3 PCRE - Perl-compatible regular expressions.
4 .SH "SYNOPSIS OF POSIX API"
7 .B #include <pcreposix.h>
10 .B int regcomp(regex_t *\fIpreg\fP, const char *\fIpattern\fP,
14 .B int regexec(regex_t *\fIpreg\fP, const char *\fIstring\fP,
16 .B size_t \fInmatch\fP, regmatch_t \fIpmatch\fP[], int \fIeflags\fP);
18 .B size_t regerror(int \fIerrcode\fP, const regex_t *\fIpreg\fP,
20 .B char *\fIerrbuf\fP, size_t \fIerrbuf_size\fP);
22 .B void regfree(regex_t *\fIpreg\fP);
27 This set of functions provides a POSIX-style API to the PCRE regular expression
32 documentation for a description of PCRE's native API, which contains much
33 additional functionality.
35 The functions described here are just wrapper functions that ultimately call
36 the PCRE native API. Their prototypes are defined in the \fBpcreposix.h\fP
37 header file, and on Unix systems the library itself is called
38 \fBpcreposix.a\fP, so can be accessed by adding \fB-lpcreposix\fP to the
39 command for linking an application that uses them. Because the POSIX functions
40 call the native ones, it is also necessary to add \fB-lpcre\fP.
42 I have implemented only those POSIX option bits that can be reasonably mapped
43 to PCRE native options. In addition, the option REG_EXTENDED is defined with
44 the value zero. This has no effect, but since programs that are written to the
45 POSIX interface often use it, this makes it easier to slot in PCRE as a
46 replacement library. Other POSIX options are not even defined.
48 There are also some other options that are not defined by POSIX. These have
49 been added at the request of users who want to make use of certain
50 PCRE-specific features via the POSIX calling interface.
52 When PCRE is called via these functions, it is only the API that is POSIX-like
53 in style. The syntax and semantics of the regular expressions themselves are
54 still those of Perl, subject to the setting of various PCRE options, as
55 described below. "POSIX-like in style" means that the API approximates to the
56 POSIX definition; it is not fully POSIX-compatible, and in multi-byte encoding
57 domains it is probably even less compatible.
59 The header for these functions is supplied as \fBpcreposix.h\fP to avoid any
60 potential clash with other POSIX libraries. It can, of course, be renamed or
61 aliased as \fBregex.h\fP, which is the "correct" name. It provides two
62 structure types, \fIregex_t\fP for compiled internal forms, and
63 \fIregmatch_t\fP for returning captured substrings. It also defines some
64 constants whose names start with "REG_"; these are used for setting options and
65 identifying error codes.
68 .SH "COMPILING A PATTERN"
71 The function \fBregcomp()\fP is called to compile a pattern into an
72 internal form. The pattern is a C string terminated by a binary zero, and
73 is passed in the argument \fIpattern\fP. The \fIpreg\fP argument is a pointer
74 to a \fBregex_t\fP structure that is used as a base for storing information
75 about the compiled regular expression.
77 The argument \fIcflags\fP is either zero, or contains one or more of the bits
78 defined by the following macros:
82 The PCRE_DOTALL option is set when the regular expression is passed for
83 compilation to the native function. Note that REG_DOTALL is not part of the
88 The PCRE_CASELESS option is set when the regular expression is passed for
89 compilation to the native function.
93 The PCRE_MULTILINE option is set when the regular expression is passed for
94 compilation to the native function. Note that this does \fInot\fP mimic the
95 defined POSIX behaviour for REG_NEWLINE (see the following section).
99 The PCRE_NO_AUTO_CAPTURE option is set when the regular expression is passed
100 for compilation to the native function. In addition, when a pattern that is
101 compiled with this flag is passed to \fBregexec()\fP for matching, the
102 \fInmatch\fP and \fIpmatch\fP arguments are ignored, and no captured strings
107 The PCRE_UCP option is set when the regular expression is passed for
108 compilation to the native function. This causes PCRE to use Unicode properties
109 when matchine \ed, \ew, etc., instead of just recognizing ASCII values. Note
110 that REG_UTF8 is not part of the POSIX standard.
114 The PCRE_UNGREEDY option is set when the regular expression is passed for
115 compilation to the native function. Note that REG_UNGREEDY is not part of the
120 The PCRE_UTF8 option is set when the regular expression is passed for
121 compilation to the native function. This causes the pattern itself and all data
122 strings used for matching it to be treated as UTF-8 strings. Note that REG_UTF8
123 is not part of the POSIX standard.
125 In the absence of these flags, no options are passed to the native function.
126 This means the the regex is compiled with PCRE default semantics. In
127 particular, the way it handles newline characters in the subject string is the
128 Perl way, not the POSIX way. Note that setting PCRE_MULTILINE has only
129 \fIsome\fP of the effects specified for REG_NEWLINE. It does not affect the way
130 newlines are matched by . (they are not) or by a negative class such as [^a]
133 The yield of \fBregcomp()\fP is zero on success, and non-zero otherwise. The
134 \fIpreg\fP structure is filled in on success, and one member of the structure
135 is public: \fIre_nsub\fP contains the number of capturing subpatterns in
136 the regular expression. Various error codes are defined in the header file.
138 NOTE: If the yield of \fBregcomp()\fP is non-zero, you must not attempt to
139 use the contents of the \fIpreg\fP structure. If, for example, you pass it to
140 \fBregexec()\fP, the result is undefined and your program is likely to crash.
143 .SH "MATCHING NEWLINE CHARACTERS"
146 This area is not simple, because POSIX and Perl take different views of things.
147 It is not possible to get PCRE to obey POSIX semantics, but then PCRE was never
148 intended to be a POSIX engine. The following table lists the different
149 possibilities for matching newline characters in PCRE:
153 . matches newline no PCRE_DOTALL
154 newline matches [^a] yes not changeable
155 $ matches \en at end yes PCRE_DOLLARENDONLY
156 $ matches \en in middle no PCRE_MULTILINE
157 ^ matches \en in middle no PCRE_MULTILINE
159 This is the equivalent table for POSIX:
163 . matches newline yes REG_NEWLINE
164 newline matches [^a] yes REG_NEWLINE
165 $ matches \en at end no REG_NEWLINE
166 $ matches \en in middle no REG_NEWLINE
167 ^ matches \en in middle no REG_NEWLINE
169 PCRE's behaviour is the same as Perl's, except that there is no equivalent for
170 PCRE_DOLLAR_ENDONLY in Perl. In both PCRE and Perl, there is no way to stop
171 newline from matching [^a].
173 The default POSIX newline handling can be obtained by setting PCRE_DOTALL and
174 PCRE_DOLLAR_ENDONLY, but there is no way to make PCRE behave exactly as for the
178 .SH "MATCHING A PATTERN"
181 The function \fBregexec()\fP is called to match a compiled pattern \fIpreg\fP
182 against a given \fIstring\fP, which is by default terminated by a zero byte
183 (but see REG_STARTEND below), subject to the options in \fIeflags\fP. These can
188 The PCRE_NOTBOL option is set when calling the underlying PCRE matching
193 The PCRE_NOTEMPTY option is set when calling the underlying PCRE matching
194 function. Note that REG_NOTEMPTY is not part of the POSIX standard. However,
195 setting this option can give more POSIX-like behaviour in some situations.
199 The PCRE_NOTEOL option is set when calling the underlying PCRE matching
204 The string is considered to start at \fIstring\fP + \fIpmatch[0].rm_so\fP and
205 to have a terminating NUL located at \fIstring\fP + \fIpmatch[0].rm_eo\fP
206 (there need not actually be a NUL at that location), regardless of the value of
207 \fInmatch\fP. This is a BSD extension, compatible with but not specified by
208 IEEE Standard 1003.2 (POSIX.2), and should be used with caution in software
209 intended to be portable to other systems. Note that a non-zero \fIrm_so\fP does
210 not imply REG_NOTBOL; REG_STARTEND affects only the location of the string, not
213 If the pattern was compiled with the REG_NOSUB flag, no data about any matched
214 strings is returned. The \fInmatch\fP and \fIpmatch\fP arguments of
215 \fBregexec()\fP are ignored.
217 If the value of \fInmatch\fP is zero, or if the value \fIpmatch\fP is NULL,
218 no data about any matched strings is returned.
220 Otherwise,the portion of the string that was matched, and also any captured
221 substrings, are returned via the \fIpmatch\fP argument, which points to an
222 array of \fInmatch\fP structures of type \fIregmatch_t\fP, containing the
223 members \fIrm_so\fP and \fIrm_eo\fP. These contain the offset to the first
224 character of each substring and the offset to the first character after the end
225 of each substring, respectively. The 0th element of the vector relates to the
226 entire portion of \fIstring\fP that was matched; subsequent elements relate to
227 the capturing subpatterns of the regular expression. Unused entries in the
228 array have both structure members set to -1.
230 A successful match yields a zero return; various error codes are defined in the
231 header file, of which REG_NOMATCH is the "expected" failure code.
237 The \fBregerror()\fP function maps a non-zero errorcode from either
238 \fBregcomp()\fP or \fBregexec()\fP to a printable message. If \fIpreg\fP is not
239 NULL, the error should have arisen from the use of that structure. A message
240 terminated by a binary zero is placed in \fIerrbuf\fP. The length of the
241 message, including the zero, is limited to \fIerrbuf_size\fP. The yield of the
242 function is the size of buffer needed to hold the whole message.
248 Compiling a regular expression causes memory to be allocated and associated
249 with the \fIpreg\fP structure. The function \fBregfree()\fP frees all such
250 memory, after which \fIpreg\fP may no longer be used as a compiled expression.
258 University Computing Service
259 Cambridge CB2 3QH, England.
267 Last updated: 16 May 2010
268 Copyright (c) 1997-2010 University of Cambridge.