Return 1 if the pattern contains any explicit matches for CR or LF characters,
990    otherwise 0. The fourth argument should point to an \fBint\fP variable. An
991    explicit match is either a literal CR or LF character, or \er or \en.
992    .sp
993      PCRE_INFO_JCHANGED
994    .sp
995    Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
996    0. The fourth argument should point to an \fBint\fP variable. (?J) and
997    (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
998    .sp
999    PCRE_INFO_LASTLITERAL    PCRE_INFO_LASTLITERAL
1000    .sp
1001  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
1002  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
1003  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
1004  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
1005  follows something of variable length. For example, for the pattern  follows something of variable length. For example, for the pattern
1006  /^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
1007  is -1.  is -1.
1008    .sp
1009      PCRE_INFO_MINLENGTH
1010    .sp
1011    If the pattern was studied and a minimum length for matching subject strings
1012    was computed, its value is returned. Otherwise the returned value is -1. The
1013    value is a number of characters, not bytes (this may be relevant in UTF-8
1014    mode). The fourth argument should point to an \fBint\fP variable. A
1015    non-negative value is a lower bound to the length of any matching string. There
1016    may not be any strings of that length that do actually match, but every string
1017    that does match is at least that long.
1018    .sp
1019    PCRE_INFO_NAMECOUNT    PCRE_INFO_NAMECOUNT
1020    PCRE_INFO_NAMEENTRYSIZE    PCRE_INFO_NAMEENTRYSIZE
1021    PCRE_INFO_NAMETABLE    PCRE_INFO_NAMETABLE
1022    .sp
1023  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
1024  names are just an additional way of identifying the parentheses, which still  names are just an additional way of identifying the parentheses, which still
1025  acquire a number. A caller that wants to extract data from a named subpattern  acquire numbers. Several convenience functions such as
1026  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
1027  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
1028  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
1029  table for the pattern.  output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1030    you need to use the name-to-number map, which is described by these three
1031    values.
1032    .P
1033  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
1034  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
1035  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
1036  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
1037  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
1038  are the number of the capturing parenthesis, most significant byte first. The  are the number of the capturing parenthesis, most significant byte first. The
1039  rest of the entry is the corresponding name, zero terminated. The names are in  rest of the entry is the corresponding name, zero terminated.
1040  alphabetical order. For example, consider the following pattern (assume  .P
1041  PCRE_EXTENDED is set, so white space - including newlines - is ignored):  The names are in alphabetical order. Duplicate names may appear if (?| is used
1042    to create multiple groups with the same number, as described in the
1043    (?P<date> (?P<year>(\\d\\d)?\\d\\d) -  .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1044    (?P<month>\\d\\d) - (?P<day>\\d\\d) )  .\" </a>
1045    section on duplicate subpattern numbers
1046    .\"
1047    in the
1048    .\" HREF
1049    \fBpcrepattern\fP
1050    .\"
1051    page. Duplicate names for subpatterns with different numbers are permitted only
1052    if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1053    table in the order in which they were found in the pattern. In the absence of
1054    (?| this is the order of increasing number; when (?| is used this is not
1055    necessarily the case because later subpatterns may have lower numbers.
1056    .P
1057    As a simple example of the name/number table, consider the following pattern
1058    (assume PCRE_EXTENDED is set, so white space - including newlines - is
1059    ignored):
1060    .sp
1061    .\" JOIN
1062      (?<date> (?<year>(\ed\ed)?\ed\ed) -
1063      (?<month>\ed\ed) - (?<day>\ed\ed) )
1064    .sp
1065  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
1066  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
1067  bytes shows in hex, and undefined bytes shown as ??:  bytes shows in hexadecimal, and undefined bytes shown as ??:
1068    .sp
1069    00 01 d  a  t  e  00 ??    00 01 d  a  t  e  00 ??
1070    00 05 d  a  y  00 ?? ??    00 05 d  a  y  00 ?? ??
1071    00 04 m  o  n  t  h  00    00 04 m  o  n  t  h  00
1072    00 02 y  e  a  r  00 ??    00 02 y  e  a  r  00 ??
1073    .sp
1074  When writing code to extract data from named subpatterns, remember that the  When writing code to extract data from named subpatterns using the
1075  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
1076    different for each compiled pattern.
1077    .sp
1078      PCRE_INFO_OKPARTIAL
1079    .sp
1080    Return 1 if the pattern can be used for partial matching with
1081    \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1082    \fBint\fP variable. From release 8.00, this always returns 1, because the
1083    restrictions that previously applied to partial matching have been lifted. The
1084    .\" HREF
1085    \fBpcrepartial\fP
1086    .\"
1087    documentation gives details of partial matching.
1088    .sp
1089    PCRE_INFO_OPTIONS    PCRE_INFO_OPTIONS
1090    .sp
1091  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
1092  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
1093  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
1094  top-level option settings within the pattern itself.  top-level option settings at the start of the pattern itself. In other words,
1095    they are the options that will be in force when matching starts. For example,
1096    if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1097    result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1098    .P
1099  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
1100  alternatives begin with one of the following:  alternatives begin with one of the following:
1101    .sp
1102    ^     unless PCRE_MULTILINE is set    ^     unless PCRE_MULTILINE is set
1103    \\A    always    \eA    always
1104    \\G    always    \eG    always
1105    .\" JOIN
1106    .*    if PCRE_DOTALL is set and there are no back    .*    if PCRE_DOTALL is set and there are no back
1107            references to the subpattern in which .* appears            references to the subpattern in which .* appears
1108    .sp
1109  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
1110  \fBpcre_fullinfo()\fR.  \fBpcre_fullinfo()\fP.
1111    .sp
1112    PCRE_INFO_SIZE    PCRE_INFO_SIZE
1113    .sp
1114  Return the size of the compiled pattern, that is, the value that was passed as  Return the size of the compiled pattern, that is, the value that was passed as
1115  the argument to \fBpcre_malloc()\fR when PCRE was getting memory in which to  the argument to \fBpcre_malloc()\fP when PCRE was getting memory in which to
1116  place the compiled data. The fourth argument should point to a \fBsize_t\fR  place the compiled data. The fourth argument should point to a \fBsize_t\fP
1117  variable.  variable.
1118    .sp
1119    PCRE_INFO_STUDYSIZE    PCRE_INFO_STUDYSIZE
1120    .sp
1121  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
1122  a \fBpcre_extra\fR block. That is, it is the value that was passed to  a \fBpcre_extra\fP block. That is, it is the value that was passed to
1123  \fBpcre_malloc()\fR when PCRE was getting memory into which to place the data  \fBpcre_malloc()\fP when PCRE was getting memory into which to place the data
1124  created by \fBpcre_study()\fR. The fourth argument should point to a  created by \fBpcre_study()\fP. If \fBpcre_extra\fP is NULL, or there is no
1125  \fBsize_t\fR variable.  study data, zero is returned. The fourth argument should point to a
1126    \fBsize_t\fP variable.
1127  .SH OBSOLETE INFO FUNCTION  .
1128    .
1129    .SH "OBSOLETE INFO FUNCTION"
1130  .rs  .rs
1131  .sp  .sp
1132  .B int pcre_info(const pcre *\fIcode\fR, int *\fIoptptr\fR, int  .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
1133  .B *\fIfirstcharptr\fR);  .B *\fIfirstcharptr\fP);
1134  .PP  .PP
1135  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
1136  restrictive to return all the available data about a compiled pattern. New  restrictive to return all the available data about a compiled pattern. New
1137  programs should use \fBpcre_fullinfo()\fR instead. The yield of  programs should use \fBpcre_fullinfo()\fP instead. The yield of
1138  \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
1139  following negative numbers:  following negative numbers:
1140    .sp
1141    PCRE_ERROR_NULL       the argument \fIcode\fR was NULL    PCRE_ERROR_NULL       the argument \fIcode\fP was NULL
1142    PCRE_ERROR_BADMAGIC   the "magic number" was not found    PCRE_ERROR_BADMAGIC   the "magic number" was not found
1143    .sp
1144  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
1145  pattern was compiled is placed in the integer it points to (see  pattern was compiled is placed in the integer it points to (see
1146  PCRE_INFO_OPTIONS above).  PCRE_INFO_OPTIONS above).
1147    .P
1148  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,
1149  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
1150  string (see PCRE_INFO_FIRSTBYTE above).  string (see PCRE_INFO_FIRSTBYTE above).
1151    .
1152  .SH MATCHING A PATTERN  .
1153    .SH "REFERENCE COUNTS"
1154  .rs  .rs
1155  .sp  .sp
1156  .B int pcre_exec(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"  .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1157    .PP
1158    The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1159    data block that contains a compiled pattern. It is provided for the benefit of
1160    applications that operate in an object-oriented manner, where different parts
1161    of the application may be using the same compiled pattern, but you want to free
1162    the block when they are all done.
1163    .P
1164    When a pattern is compiled, the reference count field is initialized to zero.
1165    It is changed only by calling this function, whose action is to add the
1166    \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1167    function is the new value. However, the value of the count is constrained to
1168    lie between 0 and 65535, inclusive. If the new value is outside these limits,
1169    it is forced to the appropriate limit value.
1170    .P
1171    Except when it is zero, the reference count is not correctly preserved if a
1172    pattern is compiled on one host and then transferred to a host whose byte-order
1173    is different. (This seems a highly unlikely scenario.)
1174    .
1175    .
1176    .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1177    .rs
1178    .sp
1179    .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1180  .ti +5n  .ti +5n
1181  .B "const char *\fIsubject\fR," int \fIlength\fR, int \fIstartoffset\fR,  .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1182  .ti +5n  .ti +5n
1183  .B int \fIoptions\fR, int *\fIovector\fR, int \fIovecsize\fR);  .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1184  .PP  .P
1185  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
1186  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
1187  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
1188  \fIextra\fR argument.  \fIextra\fP argument. This function is the main matching facility of the
1189    library, and it operates in a Perl-like manner. For specialist use there is
1190  Here is an example of a simple call to \fBpcre_exec()\fR:  also an alternative matching function, which is described
1191    .\" HTML <a href="#dfamatch">
1192    .\" </a>
1193    below
1194    .\"
1195    in the section about the \fBpcre_dfa_exec()\fP function.
1196    .P
1197    In most applications, the pattern will have been compiled (and optionally
1198    studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1199    possible to save compiled patterns and study data, and then use them later
1200    in different processes, possibly even on different hosts. For a discussion
1201    about this, see the
1202    .\" HREF
1203    \fBpcreprecompile\fP
1204    .\"
1205    documentation.
1206    .P
1207    Here is an example of a simple call to \fBpcre_exec()\fP:
1208    .sp
1209    int rc;    int rc;
1210    int ovector[30];    int ovector[30];
1211    rc = pcre_exec(    rc = pcre_exec(
# Line 659  Here is an example of a simple call to \ Line 1215  Here is an example of a simple call to \
1215      11,             /* the length of the subject string */      11,             /* the length of the subject string */
1216      0,              /* start at offset 0 in the subject */      0,              /* start at offset 0 in the subject */
1217      0,              /* default options */      0,              /* default options */
1218      ovector,        /* vector for substring information */      ovector,        /* vector of integers for substring information */
1219      30);            /* number of elements in the vector */      30);            /* number of elements (NOT size in bytes) */
1220    .
1221  If the \fIextra\fR argument is not NULL, it must point to a \fBpcre_extra\fR  .\" HTML <a name="extradata"></a>
1222  data block. The \fBpcre_study()\fR function returns such a block (when it  .SS "Extra data for \fBpcre_exec()\fR"
1223    .rs
1224    .sp
1225    If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1226    data block. The \fBpcre_study()\fP function returns such a block (when it
1227  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
1228  additional information in it. The fields in the block are as follows:  additional information in it. The \fBpcre_extra\fP block contains the following
1229    fields (not necessarily in this order):
1230    unsigned long int \fIflags\fR;  .sp
1231    void *\fIstudy_data\fR;    unsigned long int \fIflags\fP;
1232    unsigned long int \fImatch_limit\fR;    void *\fIstudy_data\fP;
1233    void *\fIcallout_data\fR;    unsigned long int \fImatch_limit\fP;
1234      unsigned long int \fImatch_limit_recursion\fP;
1235  The \fIflags\fR field is a bitmap that specifies which of the other fields    void *\fIcallout_data\fP;
1236      const unsigned char *\fItables\fP;
1237      unsigned char **\fImark\fP;
1238    .sp
1239    The \fIflags\fP field is a bitmap that specifies which of the other fields
1240  are set. The flag bits are:  are set. The flag bits are:
1241    .sp
1242    PCRE_EXTRA_STUDY_DATA    PCRE_EXTRA_STUDY_DATA
1243    PCRE_EXTRA_MATCH_LIMIT    PCRE_EXTRA_MATCH_LIMIT
1244      PCRE_EXTRA_MATCH_LIMIT_RECURSION
1245    PCRE_EXTRA_CALLOUT_DATA    PCRE_EXTRA_CALLOUT_DATA
1246      PCRE_EXTRA_TABLES
1247  Other flag bits should be set to zero. The \fIstudy_data\fR field is set in the    PCRE_EXTRA_MARK
1248  \fBpcre_extra\fR block that is returned by \fBpcre_study()\fR, together with  .sp
1249  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 is set in the
1250  the block by setting the other fields.  \fBpcre_extra\fP block that is returned by \fBpcre_study()\fP, together with
1251    the appropriate flag bit. You should not set this yourself, but you may add to
1252  The \fImatch_limit\fR field provides a means of preventing PCRE from using up a  the block by setting the other fields and their corresponding flag bits.
1253    .P
1254    The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1255  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,
1256  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
1257  classic example is the use of nested unlimited repeats. Internally, PCRE uses a  classic example is a pattern that uses nested unlimited repeats.
1258  function called \fBmatch()\fR which it calls repeatedly (sometimes  .P
1259  recursively). The limit is imposed on the number of times this function is  Internally, PCRE uses a function called \fBmatch()\fP which it calls repeatedly
1260  called during a match, which has the effect of limiting the amount of recursion  (sometimes recursively). The limit set by \fImatch_limit\fP is imposed on the
1261  and backtracking that can take place. For patterns that are not anchored, the  number of times this function is called during a match, which has the effect of
1262  count starts from zero for each position in the subject string.  limiting the amount of backtracking that can take place. For patterns that are
1263    not anchored, the count restarts from zero for each position in the subject
1264  The default limit for the library can be set when PCRE is built; the default  string.
1265    .P
1266    The default value for the limit can be set when PCRE is built; the default
1267  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
1268  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
1269  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
1270  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
1271  exceeded, \fBpcre_exec()\fR returns PCRE_ERROR_MATCHLIMIT.  PCRE_ERROR_MATCHLIMIT.
1272    .P
1273  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
1274  which is described in the \fBpcrecallout\fR documentation.  instead of limiting the total number of times that \fBmatch()\fP is called, it
1275    limits the depth of recursion. The recursion depth is a smaller number than the
1276  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.
1277  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.
1278  first matching position. However, if a pattern was compiled with PCRE_ANCHORED,  .P
1279  or turned out to be anchored by virtue of its contents, it cannot be made  Limiting the recursion depth limits the amount of stack that can be used, or,
1280  unachored at matching time.  when PCRE has been compiled to use memory on the heap instead of the stack, the
1281    amount of heap memory that can be used.
1282  When PCRE_UTF8 was set at compile time, the validity of the subject as a UTF-8  .P
1283  string is automatically checked. If an invalid UTF-8 sequence of bytes is  The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1284  found, \fBpcre_exec()\fR returns the error PCRE_ERROR_BADUTF8. If you already  built; the default default is the same value as the default for
1285  know that your subject is valid, and you want to skip this check for  \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1286  performance reasons, you can set the PCRE_NO_UTF8_CHECK option when calling  with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1287  \fBpcre_exec()\fR. When this option is set, the effect of passing an invalid  PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1288  UTF-8 string as a subject is undefined. It may cause your program to crash.  is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1289    .P
1290  There are also three further options that can be set only at matching time:  The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1291    and is described in the
1292    .\" HREF
1293    \fBpcrecallout\fP
1294    .\"
1295    documentation.
1296    .P
1297    The \fItables\fP field is used to pass a character tables pointer to
1298    \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1299    pattern. A non-NULL value is stored with the compiled pattern only if custom
1300    tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1301    If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1302    internal tables to be used. This facility is helpful when re-using patterns
1303    that have been saved after compiling with an external set of tables, because
1304    the external tables might be at a different address when \fBpcre_exec()\fP is
1305    called. See the
1306    .\" HREF
1307    \fBpcreprecompile\fP
1308    .\"
1309    documentation for a discussion of saving compiled patterns for later use.
1310    .P
1311    If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1312    be set to point to a \fBchar *\fP variable. If the pattern contains any
1313    backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1314    a name to pass back, a pointer to the name string (zero terminated) is placed
1315    in the variable pointed to by the \fImark\fP field. The names are within the
1316    compiled pattern; if you wish to retain such a name you must copy it before
1317    freeing the memory of a compiled pattern. If there is no name to pass back, the
1318    variable pointed to by the \fImark\fP field set to NULL. For details of the
1319    backtracking control verbs, see the section entitled
1320    .\" HTML <a href="pcrepattern#backtrackcontrol">
1321    .\" </a>
1322    "Backtracking control"
1323    .\"
1324    in the
1325    .\" HREF
1326    \fBpcrepattern\fP
1327    .\"
1328    documentation.
1329    .
1330    .
1331    .\" HTML <a name="execoptions"></a>
1332    .SS "Option bits for \fBpcre_exec()\fP"
1333    .rs
1334    .sp
1335    The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1336    zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1337    PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1338    PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_SOFT, and
1339    PCRE_PARTIAL_HARD.
1340    .sp
1341      PCRE_ANCHORED
1342    .sp
1343    The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1344    matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1345    to be anchored by virtue of its contents, it cannot be made unachored at
1346    matching time.
1347    .sp
1348      PCRE_BSR_ANYCRLF
1349      PCRE_BSR_UNICODE
1350    .sp
1351    These options (which are mutually exclusive) control what the \eR escape
1352    sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1353    match any Unicode newline sequence. These options override the choice that was
1354    made or defaulted when the pattern was compiled.
1355    .sp
1356      PCRE_NEWLINE_CR
1357      PCRE_NEWLINE_LF
1358      PCRE_NEWLINE_CRLF
1359      PCRE_NEWLINE_ANYCRLF
1360      PCRE_NEWLINE_ANY
1361    .sp
1362    These options override the newline definition that was chosen or defaulted when
1363    the pattern was compiled. For details, see the description of
1364    \fBpcre_compile()\fP above. During matching, the newline choice affects the
1365    behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1366    the way the match position is advanced after a match failure for an unanchored
1367    pattern.
1368    .P
1369    When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1370    match attempt for an unanchored pattern fails when the current position is at a
1371    CRLF sequence, and the pattern contains no explicit matches for CR or LF
1372    characters, the match position is advanced by two characters instead of one, in
1373    other words, to after the CRLF.
1374    .P
1375    The above rule is a compromise that makes the most common cases work as
1376    expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1377    set), it does not match the string "\er\enA" because, after failing at the
1378    start, it skips both the CR and the LF before retrying. However, the pattern
1379    [\er\en]A does match that string, because it contains an explicit CR or LF
1380    reference, and so advances only by one character after the first failure.
1381    .P
1382    An explicit match for CR of LF is either a literal appearance of one of those
1383    characters, or one of the \er or \en escape sequences. Implicit matches such as
1384    [^X] do not count, nor does \es (which includes CR and LF in the characters
1385    that it matches).
1386    .P
1387    Notwithstanding the above, anomalous effects may still occur when CRLF is a
1388    valid newline sequence and explicit \er or \en escapes appear in the pattern.
1389    .sp
1390    PCRE_NOTBOL    PCRE_NOTBOL
1391    .sp
1392  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
1393  circumflex metacharacter should not match before it. Setting this without  beginning of a line, so the circumflex metacharacter should not match before
1394  PCRE_MULTILINE (at compile time) causes circumflex never to match.  it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1395    never to match. This option affects only the behaviour of the circumflex
1396    metacharacter. It does not affect \eA.
1397    .sp
1398    PCRE_NOTEOL    PCRE_NOTEOL
1399    .sp
1400  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
1401  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
1402  it. Setting this without PCRE_MULTILINE (at compile time) causes dollar never  mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1403  to match.  compile time) causes dollar never to match. This option affects only the
1404    behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1405    .sp
1406    PCRE_NOTEMPTY    PCRE_NOTEMPTY
1407    .sp
1408  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
1409  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
1410  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
1411    .sp
1412    a?b?    a?b?
1413    .sp
1414  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
1415  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
1416  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".
1417    .sp
1418  Perl has no direct equivalent of PCRE_NOTEMPTY, but it does make a special case    PCRE_NOTEMPTY_ATSTART
1419  of a pattern match of the empty string within its \fBsplit()\fR function, and  .sp
1420  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
1421  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
1422  PCRE_NOTEMPTY set, and then if that fails by advancing the starting offset (see  can occur only if the pattern contains \eK.
1423  below) and trying an ordinary match again.  .P
1424    Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1425  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
1426  \fIsubject\fR, a length in \fIlength\fR, and a starting offset in  \fBsplit()\fP function, and when using the /g modifier. It is possible to
1427  \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
1428    again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1429    if that fails, by advancing the starting offset (see below) and trying an
1430    ordinary match again. There is some code that demonstrates how to do this in
1431    the
1432    .\" HREF
1433    \fBpcredemo\fP
1434    .\"
1435    sample program.
1436    .sp
1437      PCRE_NO_START_OPTIMIZE
1438    .sp
1439    There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1440    a match, in order to speed up the process. For example, if it is known that a
1441    match must start with a specific character, it searches the subject for that
1442    character, and fails immediately if it cannot find it, without actually running
1443    the main matching function. When callouts are in use, these optimizations can
1444    cause them to be skipped. This option disables the "start-up" optimizations,
1445    causing performance to suffer, but ensuring that the callouts do occur.
1446    .sp
1447      PCRE_NO_UTF8_CHECK
1448    .sp
1449    When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1450    string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1451    The value of \fIstartoffset\fP is also checked to ensure that it points to the
1452    start of a UTF-8 character. There is a discussion about the validity of UTF-8
1453    strings in the
1454    .\" HTML <a href="pcre.html#utf8strings">
1455    .\" </a>
1456    section on UTF-8 support
1457    .\"
1458    in the main
1459    .\" HREF
1460    \fBpcre\fP
1461    .\"
1462    page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_exec()\fP returns
1463    the error PCRE_ERROR_BADUTF8. If \fIstartoffset\fP contains an invalid value,
1464    PCRE_ERROR_BADUTF8_OFFSET is returned.
1465    .P
1466    If you already know that your subject is valid, and you want to skip these
1467    checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1468    calling \fBpcre_exec()\fP. You might want to do this for the second and
1469    subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1470    all the matches in a single subject string. However, you should be sure that
1471    the value of \fIstartoffset\fP points to the start of a UTF-8 character. When
1472    PCRE_NO_UTF8_CHECK is set, the effect of passing an invalid UTF-8 string as a
1473    subject, or a value of \fIstartoffset\fP that does not point to the start of a
1474    UTF-8 character, is undefined. Your program may crash.
1475    .sp
1476      PCRE_PARTIAL_HARD
1477      PCRE_PARTIAL_SOFT
1478    .sp
1479    These options turn on the partial matching feature. For backwards
1480    compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1481    occurs if the end of the subject string is reached successfully, but there are
1482    not enough subject characters to complete the match. If this happens when
1483    PCRE_PARTIAL_HARD is set, \fBpcre_exec()\fP immediately returns
1484    PCRE_ERROR_PARTIAL. Otherwise, if PCRE_PARTIAL_SOFT is set, matching continues
1485    by testing any other alternatives. Only if they all fail is PCRE_ERROR_PARTIAL
1486    returned (instead of PCRE_ERROR_NOMATCH). The portion of the string that
1487    was inspected when the partial match was found is set as the first matching
1488    string. There is a more detailed discussion in the
1489    .\" HREF
1490    \fBpcrepartial\fP
1491    .\"
1492    documentation.
1493    .
1494    .SS "The string to be matched by \fBpcre_exec()\fP"
1495    .rs
1496    .sp
1497    The subject string is passed to \fBpcre_exec()\fP as a pointer in
1498    \fIsubject\fP, a length (in bytes) in \fIlength\fP, and a starting byte offset
1499    in \fIstartoffset\fP. In UTF-8 mode, the byte offset must point to the start of
1500    a UTF-8 character. Unlike the pattern string, the subject may contain binary
1501  zero bytes. When the starting offset is zero, the search for a match starts at  zero bytes. When the starting offset is zero, the search for a match starts at
1502  the beginning of the subject, and this is by far the most common case.  the beginning of the subject, and this is by far the most common case.
1503    .P
 If the pattern was compiled with the PCRE_UTF8 option, the subject must be a  
 sequence of bytes that is a valid UTF-8 string. If an invalid UTF-8 string is  
 passed, PCRE's behaviour is not defined.  
   
1504  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
1505  same subject by calling \fBpcre_exec()\fR again after a previous success.  same subject by calling \fBpcre_exec()\fP again after a previous success.
1506  Setting \fIstartoffset\fR differs from just passing over a shortened string and  Setting \fIstartoffset\fP differs from just passing over a shortened string and
1507  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
1508  lookbehind. For example, consider the pattern  lookbehind. For example, consider the pattern
1509    .sp
1510    \\Biss\\B    \eBiss\eB
1511    .sp
1512  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
1513  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
1514  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
1515  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
1516  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
1517  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
1518  \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
1519  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
1520  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.
1521    .P
1522  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
1523  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
1524  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.
1525    .
1526    .SS "How \fBpcre_exec()\fP returns captured substrings"
1527    .rs
1528    .sp
1529  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
1530  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
1531  pattern. Following the usage in Jeffrey Friedl's book, this is called  pattern. Following the usage in Jeffrey Friedl's book, this is called
1532  "capturing" in what follows, and the phrase "capturing subpattern" is used for  "capturing" in what follows, and the phrase "capturing subpattern" is used for
1533  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
1534  kinds of parenthesized subpattern that do not cause substrings to be captured.  kinds of parenthesized subpattern that do not cause substrings to be captured.
1535    .P
1536  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
1537  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
1538  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
1539  back captured substrings, each substring using a pair of integers. The  argument is NOT the size of \fIovector\fP in bytes.
1540  remaining third of the vector is used as workspace by \fBpcre_exec()\fR while  .P
1541  matching capturing subpatterns, and is not available for passing back  The first two-thirds of the vector is used to pass back captured substrings,
1542  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
1543  three. If it is not, it is rounded down.  used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1544    and is not available for passing back information. The number passed in
1545  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
1546  returned in pairs of integers, starting at the beginning of \fIovector\fR, and  rounded down.
1547  continuing up to two-thirds of its length at the most. The first element of a  .P
1548  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
1549  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
1550  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
1551  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
1552  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
1553  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
1554  subpatterns, the return value from a successful match is 1, indicating that  mode. They are not character counts.
1555  just the first pair of offsets has been set.  .P
1556    The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
1557  Some convenience functions are provided for extracting the captured substrings  portion of the subject string matched by the entire pattern. The next pair is
1558  as separate strings. These are described in the following section.  used for the first capturing subpattern, and so on. The value returned by
1559    \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
1560  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
1561  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
1562  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.
1563  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.  
   
1564  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
1565  string that it matched that gets returned.  string that it matched that is returned.
1566    .P
1567  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
1568  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
1569  value of zero. In particular, if the substring offsets are not of interest,  returns a value of zero. If the substring offsets are not of interest,
1570  \fBpcre_exec()\fR may be called with \fIovector\fR passed as NULL and  \fBpcre_exec()\fP may be called with \fIovector\fP passed as NULL and
1571  \fIovecsize\fR as zero. However, if the pattern contains back references and  \fIovecsize\fP as zero. However, if the pattern contains back references and
1572  the \fIovector\fR isn't big enough to remember the related substrings, PCRE has  the \fIovector\fP is not big enough to remember the related substrings, PCRE
1573  to get additional memory for use during matching. Thus it is usually advisable  has to get additional memory for use during matching. Thus it is usually
1574  to supply an \fIovector\fR.  advisable to supply an \fIovector\fP.
1575    .P
1576  Note that \fBpcre_info()\fR can be used to find out how many capturing  The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
1577  subpatterns there are in a compiled pattern. The smallest size for  subpatterns there are in a compiled pattern. The smallest size for
1578  \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
1579  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.
1580    .P
1581  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
1582    the subject when subpattern \fIn\fP has not been used at all. For example, if
1583    the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1584    function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1585    happens, both values in the offset pairs corresponding to unused subpatterns
1586    are set to -1.
1587    .P
1588    Offset values that correspond to unused subpatterns at the end of the
1589    expression are also set to -1. For example, if the string "abc" is matched
1590    against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1591    return from the function is 2, because the highest used capturing subpattern
1592    number is 1. However, you can refer to the offsets for the second and third
1593    capturing subpatterns if you wish (assuming the vector is large enough, of
1594    course).
1595    .P
1596    Some convenience functions are provided for extracting the captured substrings
1597    as separate strings. These are described below.
1598    .
1599    .\" HTML <a name="errorlist"></a>
1600    .SS "Error return values from \fBpcre_exec()\fP"
1601    .rs
1602    .sp
1603    If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1604  defined in the header file:  defined in the header file:
1605    .sp
1606    PCRE_ERROR_NOMATCH        (-1)    PCRE_ERROR_NOMATCH        (-1)
1607    .sp
1608  The subject string did not match the pattern.  The subject string did not match the pattern.
1609    .sp
1610    PCRE_ERROR_NULL           (-2)    PCRE_ERROR_NULL           (-2)
1611    .sp
1612  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
1613  NULL and \fIovecsize\fR was not zero.  NULL and \fIovecsize\fP was not zero.
1614    .sp
1615    PCRE_ERROR_BADOPTION      (-3)    PCRE_ERROR_BADOPTION      (-3)
1616    .sp
1617  An unrecognized bit was set in the \fIoptions\fR argument.  An unrecognized bit was set in the \fIoptions\fP argument.
1618    .sp
1619    PCRE_ERROR_BADMAGIC       (-4)    PCRE_ERROR_BADMAGIC       (-4)
1620    .sp
1621  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
1622  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
1623  magic number isn't present.  compiled in an environment of one endianness is run in an environment with the
1624    other endianness. This is the error that PCRE gives when the magic number is
1625    PCRE_ERROR_UNKNOWN_NODE   (-5)  not present.
1626    .sp
1627      PCRE_ERROR_UNKNOWN_OPCODE (-5)
1628    .sp
1629  While running the pattern match, an unknown item was encountered in the  While running the pattern match, an unknown item was encountered in the
1630  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
1631  of the compiled pattern.  of the compiled pattern.
1632    .sp
1633    PCRE_ERROR_NOMEMORY       (-6)    PCRE_ERROR_NOMEMORY       (-6)
1634    .sp
1635  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
1636  \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
1637  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
1638  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
1639  the end of matching.  automatically freed at the end of matching.
1640    .sp
1641    PCRE_ERROR_NOSUBSTRING    (-7)    PCRE_ERROR_NOSUBSTRING    (-7)
1642    .sp
1643  This error is used by the \fBpcre_copy_substring()\fR,  This error is used by the \fBpcre_copy_substring()\fP,
1644  \fBpcre_get_substring()\fR, and \fBpcre_get_substring_list()\fR functions (see  \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
1645  below). It is never returned by \fBpcre_exec()\fR.  below). It is never returned by \fBpcre_exec()\fP.
1646    .sp
1647    PCRE_ERROR_MATCHLIMIT     (-8)    PCRE_ERROR_MATCHLIMIT     (-8)
1648    .sp
1649  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
1650  field in a \fBpcre_extra\fR structure (or defaulted) was reached. See the  \fBpcre_extra\fP structure (or defaulted) was reached. See the description
1651  description above.  above.
1652    .sp
1653    PCRE_ERROR_CALLOUT        (-9)    PCRE_ERROR_CALLOUT        (-9)
1654    .sp
1655  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
1656  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
1657  \fBpcrecallout\fR documentation for details.  .\" HREF
1658    \fBpcrecallout\fP
1659    PCRE_ERROR_BADUTF8       (-10)  .\"
1660    documentation for details.
1661    .sp
1662      PCRE_ERROR_BADUTF8        (-10)
1663    .sp
1664  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.
1665    .sp
1666  .SH EXTRACTING CAPTURED SUBSTRINGS BY NUMBER    PCRE_ERROR_BADUTF8_OFFSET (-11)
1667    .sp
1668    The UTF-8 byte sequence that was passed as a subject was valid, but the value
1669    of \fIstartoffset\fP did not point to the beginning of a UTF-8 character.
1670    .sp
1671      PCRE_ERROR_PARTIAL        (-12)
1672    .sp
1673    The subject string did not match, but it did match partially. See the
1674    .\" HREF
1675    \fBpcrepartial\fP
1676    .\"
1677    documentation for details of partial matching.
1678    .sp
1679      PCRE_ERROR_BADPARTIAL     (-13)
1680    .sp
1681    This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1682    option was used with a compiled pattern containing items that were not
1683    supported for partial matching. From release 8.00 onwards, there are no
1684    restrictions on partial matching.
1685    .sp
1686      PCRE_ERROR_INTERNAL       (-14)
1687    .sp
1688    An unexpected internal error has occurred. This error could be caused by a bug
1689    in PCRE or by overwriting of the compiled pattern.
1690    .sp
1691      PCRE_ERROR_BADCOUNT       (-15)
1692    .sp
1693    This error is given if the value of the \fIovecsize\fP argument is negative.
1694    .sp
1695      PCRE_ERROR_RECURSIONLIMIT (-21)
1696    .sp
1697    The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
1698    field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
1699    description above.
1700    .sp
1701      PCRE_ERROR_BADNEWLINE     (-23)
1702    .sp
1703    An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
1704    .P
1705    Error numbers -16 to -20 and -22 are not used by \fBpcre_exec()\fP.
1706    .
1707    .
1708    .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
1709  .rs  .rs
1710  .sp  .sp
1711  .B int pcre_copy_substring(const char *\fIsubject\fR, int *\fIovector\fR,  .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1712  .ti +5n  .ti +5n
1713  .B int \fIstringcount\fR, int \fIstringnumber\fR, char *\fIbuffer\fR,  .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
1714  .ti +5n  .ti +5n
1715  .B int \fIbuffersize\fR);  .B int \fIbuffersize\fP);
1716  .PP  .PP
1717  .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,  
1718  .ti +5n  .ti +5n
1719  .B int \fIstringcount\fR, int \fIstringnumber\fR,  .B int \fIstringcount\fP, int \fIstringnumber\fP,
1720  .ti +5n  .ti +5n
1721  .B const char **\fIstringptr\fR);  .B const char **\fIstringptr\fP);
1722  .PP  .PP
1723  .br  .B int pcre_get_substring_list(const char *\fIsubject\fP,
 .B int pcre_get_substring_list(const char *\fIsubject\fR,  
1724  .ti +5n  .ti +5n
1725  .B int *\fIovector\fR, int \fIstringcount\fR, "const char ***\fIlistptr\fR);"  .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
1726  .PP  .PP
1727  Captured substrings can be accessed directly by using the offsets returned by  Captured substrings can be accessed directly by using the offsets returned by
1728  \fBpcre_exec()\fR in \fIovector\fR. For convenience, the functions  \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
1729  \fBpcre_copy_substring()\fR, \fBpcre_get_substring()\fR, and  \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
1730  \fBpcre_get_substring_list()\fR are provided for extracting captured substrings  \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
1731  as new, separate, zero-terminated strings. These functions identify substrings  as new, separate, zero-terminated strings. These functions identify substrings
1732  by number. The next section describes functions for extracting named  by number. The next section describes functions for extracting named
1733  substrings. A substring that contains a binary zero is correctly extracted and  substrings.
1734  has a further zero added on the end, but the result is not, of course,  .P
1735  a C string.  A substring that contains a binary zero is correctly extracted and has a
1736    further zero added on the end, but the result is not, of course, a C string.
1737    However, you can process such a string by referring to the length that is
1738    returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
1739    Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
1740    for handling strings containing binary zeros, because the end of the final
1741    string is not independently indicated.
1742    .P
1743  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:
1744  \fIsubject\fR is the subject string which has just been successfully matched,  \fIsubject\fP is the subject string that has just been successfully matched,
1745  \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
1746  \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
1747  captured by the match, including the substring that matched the entire regular  captured by the match, including the substring that matched the entire regular
1748  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
1749  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
1750  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
1751  the vector divided by three.  number of elements in the vector divided by three.
1752    .P
1753  The functions \fBpcre_copy_substring()\fR and \fBpcre_get_substring()\fR  The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
1754  extract a single substring, whose number is given as \fIstringnumber\fR. A  extract a single substring, whose number is given as \fIstringnumber\fP. A
1755  value of zero extracts the substring that matched the entire pattern, while  value of zero extracts the substring that matched the entire pattern, whereas
1756  higher values extract the captured substrings. For \fBpcre_copy_substring()\fR,  higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
1757  the string is placed in \fIbuffer\fR, whose length is given by  the string is placed in \fIbuffer\fP, whose length is given by
1758  \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
1759  obtained via \fBpcre_malloc\fR, and its address is returned via  obtained via \fBpcre_malloc\fP, and its address is returned via
1760  \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
1761  including the terminating zero, or one of  including the terminating zero, or one of these error codes:
1762    .sp
1763    PCRE_ERROR_NOMEMORY       (-6)    PCRE_ERROR_NOMEMORY       (-6)
1764    .sp
1765  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
1766  memory failed for \fBpcre_get_substring()\fR.  memory failed for \fBpcre_get_substring()\fP.
1767    .sp
1768    PCRE_ERROR_NOSUBSTRING    (-7)    PCRE_ERROR_NOSUBSTRING    (-7)
1769    .sp
1770  There is no substring whose number is \fIstringnumber\fR.  There is no substring whose number is \fIstringnumber\fP.
1771    .P
1772  The \fBpcre_get_substring_list()\fR function extracts all available substrings  The \fBpcre_get_substring_list()\fP function extracts all available substrings
1773  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
1774  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
1775  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
1776  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
1777  function is zero if all went well, or  function is zero if all went well, or the error code
1778    .sp
1779    PCRE_ERROR_NOMEMORY       (-6)    PCRE_ERROR_NOMEMORY       (-6)
1780    .sp
1781  if the attempt to get the memory block failed.  if the attempt to get the memory block failed.
1782    .P
1783  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
1784  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
1785  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
1786  string. This can be distinguished from a genuine zero-length substring by  string. This can be distinguished from a genuine zero-length substring by
1787  inspecting the appropriate offset in \fIovector\fR, which is negative for unset  inspecting the appropriate offset in \fIovector\fP, which is negative for unset
1788  substrings.  substrings.
1789    .P
1790  The two convenience functions \fBpcre_free_substring()\fR and  The two convenience functions \fBpcre_free_substring()\fP and
1791  \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
1792  a previous call of \fBpcre_get_substring()\fR or  a previous call of \fBpcre_get_substring()\fP or
1793  \fBpcre_get_substring_list()\fR, respectively. They do nothing more than call  \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
1794  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
1795  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
1796  linked via a special interface to another programming language which cannot use  linked via a special interface to another programming language that cannot use
1797  \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
1798  provided.  provided.
1799    .
1800  .SH EXTRACTING CAPTURED SUBSTRINGS BY NAME  .
1801    .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
1802  .rs  .rs
1803  .sp  .sp
1804  .B int pcre_copy_named_substring(const pcre *\fIcode\fR,  .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
1805  .ti +5n  .ti +5n
1806  .B const char *\fIsubject\fR, int *\fIovector\fR,  .B const char *\fIname\fP);
1807    .PP
1808    .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
1809  .ti +5n  .ti +5n
1810  .B int \fIstringcount\fR, const char *\fIstringname\fR,  .B const char *\fIsubject\fP, int *\fIovector\fP,
1811  .ti +5n  .ti +5n
1812  .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,  
1813  .ti +5n  .ti +5n
1814  .B const char *\fIname\fR);  .B char *\fIbuffer\fP, int \fIbuffersize\fP);
1815  .PP  .PP
1816  .br  .B int pcre_get_named_substring(const pcre *\fIcode\fP,
 .B int pcre_get_named_substring(const pcre *\fIcode\fR,  
1817  .ti +5n  .ti +5n
1818  .B const char *\fIsubject\fR, int *\fIovector\fR,  .B const char *\fIsubject\fP, int *\fIovector\fP,
1819  .ti +5n  .ti +5n
1820  .B int \fIstringcount\fR, const char *\fIstringname\fR,  .B int \fIstringcount\fP, const char *\fIstringname\fP,
1821  .ti +5n  .ti +5n
1822  .B const char **\fIstringptr\fR);  .B const char **\fIstringptr\fP);
1823  .PP  .PP
1824  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.
1825  can be done by calling \fBpcre_get_stringnumber()\fR. The first argument is the  For example, for this pattern
1826  compiled pattern, and the second is the name. For example, for this pattern  .sp
1827      (a+)b(?<xxx>\ed+)...
1828    ab(?<xxx>\\d+)...  .sp
1829    the number of the subpattern called "xxx" is 2. If the name is known to be
1830  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
1831  extract the substring directly, or use one of the functions described in the  calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
1832  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
1833  whole job.  subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
1834    that name.
1835  Most of the arguments of \fIpcre_copy_named_substring()\fR and  .P
1836  \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
1837  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
1838  differences.  two functions that do the whole job.
1839    .P
1840    Most of the arguments of \fBpcre_copy_named_substring()\fP and
1841    \fBpcre_get_named_substring()\fP are the same as those for the similarly named
1842    functions that extract by number. As these are described in the previous
1843    section, they are not re-described here. There are just two differences:
1844    .P
1845  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
1846  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
1847  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
1848  translation table.  translation table.
1849    .P
1850  These functions call \fBpcre_get_stringnumber()\fR, and if it succeeds, they  These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
1851  then call \fIpcre_copy_substring()\fR or \fIpcre_get_substring()\fR, as  then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
1852  appropriate.  appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
1853    the behaviour may not be what you want (see the next section).
1854  .in 0  .P
1855  Last updated: 20 August 2003  \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
1856  .br  subpatterns with the same number, as described in the
1857  Copyright (c) 1997-2003 University of Cambridge.  .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1858    .\" </a>
1859    section on duplicate subpattern numbers
1860    .\"
1861    in the
1862    .\" HREF
1863    \fBpcrepattern\fP
1864    .\"
1865    page, you cannot use names to distinguish the different subpatterns, because
1866    names are not included in the compiled code. The matching process uses only
1867    numbers. For this reason, the use of different names for subpatterns of the
1868    same number causes an error at compile time.
1869    .
1870    .SH "DUPLICATE SUBPATTERN NAMES"
1871    .rs
1872    .sp
1873    .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
1874    .ti +5n
1875    .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
1876    .PP
1877    When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
1878    are not required to be unique. (Duplicate names are always allowed for
1879    subpatterns with the same number, created by using the (?| feature. Indeed, if
1880    such subpatterns are named, they are required to use the same names.)
1881    .P
1882    Normally, patterns with duplicate names are such that in any one match, only
1883    one of the named subpatterns participates. An example is shown in the
1884    .\" HREF
1885    \fBpcrepattern\fP
1886    .\"
1887    documentation.
1888    .P
1889    When duplicates are present, \fBpcre_copy_named_substring()\fP and
1890    \fBpcre_get_named_substring()\fP return the first substring corresponding to
1891    the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
1892    returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
1893    returns one of the numbers that are associated with the name, but it is not
1894    defined which it is.
1895    .P
1896    If you want to get full details of all captured substrings for a given name,
1897    you must use the \fBpcre_get_stringtable_entries()\fP function. The first
1898    argument is the compiled pattern, and the second is the name. The third and
1899    fourth are pointers to variables which are updated by the function. After it
1900    has run, they point to the first and last entries in the name-to-number table
1901    for the given name. The function itself returns the length of each entry, or
1902    PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
1903    described above in the section entitled \fIInformation about a pattern\fP.
1904    Given all the relevant entries for the name, you can extract each of their
1905    numbers, and hence the captured data, if any.
1906    .
1907    .
1908    .SH "FINDING ALL POSSIBLE MATCHES"
1909    .rs
1910    .sp
1911    The traditional matching function uses a similar algorithm to Perl, which stops
1912    when it finds the first match, starting at a given point in the subject. If you
1913    want to find all possible matches, or the longest possible match, consider
1914    using the alternative matching function (see below) instead. If you cannot use
1915    the alternative function, but still need to find all possible matches, you
1916    can kludge it up by making use of the callout facility, which is described in
1917    the
1918    .\" HREF
1919    \fBpcrecallout\fP
1920    .\"
1921    documentation.
1922    .P
1923    What you have to do is to insert a callout right at the end of the pattern.
1924    When your callout function is called, extract and save the current matched
1925    substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
1926    other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
1927    will yield PCRE_ERROR_NOMATCH.
1928    .
1929    .
1930    .\" HTML <a name="dfamatch"></a>
1931    .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
1932    .rs
1933    .sp
1934    .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1935    .ti +5n
1936    .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1937    .ti +5n
1938    .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
1939    .ti +5n
1940    .B int *\fIworkspace\fP, int \fIwscount\fP);
1941    .P
1942    The function \fBpcre_dfa_exec()\fP is called to match a subject string against
1943    a compiled pattern, using a matching algorithm that scans the subject string
1944    just once, and does not backtrack. This has different characteristics to the
1945    normal algorithm, and is not compatible with Perl. Some of the features of PCRE
1946    patterns are not supported. Nevertheless, there are times when this kind of
1947    matching can be useful. For a discussion of the two matching algorithms, and a
1948    list of features that \fBpcre_dfa_exec()\fP does not support, see the
1949    .\" HREF
1950    \fBpcrematching\fP
1951    .\"
1952    documentation.
1953    .P
1954    The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
1955    \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
1956    different way, and this is described below. The other common arguments are used
1957    in the same way as for \fBpcre_exec()\fP, so their description is not repeated
1958    here.
1959    .P
1960    The two additional arguments provide workspace for the function. The workspace
1961    vector should contain at least 20 elements. It is used for keeping track of
1962    multiple paths through the pattern tree. More workspace will be needed for
1963    patterns and subjects where there are a lot of potential matches.
1964    .P
1965    Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
1966    .sp
1967      int rc;
1968      int ovector[10];
1969      int wspace[20];
1970      rc = pcre_dfa_exec(
1971        re,             /* result of pcre_compile() */
1972        NULL,           /* we didn't study the pattern */
1973        "some string",  /* the subject string */
1974        11,             /* the length of the subject string */
1975        0,              /* start at offset 0 in the subject */
1976        0,              /* default options */
1977        ovector,        /* vector of integers for substring information */
1978        10,             /* number of elements (NOT size in bytes) */
1979        wspace,         /* working space vector */
1980        20);            /* number of elements (NOT size in bytes) */
1981    .
1982    .SS "Option bits for \fBpcre_dfa_exec()\fP"
1983    .rs
1984    .sp
1985    The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
1986    zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1987    PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1988    PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST,
1989    and PCRE_DFA_RESTART. All but the last four of these are exactly the same as
1990    for \fBpcre_exec()\fP, so their description is not repeated here.
1991    .sp
1992      PCRE_PARTIAL_HARD
1993      PCRE_PARTIAL_SOFT
1994    .sp
1995    These have the same general effect as they do for \fBpcre_exec()\fP, but the
1996    details are slightly different. When PCRE_PARTIAL_HARD is set for
1997    \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
1998    is reached and there is still at least one matching possibility that requires
1999    additional characters. This happens even if some complete matches have also
2000    been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2001    is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2002    there have been no complete matches, but there is still at least one matching
2003    possibility. The portion of the string that was inspected when the longest
2004    partial match was found is set as the first matching string in both cases.
2005    .sp
2006      PCRE_DFA_SHORTEST
2007    .sp
2008    Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2009    soon as it has found one match. Because of the way the alternative algorithm
2010    works, this is necessarily the shortest possible match at the first possible
2011    matching point in the subject string.
2012    .sp
2013      PCRE_DFA_RESTART
2014    .sp
2015    When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2016    again, with additional subject characters, and have it continue with the same
2017    match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2018    \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2019    before because data about the match so far is left in them after a partial
2020    match. There is more discussion of this facility in the
2021    .\" HREF
2022    \fBpcrepartial\fP
2023    .\"
2024    documentation.
2025    .
2026    .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2027    .rs
2028    .sp
2029    When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2030    substring in the subject. Note, however, that all the matches from one run of
2031    the function start at the same point in the subject. The shorter matches are
2032    all initial substrings of the longer matches. For example, if the pattern
2033    .sp
2034      <.*>
2035    .sp
2036    is matched against the string
2037    .sp
2038      This is <something> <something else> <something further> no more
2039    .sp
2040    the three matched strings are
2041    .sp
2042      <something>
2043      <something> <something else>
2044      <something> <something else> <something further>
2045    .sp
2046    On success, the yield of the function is a number greater than zero, which is
2047    the number of matched substrings. The substrings themselves are returned in
2048    \fIovector\fP. Each string uses two elements; the first is the offset to the
2049    start, and the second is the offset to the end. In fact, all the strings have
2050    the same start offset. (Space could have been saved by giving this only once,
2051    but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2052    returns data, even though the meaning of the strings is different.)
2053    .P
2054    The strings are returned in reverse order of length; that is, the longest
2055    matching string is given first. If there were too many matches to fit into
2056    \fIovector\fP, the yield of the function is zero, and the vector is filled with
2057    the longest matches.
2058    .
2059    .SS "Error returns from \fBpcre_dfa_exec()\fP"
2060    .rs
2061    .sp
2062    The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2063    Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2064    described
2065    .\" HTML <a href="#errorlist">
2066    .\" </a>
2067    above.
2068    .\"
2069    There are in addition the following errors that are specific to
2070    \fBpcre_dfa_exec()\fP:
2071    .sp
2072      PCRE_ERROR_DFA_UITEM      (-16)
2073    .sp
2074    This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2075    that it does not support, for instance, the use of \eC or a back reference.
2076    .sp
2077      PCRE_ERROR_DFA_UCOND      (-17)
2078    .sp
2079    This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2080    uses a back reference for the condition, or a test for recursion in a specific
2081    group. These are not supported.
2082    .sp
2083      PCRE_ERROR_DFA_UMLIMIT    (-18)
2084    .sp
2085    This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2086    block that contains a setting of the \fImatch_limit\fP field. This is not
2087    supported (it is meaningless).
2088    .sp
2089      PCRE_ERROR_DFA_WSSIZE     (-19)
2090    .sp
2091    This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2092    \fIworkspace\fP vector.
2093    .sp
2094      PCRE_ERROR_DFA_RECURSE    (-20)
2095    .sp
2096    When a recursive subpattern is processed, the matching function calls itself
2097    recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2098    error is given if the output vector is not large enough. This should be
2099    extremely rare, as a vector of size 1000 is used.
2100    .
2101    .
2102    .SH "SEE ALSO"
2103    .rs
2104    .sp
2105    \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
2106    \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
2107    \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
2108    .
2109    .
2110    .SH AUTHOR
2111    .rs
2112    .sp
2113    .nf
2114    Philip Hazel
2115    University Computing Service
2116    Cambridge CB2 3QH, England.
2117    .fi
2118    .
2119    .
2120    .SH REVISION
2121    .rs
2122    .sp
2123    .nf
2124    Last updated: 16 May 2010
2125    Copyright (c) 1997-2010 University of Cambridge.
2126    .fi

Legend:
Removed from v.71  
changed lines
  Added in v.518

  ViewVC Help
Powered by ViewVC 1.1.5