trunk/doc/pcreprecompile.3

.TH PCREPRECOMPILE 3 "24 June 2012" "PCRE 8.30"
.SH NAME
PCRE - Perl-compatible regular expressions
.SH "SAVING AND RE-USING PRECOMPILED PCRE PATTERNS"
.rs
.sp
If you are running an application that uses a large number of regular
expression patterns, it may be useful to store them in a precompiled form
instead of having to compile them every time the application is run.
If you are not using any private character tables (see the
.\" HREF
\fBpcre_maketables()\fP
.\"
documentation), this is relatively straightforward. If you are using private
tables, it is a little bit more complicated. However, if you are using the
just-in-time optimization feature, it is not possible to save and reload the
JIT data.
.P
If you save compiled patterns to a file, you can copy them to a different host
and run them there. If the two hosts have different endianness (byte order),
you should run the \fBpcre[16|32]_pattern_to_host_byte_order()\fP function on the
new host before trying to match the pattern. The matching functions return
PCRE_ERROR_BADENDIANNESS if they detect a pattern with the wrong endianness.
.P
Compiling regular expressions with one version of PCRE for use with a different
version is not guaranteed to work and may cause crashes, and saving and
restoring a compiled pattern loses any JIT optimization data.
.
.
.SH "SAVING A COMPILED PATTERN"
.rs
.sp
The value returned by \fBpcre[16|32]_compile()\fP points to a single block of
memory that holds the compiled pattern and associated data. You can find the
length of this block in bytes by calling \fBpcre[16|32]_fullinfo()\fP with an
argument of PCRE_INFO_SIZE. You can then save the data in any appropriate
manner. Here is sample code for the 8-bit library that compiles a pattern and
writes it to a file. It assumes that the variable \fIfd\fP refers to a file
that is open for output:
.sp
  int erroroffset, rc, size;
  char *error;
  pcre *re;
.sp
  re = pcre_compile("my pattern", 0, &error, &erroroffset, NULL);
  if (re == NULL) { ... handle errors ... }
  rc = pcre_fullinfo(re, NULL, PCRE_INFO_SIZE, &size);
  if (rc < 0) { ... handle errors ... }
  rc = fwrite(re, 1, size, fd);
  if (rc != size) { ... handle errors ... }
.sp
In this example, the bytes that comprise the compiled pattern are copied
exactly. Note that this is binary data that may contain any of the 256 possible
byte values. On systems that make a distinction between binary and non-binary
data, be sure that the file is opened for binary output.
.P
If you want to write more than one pattern to a file, you will have to devise a
way of separating them. For binary data, preceding each pattern with its length
is probably the most straightforward approach. Another possibility is to write
out the data in hexadecimal instead of binary, one pattern to a line.
.P
Saving compiled patterns in a file is only one possible way of storing them for
later use. They could equally well be saved in a database, or in the memory of
some daemon process that passes them via sockets to the processes that want
them.
.P
If the pattern has been studied, it is also possible to save the normal study
data in a similar way to the compiled pattern itself. However, if the
PCRE_STUDY_JIT_COMPILE was used, the just-in-time data that is created cannot
be saved because it is too dependent on the current environment. When studying
generates additional information, \fBpcre[16|32]_study()\fP returns a pointer to a
\fBpcre[16|32]_extra\fP data block. Its format is defined in the
.\" HTML <a href="pcreapi.html#extradata">
.\" </a>
section on matching a pattern
.\"
in the
.\" HREF
\fBpcreapi\fP
.\"
documentation. The \fIstudy_data\fP field points to the binary study data, and
this is what you must save (not the \fBpcre[16|32]_extra\fP block itself). The
length of the study data can be obtained by calling \fBpcre[16|32]_fullinfo()\fP
with an argument of PCRE_INFO_STUDYSIZE. Remember to check that
\fBpcre[16|32]_study()\fP did return a non-NULL value before trying to save the
study data.
.
.
.SH "RE-USING A PRECOMPILED PATTERN"
.rs
.sp
Re-using a precompiled pattern is straightforward. Having reloaded it into main
memory, called \fBpcre[16|32]_pattern_to_host_byte_order()\fP if necessary,
you pass its pointer to \fBpcre[16|32]_exec()\fP or \fBpcre[16|32]_dfa_exec()\fP in
the usual way.
.P
However, if you passed a pointer to custom character tables when the pattern
was compiled (the \fItableptr\fP argument of \fBpcre[16|32]_compile()\fP), you
must now pass a similar pointer to \fBpcre[16|32]_exec()\fP or
\fBpcre[16|32]_dfa_exec()\fP, because the value saved with the compiled pattern
will obviously be nonsense. A field in a \fBpcre[16|32]_extra()\fP block is used
to pass this data, as described in the
.\" HTML <a href="pcreapi.html#extradata">
.\" </a>
section on matching a pattern
.\"
in the
.\" HREF
\fBpcreapi\fP
.\"
documentation.
.P
If you did not provide custom character tables when the pattern was compiled,
the pointer in the compiled pattern is NULL, which causes the matching
functions to use PCRE's internal tables. Thus, you do not need to take any
special action at run time in this case.
.P
If you saved study data with the compiled pattern, you need to create your own
\fBpcre[16|32]_extra\fP data block and set the \fIstudy_data\fP field to point to the
reloaded study data. You must also set the PCRE_EXTRA_STUDY_DATA bit in the
\fIflags\fP field to indicate that study data is present. Then pass the
\fBpcre[16|32]_extra\fP block to the matching function in the usual way. If the
pattern was studied for just-in-time optimization, that data cannot be saved,
and so is lost by a save/restore cycle.
.
.
.SH "COMPATIBILITY WITH DIFFERENT PCRE RELEASES"
.rs
.sp
In general, it is safest to recompile all saved patterns when you update to a
new PCRE release, though not all updates actually require this.
.
.
.
.SH AUTHOR
.rs
.sp
.nf
Philip Hazel
University Computing Service
Cambridge CB2 3QH, England.
.fi
.
.
.SH REVISION
.rs
.sp
.nf
Last updated: 24 June 2012
Copyright (c) 1997-2012 University of Cambridge.
.fi
1	chpe	1055	.TH PCREPRECOMPILE 3 "24 June 2012" "PCRE 8.30"
2	nigel	75	.SH NAME
3			PCRE - Perl-compatible regular expressions
4			.SH "SAVING AND RE-USING PRECOMPILED PCRE PATTERNS"
5			.rs
6			.sp
7			If you are running an application that uses a large number of regular
8			expression patterns, it may be useful to store them in a precompiled form
9			instead of having to compile them every time the application is run.
10			If you are not using any private character tables (see the
11			.\" HREF
12			\fBpcre_maketables()\fP
13			.\"
14			documentation), this is relatively straightforward. If you are using private
15	ph10	691	tables, it is a little bit more complicated. However, if you are using the
16	ph10	861	just-in-time optimization feature, it is not possible to save and reload the
17			JIT data.
18	nigel	75	.P
19			If you save compiled patterns to a file, you can copy them to a different host
20	ph10	861	and run them there. If the two hosts have different endianness (byte order),
21	chpe	1055	you should run the \fBpcre[16\|32]_pattern_to_host_byte_order()\fP function on the
22	ph10	861	new host before trying to match the pattern. The matching functions return
23			PCRE_ERROR_BADENDIANNESS if they detect a pattern with the wrong endianness.
24			.P
25			Compiling regular expressions with one version of PCRE for use with a different
26			version is not guaranteed to work and may cause crashes, and saving and
27			restoring a compiled pattern loses any JIT optimization data.
28	nigel	75	.
29			.
30			.SH "SAVING A COMPILED PATTERN"
31			.rs
32	ph10	583	.sp
33	chpe	1055	The value returned by \fBpcre[16\|32]_compile()\fP points to a single block of
34	ph10	861	memory that holds the compiled pattern and associated data. You can find the
35	chpe	1055	length of this block in bytes by calling \fBpcre[16\|32]_fullinfo()\fP with an
36	ph10	861	argument of PCRE_INFO_SIZE. You can then save the data in any appropriate
37			manner. Here is sample code for the 8-bit library that compiles a pattern and
38			writes it to a file. It assumes that the variable \fIfd\fP refers to a file
39			that is open for output:
40	nigel	75	.sp
41			int erroroffset, rc, size;
42			char *error;
43			pcre *re;
44			.sp
45			re = pcre_compile("my pattern", 0, &error, &erroroffset, NULL);
46			if (re == NULL) { ... handle errors ... }
47			rc = pcre_fullinfo(re, NULL, PCRE_INFO_SIZE, &size);
48			if (rc < 0) { ... handle errors ... }
49			rc = fwrite(re, 1, size, fd);
50			if (rc != size) { ... handle errors ... }
51			.sp
52			In this example, the bytes that comprise the compiled pattern are copied
53			exactly. Note that this is binary data that may contain any of the 256 possible
54			byte values. On systems that make a distinction between binary and non-binary
55			data, be sure that the file is opened for binary output.
56			.P
57			If you want to write more than one pattern to a file, you will have to devise a
58			way of separating them. For binary data, preceding each pattern with its length
59			is probably the most straightforward approach. Another possibility is to write
60			out the data in hexadecimal instead of binary, one pattern to a line.
61			.P
62			Saving compiled patterns in a file is only one possible way of storing them for
63			later use. They could equally well be saved in a database, or in the memory of
64			some daemon process that passes them via sockets to the processes that want
65			them.
66			.P
67	ph10	678	If the pattern has been studied, it is also possible to save the normal study
68	ph10	691	data in a similar way to the compiled pattern itself. However, if the
69	ph10	678	PCRE_STUDY_JIT_COMPILE was used, the just-in-time data that is created cannot
70			be saved because it is too dependent on the current environment. When studying
71	chpe	1055	generates additional information, \fBpcre[16\|32]_study()\fP returns a pointer to a
72			\fBpcre[16\|32]_extra\fP data block. Its format is defined in the
73	nigel	75	.\" HTML <a href="pcreapi.html#extradata">
74			.\" </a>
75			section on matching a pattern
76			.\"
77			in the
78			.\" HREF
79			\fBpcreapi\fP
80			.\"
81			documentation. The \fIstudy_data\fP field points to the binary study data, and
82	chpe	1055	this is what you must save (not the \fBpcre[16\|32]_extra\fP block itself). The
83			length of the study data can be obtained by calling \fBpcre[16\|32]_fullinfo()\fP
84	ph10	861	with an argument of PCRE_INFO_STUDYSIZE. Remember to check that
85	chpe	1055	\fBpcre[16\|32]_study()\fP did return a non-NULL value before trying to save the
86	ph10	861	study data.
87	nigel	75	.
88			.
89			.SH "RE-USING A PRECOMPILED PATTERN"
90			.rs
91			.sp
92			Re-using a precompiled pattern is straightforward. Having reloaded it into main
93	chpe	1055	memory, called \fBpcre[16\|32]_pattern_to_host_byte_order()\fP if necessary,
94			you pass its pointer to \fBpcre[16\|32]_exec()\fP or \fBpcre[16\|32]_dfa_exec()\fP in
95	ph10	861	the usual way.
96	nigel	75	.P
97			However, if you passed a pointer to custom character tables when the pattern
98	chpe	1055	was compiled (the \fItableptr\fP argument of \fBpcre[16\|32]_compile()\fP), you
99			must now pass a similar pointer to \fBpcre[16\|32]_exec()\fP or
100			\fBpcre[16\|32]_dfa_exec()\fP, because the value saved with the compiled pattern
101			will obviously be nonsense. A field in a \fBpcre[16\|32]_extra()\fP block is used
102	ph10	861	to pass this data, as described in the
103	nigel	75	.\" HTML <a href="pcreapi.html#extradata">
104			.\" </a>
105			section on matching a pattern
106			.\"
107			in the
108			.\" HREF
109			\fBpcreapi\fP
110			.\"
111			documentation.
112			.P
113			If you did not provide custom character tables when the pattern was compiled,
114	ph10	903	the pointer in the compiled pattern is NULL, which causes the matching
115	ph10	861	functions to use PCRE's internal tables. Thus, you do not need to take any
116			special action at run time in this case.
117	nigel	75	.P
118			If you saved study data with the compiled pattern, you need to create your own
119	chpe	1055	\fBpcre[16\|32]_extra\fP data block and set the \fIstudy_data\fP field to point to the
120	nigel	75	reloaded study data. You must also set the PCRE_EXTRA_STUDY_DATA bit in the
121			\fIflags\fP field to indicate that study data is present. Then pass the
122	chpe	1055	\fBpcre[16\|32]_extra\fP block to the matching function in the usual way. If the
123	ph10	861	pattern was studied for just-in-time optimization, that data cannot be saved,
124			and so is lost by a save/restore cycle.
125	nigel	75	.
126			.
127			.SH "COMPATIBILITY WITH DIFFERENT PCRE RELEASES"
128			.rs
129			.sp
130	ph10	182	In general, it is safest to recompile all saved patterns when you update to a
131	ph10	572	new PCRE release, though not all updates actually require this.
132	ph10	99	.
133			.
134			.
135			.SH AUTHOR
136			.rs
137			.sp
138			.nf
139			Philip Hazel
140			University Computing Service
141			Cambridge CB2 3QH, England.
142			.fi
143			.
144			.
145			.SH REVISION
146			.rs
147			.sp
148			.nf
149	chpe	1055	Last updated: 24 June 2012
150	ph10	861	Copyright (c) 1997-2012 University of Cambridge.
151	ph10	99	.fi
Name	Value
svn:eol-style	native
svn:keywords	"Author Date Id Revision Url"