trunk/doc/pcreposix.3

.TH PCRE 3
.SH NAME
PCRE - Perl-compatible regular expressions.
.SH SYNOPSIS OF POSIX API
.B #include <pcreposix.h>
.PP
.SM
.br
.B int regcomp(regex_t *\fIpreg\fR, const char *\fIpattern\fR,
.ti +5n
.B int \fIcflags\fR);
.PP
.br
.B int regexec(regex_t *\fIpreg\fR, const char *\fIstring\fR,
.ti +5n
.B size_t \fInmatch\fR, regmatch_t \fIpmatch\fR[], int \fIeflags\fR);
.PP
.br
.B size_t regerror(int \fIerrcode\fR, const regex_t *\fIpreg\fR,
.ti +5n
.B char *\fIerrbuf\fR, size_t \fIerrbuf_size\fR);
.PP
.br
.B void regfree(regex_t *\fIpreg\fR);

.SH DESCRIPTION
.rs
.sp
This set of functions provides a POSIX-style API to the PCRE regular expression
package. See the
.\" HREF
\fBpcreapi\fR
.\"
documentation for a description of the native API, which contains additional
functionality.

The functions described here are just wrapper functions that ultimately call
the PCRE native API. Their prototypes are defined in the \fBpcreposix.h\fR
header file, and on Unix systems the library itself is called
\fBpcreposix.a\fR, so can be accessed by adding \fB-lpcreposix\fR to the
command for linking an application which uses them. Because the POSIX functions
call the native ones, it is also necessary to add \fR-lpcre\fR.

I have implemented only those option bits that can be reasonably mapped to PCRE
native options. In addition, the options REG_EXTENDED and REG_NOSUB are defined
with the value zero. They have no effect, but since programs that are written
to the POSIX interface often use them, this makes it easier to slot in PCRE as
a replacement library. Other POSIX options are not even defined.

When PCRE is called via these functions, it is only the API that is POSIX-like
in style. The syntax and semantics of the regular expressions themselves are
still those of Perl, subject to the setting of various PCRE options, as
described below.

The header for these functions is supplied as \fBpcreposix.h\fR to avoid any
potential clash with other POSIX libraries. It can, of course, be renamed or
aliased as \fBregex.h\fR, which is the "correct" name. It provides two
structure types, \fIregex_t\fR for compiled internal forms, and
\fIregmatch_t\fR for returning captured substrings. It also defines some
constants whose names start with "REG_"; these are used for setting options and
identifying error codes.

.SH COMPILING A PATTERN
.rs
.sp
The function \fBregcomp()\fR is called to compile a pattern into an
internal form. The pattern is a C string terminated by a binary zero, and
is passed in the argument \fIpattern\fR. The \fIpreg\fR argument is a pointer
to a regex_t structure which is used as a base for storing information about
the compiled expression.

The argument \fIcflags\fR is either zero, or contains one or more of the bits
defined by the following macros:

  REG_ICASE

The PCRE_CASELESS option is set when the expression is passed for compilation
to the native function.

  REG_NEWLINE

The PCRE_MULTILINE option is set when the expression is passed for compilation
to the native function. Note that this does \fInot\fR mimic the defined POSIX
behaviour for REG_NEWLINE (see the following section).

In the absence of these flags, no options are passed to the native function.
This means the the regex is compiled with PCRE default semantics. In
particular, the way it handles newline characters in the subject string is the
Perl way, not the POSIX way. Note that setting PCRE_MULTILINE has only
\fIsome\fR of the effects specified for REG_NEWLINE. It does not affect the way
newlines are matched by . (they aren't) or by a negative class such as [^a]
(they are).

The yield of \fBregcomp()\fR is zero on success, and non-zero otherwise. The
\fIpreg\fR structure is filled in on success, and one member of the structure
is public: \fIre_nsub\fR contains the number of capturing subpatterns in
the regular expression. Various error codes are defined in the header file.

.SH MATCHING NEWLINE CHARACTERS
.rs
.sp
This area is not simple, because POSIX and Perl take different views of things.
It is not possible to get PCRE to obey POSIX semantics, but then PCRE was never
intended to be a POSIX engine. The following table lists the different
possibilities for matching newline characters in PCRE:

                          Default   Change with

  . matches newline          no     PCRE_DOTALL
  newline matches [^a]       yes    not changeable
  $ matches \\n at end        yes    PCRE_DOLLARENDONLY
  $ matches \\n in middle     no     PCRE_MULTILINE
  ^ matches \\n in middle     no     PCRE_MULTILINE

This is the equivalent table for POSIX:

                          Default   Change with

  . matches newline          yes      REG_NEWLINE
  newline matches [^a]       yes      REG_NEWLINE
  $ matches \\n at end        no       REG_NEWLINE
  $ matches \\n in middle     no       REG_NEWLINE
  ^ matches \\n in middle     no       REG_NEWLINE

PCRE's behaviour is the same as Perl's, except that there is no equivalent for
PCRE_DOLLARENDONLY in Perl. In both PCRE and Perl, there is no way to stop
newline from matching [^a].

The default POSIX newline handling can be obtained by setting PCRE_DOTALL and
PCRE_DOLLARENDONLY, but there is no way to make PCRE behave exactly as for the
REG_NEWLINE action.

.SH MATCHING A PATTERN
.rs
.sp
The function \fBregexec()\fR is called to match a pre-compiled pattern
\fIpreg\fR against a given \fIstring\fR, which is terminated by a zero byte,
subject to the options in \fIeflags\fR. These can be:

  REG_NOTBOL

The PCRE_NOTBOL option is set when calling the underlying PCRE matching
function.

  REG_NOTEOL

The PCRE_NOTEOL option is set when calling the underlying PCRE matching
function.

The portion of the string that was matched, and also any captured substrings,
are returned via the \fIpmatch\fR argument, which points to an array of
\fInmatch\fR structures of type \fIregmatch_t\fR, containing the members
\fIrm_so\fR and \fIrm_eo\fR. These contain the offset to the first character of
each substring and the offset to the first character after the end of each
substring, respectively. The 0th element of the vector relates to the entire
portion of \fIstring\fR that was matched; subsequent elements relate to the
capturing subpatterns of the regular expression. Unused entries in the array
have both structure members set to -1.

A successful match yields a zero return; various error codes are defined in the
header file, of which REG_NOMATCH is the "expected" failure code.

.SH ERROR MESSAGES
.rs
.sp
The \fBregerror()\fR function maps a non-zero errorcode from either
\fBregcomp()\fR or \fBregexec()\fR to a printable message. If \fIpreg\fR is not
NULL, the error should have arisen from the use of that structure. A message
terminated by a binary zero is placed in \fIerrbuf\fR. The length of the
message, including the zero, is limited to \fIerrbuf_size\fR. The yield of the
function is the size of buffer needed to hold the whole message.

.SH STORAGE
.rs
.sp
Compiling a regular expression causes memory to be allocated and associated
with the \fIpreg\fR structure. The function \fBregfree()\fR frees all such
memory, after which \fIpreg\fR may no longer be used as a compiled expression.

.SH AUTHOR
.rs
.sp
Philip Hazel <ph10@cam.ac.uk>
.br
University Computing Service,
.br
Cambridge CB2 3QG, England.

.in 0
Last updated: 03 February 2003
.br
Copyright (c) 1997-2003 University of Cambridge.
1	nigel	41	.TH PCRE 3
2			.SH NAME
3	nigel	63	PCRE - Perl-compatible regular expressions.
4			.SH SYNOPSIS OF POSIX API
5	nigel	41	.B #include <pcreposix.h>
6			.PP
7			.SM
8			.br
9			.B int regcomp(regex_t \fIpreg\fR, const char \fIpattern\fR,
10			.ti +5n
11			.B int \fIcflags\fR);
12			.PP
13			.br
14			.B int regexec(regex_t \fIpreg\fR, const char \fIstring\fR,
15			.ti +5n
16			.B size_t \fInmatch\fR, regmatch_t \fIpmatch\fR[], int \fIeflags\fR);
17			.PP
18			.br
19			.B size_t regerror(int \fIerrcode\fR, const regex_t *\fIpreg\fR,
20			.ti +5n
21			.B char *\fIerrbuf\fR, size_t \fIerrbuf_size\fR);
22			.PP
23			.br
24			.B void regfree(regex_t *\fIpreg\fR);
25
26			.SH DESCRIPTION
27	nigel	63	.rs
28			.sp
29	nigel	41	This set of functions provides a POSIX-style API to the PCRE regular expression
30	nigel	63	package. See the
31			.\" HREF
32			\fBpcreapi\fR
33			.\"
34			documentation for a description of the native API, which contains additional
35			functionality.
36	nigel	41
37			The functions described here are just wrapper functions that ultimately call
38	nigel	63	the PCRE native API. Their prototypes are defined in the \fBpcreposix.h\fR
39			header file, and on Unix systems the library itself is called
40			\fBpcreposix.a\fR, so can be accessed by adding \fB-lpcreposix\fR to the
41			command for linking an application which uses them. Because the POSIX functions
42			call the native ones, it is also necessary to add \fR-lpcre\fR.
43	nigel	41
44	nigel	43	I have implemented only those option bits that can be reasonably mapped to PCRE
45			native options. In addition, the options REG_EXTENDED and REG_NOSUB are defined
46			with the value zero. They have no effect, but since programs that are written
47			to the POSIX interface often use them, this makes it easier to slot in PCRE as
48			a replacement library. Other POSIX options are not even defined.
49	nigel	41
50			When PCRE is called via these functions, it is only the API that is POSIX-like
51			in style. The syntax and semantics of the regular expressions themselves are
52			still those of Perl, subject to the setting of various PCRE options, as
53			described below.
54
55			The header for these functions is supplied as \fBpcreposix.h\fR to avoid any
56			potential clash with other POSIX libraries. It can, of course, be renamed or
57			aliased as \fBregex.h\fR, which is the "correct" name. It provides two
58			structure types, \fIregex_t\fR for compiled internal forms, and
59			\fIregmatch_t\fR for returning captured substrings. It also defines some
60			constants whose names start with "REG_"; these are used for setting options and
61			identifying error codes.
62
63			.SH COMPILING A PATTERN
64	nigel	63	.rs
65			.sp
66	nigel	41	The function \fBregcomp()\fR is called to compile a pattern into an
67			internal form. The pattern is a C string terminated by a binary zero, and
68			is passed in the argument \fIpattern\fR. The \fIpreg\fR argument is a pointer
69			to a regex_t structure which is used as a base for storing information about
70			the compiled expression.
71
72			The argument \fIcflags\fR is either zero, or contains one or more of the bits
73			defined by the following macros:
74
75			REG_ICASE
76
77			The PCRE_CASELESS option is set when the expression is passed for compilation
78			to the native function.
79
80			REG_NEWLINE
81
82			The PCRE_MULTILINE option is set when the expression is passed for compilation
83	nigel	63	to the native function. Note that this does \fInot\fR mimic the defined POSIX
84			behaviour for REG_NEWLINE (see the following section).
85	nigel	41
86	nigel	49	In the absence of these flags, no options are passed to the native function.
87			This means the the regex is compiled with PCRE default semantics. In
88			particular, the way it handles newline characters in the subject string is the
89			Perl way, not the POSIX way. Note that setting PCRE_MULTILINE has only
90			\fIsome\fR of the effects specified for REG_NEWLINE. It does not affect the way
91	nigel	63	newlines are matched by . (they aren't) or by a negative class such as [^a]
92			(they are).
93	nigel	49
94	nigel	41	The yield of \fBregcomp()\fR is zero on success, and non-zero otherwise. The
95			\fIpreg\fR structure is filled in on success, and one member of the structure
96	nigel	63	is public: \fIre_nsub\fR contains the number of capturing subpatterns in
97	nigel	41	the regular expression. Various error codes are defined in the header file.
98
99	nigel	63	.SH MATCHING NEWLINE CHARACTERS
100			.rs
101			.sp
102			This area is not simple, because POSIX and Perl take different views of things.
103			It is not possible to get PCRE to obey POSIX semantics, but then PCRE was never
104			intended to be a POSIX engine. The following table lists the different
105			possibilities for matching newline characters in PCRE:
106	nigel	41
107	nigel	63	Default Change with
108
109			. matches newline no PCRE_DOTALL
110			newline matches [^a] yes not changeable
111			$ matches \\n at end yes PCRE_DOLLARENDONLY
112			$ matches \\n in middle no PCRE_MULTILINE
113			^ matches \\n in middle no PCRE_MULTILINE
114
115			This is the equivalent table for POSIX:
116
117			Default Change with
118
119			. matches newline yes REG_NEWLINE
120			newline matches [^a] yes REG_NEWLINE
121			$ matches \\n at end no REG_NEWLINE
122			$ matches \\n in middle no REG_NEWLINE
123			^ matches \\n in middle no REG_NEWLINE
124
125			PCRE's behaviour is the same as Perl's, except that there is no equivalent for
126			PCRE_DOLLARENDONLY in Perl. In both PCRE and Perl, there is no way to stop
127			newline from matching [^a].
128
129			The default POSIX newline handling can be obtained by setting PCRE_DOTALL and
130			PCRE_DOLLARENDONLY, but there is no way to make PCRE behave exactly as for the
131			REG_NEWLINE action.
132
133	nigel	41	.SH MATCHING A PATTERN
134	nigel	63	.rs
135			.sp
136	nigel	41	The function \fBregexec()\fR is called to match a pre-compiled pattern
137			\fIpreg\fR against a given \fIstring\fR, which is terminated by a zero byte,
138			subject to the options in \fIeflags\fR. These can be:
139
140			REG_NOTBOL
141
142			The PCRE_NOTBOL option is set when calling the underlying PCRE matching
143			function.
144
145			REG_NOTEOL
146
147			The PCRE_NOTEOL option is set when calling the underlying PCRE matching
148			function.
149
150			The portion of the string that was matched, and also any captured substrings,
151			are returned via the \fIpmatch\fR argument, which points to an array of
152			\fInmatch\fR structures of type \fIregmatch_t\fR, containing the members
153			\fIrm_so\fR and \fIrm_eo\fR. These contain the offset to the first character of
154			each substring and the offset to the first character after the end of each
155			substring, respectively. The 0th element of the vector relates to the entire
156			portion of \fIstring\fR that was matched; subsequent elements relate to the
157			capturing subpatterns of the regular expression. Unused entries in the array
158			have both structure members set to -1.
159
160			A successful match yields a zero return; various error codes are defined in the
161			header file, of which REG_NOMATCH is the "expected" failure code.
162
163			.SH ERROR MESSAGES
164	nigel	63	.rs
165			.sp
166	nigel	41	The \fBregerror()\fR function maps a non-zero errorcode from either
167	nigel	63	\fBregcomp()\fR or \fBregexec()\fR to a printable message. If \fIpreg\fR is not
168	nigel	41	NULL, the error should have arisen from the use of that structure. A message
169			terminated by a binary zero is placed in \fIerrbuf\fR. The length of the
170			message, including the zero, is limited to \fIerrbuf_size\fR. The yield of the
171			function is the size of buffer needed to hold the whole message.
172
173			.SH STORAGE
174	nigel	63	.rs
175			.sp
176	nigel	41	Compiling a regular expression causes memory to be allocated and associated
177			with the \fIpreg\fR structure. The function \fBregfree()\fR frees all such
178			memory, after which \fIpreg\fR may no longer be used as a compiled expression.
179
180			.SH AUTHOR
181	nigel	63	.rs
182			.sp
183	nigel	41	Philip Hazel <ph10@cam.ac.uk>
184			.br
185			University Computing Service,
186			.br
187			Cambridge CB2 3QG, England.
188	nigel	63
189			.in 0
190			Last updated: 03 February 2003
191	nigel	41	.br
192	nigel	63	Copyright (c) 1997-2003 University of Cambridge.