ViewVC logotype

Contents of /code/trunk/doc/pcrebuild.3

Parent Directory Parent Directory | Revision Log Revision Log

Revision 685 - (show annotations)
Tue Sep 6 15:02:07 2011 UTC (9 years, 10 months ago) by ph10
File size: 14099 byte(s)
Update pcregrep to use JIT by default with options to disable.
3 PCRE - Perl-compatible regular expressions
4 .
5 .
7 .rs
8 .sp
9 This document describes the optional features of PCRE that can be selected when
10 the library is compiled. It assumes use of the \fBconfigure\fP script, where
11 the optional features are selected or deselected by providing options to
12 \fBconfigure\fP before running the \fBmake\fP command. However, the same
13 options can be selected in both Unix-like and non-Unix-like environments using
14 the GUI facility of \fBcmake-gui\fP if you are using \fBCMake\fP instead of
15 \fBconfigure\fP to build PCRE.
16 .P
17 There is a lot more information about building PCRE in non-Unix-like
18 environments in the file called \fINON_UNIX_USE\fP, which is part of the PCRE
19 distribution. You should consult this file as well as the \fIREADME\fP file if
20 you are building in a non-Unix-like environment.
21 .P
22 The complete list of options for \fBconfigure\fP (which includes the standard
23 ones such as the selection of the installation directory) can be obtained by
24 running
25 .sp
26 ./configure --help
27 .sp
28 The following sections include descriptions of options whose names begin with
29 --enable or --disable. These settings specify changes to the defaults for the
30 \fBconfigure\fP command. Because of the way that \fBconfigure\fP works,
31 --enable and --disable always come in pairs, so the complementary option always
32 exists as well, but as it specifies the default, it is not described.
33 .
34 .
36 .rs
37 .sp
38 The PCRE building process uses \fBlibtool\fP to build both shared and static
39 Unix libraries by default. You can suppress one of these by adding one of
40 .sp
41 --disable-shared
42 --disable-static
43 .sp
44 to the \fBconfigure\fP command, as required.
45 .
46 .
47 .SH "C++ SUPPORT"
48 .rs
49 .sp
50 By default, the \fBconfigure\fP script will search for a C++ compiler and C++
51 header files. If it finds them, it automatically builds the C++ wrapper library
52 for PCRE. You can disable this by adding
53 .sp
54 --disable-cpp
55 .sp
56 to the \fBconfigure\fP command.
57 .
58 .
60 .rs
61 .sp
62 To build PCRE with support for UTF-8 Unicode character strings, add
63 .sp
64 --enable-utf8
65 .sp
66 to the \fBconfigure\fP command. Of itself, this does not make PCRE treat
67 strings as UTF-8. As well as compiling PCRE with this option, you also have
68 have to set the PCRE_UTF8 option when you call the \fBpcre_compile()\fP
69 or \fBpcre_compile2()\fP functions.
70 .P
71 If you set --enable-utf8 when compiling in an EBCDIC environment, PCRE expects
72 its input to be either ASCII or UTF-8 (depending on the runtime option). It is
73 not possible to support both EBCDIC and UTF-8 codes in the same version of the
74 library. Consequently, --enable-utf8 and --enable-ebcdic are mutually
75 exclusive.
76 .
77 .
79 .rs
80 .sp
81 UTF-8 support allows PCRE to process character values greater than 255 in the
82 strings that it handles. On its own, however, it does not provide any
83 facilities for accessing the properties of such characters. If you want to be
84 able to use the pattern escapes \eP, \ep, and \eX, which refer to Unicode
85 character properties, you must add
86 .sp
87 --enable-unicode-properties
88 .sp
89 to the \fBconfigure\fP command. This implies UTF-8 support, even if you have
90 not explicitly requested it.
91 .P
92 Including Unicode property support adds around 30K of tables to the PCRE
93 library. Only the general category properties such as \fILu\fP and \fINd\fP are
94 supported. Details are given in the
95 .\" HREF
96 \fBpcrepattern\fP
97 .\"
98 documentation.
99 .
100 .
102 .rs
103 .sp
104 Just-in-time compiler support is included in the build by specifying
105 .sp
106 --enable-jit
107 .sp
108 This support is available only for certain hardware architectures. If this
109 option is set for an unsupported architecture, a compile time error occurs.
110 See the
111 .\" HREF
112 \fBpcrejit\fP
113 .\"
114 documentation for a discussion of JIT usage. When JIT support is enabled,
115 pcregrep automatically makes use of it, unless you add
116 .sp
117 --disable-pcregrep-jit
118 .sp
119 to the "configure" command.
120 .
121 .
123 .rs
124 .sp
125 By default, PCRE interprets the linefeed (LF) character as indicating the end
126 of a line. This is the normal newline character on Unix-like systems. You can
127 compile PCRE to use carriage return (CR) instead, by adding
128 .sp
129 --enable-newline-is-cr
130 .sp
131 to the \fBconfigure\fP command. There is also a --enable-newline-is-lf option,
132 which explicitly specifies linefeed as the newline character.
133 .sp
134 Alternatively, you can specify that line endings are to be indicated by the two
135 character sequence CRLF. If you want this, add
136 .sp
137 --enable-newline-is-crlf
138 .sp
139 to the \fBconfigure\fP command. There is a fourth option, specified by
140 .sp
141 --enable-newline-is-anycrlf
142 .sp
143 which causes PCRE to recognize any of the three sequences CR, LF, or CRLF as
144 indicating a line ending. Finally, a fifth option, specified by
145 .sp
146 --enable-newline-is-any
147 .sp
148 causes PCRE to recognize any Unicode newline sequence.
149 .P
150 Whatever line ending convention is selected when PCRE is built can be
151 overridden when the library functions are called. At build time it is
152 conventional to use the standard for your operating system.
153 .
154 .
156 .rs
157 .sp
158 By default, the sequence \eR in a pattern matches any Unicode newline sequence,
159 whatever has been selected as the line ending sequence. If you specify
160 .sp
161 --enable-bsr-anycrlf
162 .sp
163 the default is changed so that \eR matches only CR, LF, or CRLF. Whatever is
164 selected when PCRE is built can be overridden when the library functions are
165 called.
166 .
167 .
169 .rs
170 .sp
171 When PCRE is called through the POSIX interface (see the
172 .\" HREF
173 \fBpcreposix\fP
174 .\"
175 documentation), additional working storage is required for holding the pointers
176 to capturing substrings, because PCRE requires three integers per substring,
177 whereas the POSIX interface provides only two. If the number of expected
178 substrings is small, the wrapper function uses space on the stack, because this
179 is faster than using \fBmalloc()\fP for each call. The default threshold above
180 which the stack is no longer used is 10; it can be changed by adding a setting
181 such as
182 .sp
183 --with-posix-malloc-threshold=20
184 .sp
185 to the \fBconfigure\fP command.
186 .
187 .
189 .rs
190 .sp
191 Within a compiled pattern, offset values are used to point from one part to
192 another (for example, from an opening parenthesis to an alternation
193 metacharacter). By default, two-byte values are used for these offsets, leading
194 to a maximum size for a compiled pattern of around 64K. This is sufficient to
195 handle all but the most gigantic patterns. Nevertheless, some people do want to
196 process truyl enormous patterns, so it is possible to compile PCRE to use
197 three-byte or four-byte offsets by adding a setting such as
198 .sp
199 --with-link-size=3
200 .sp
201 to the \fBconfigure\fP command. The value given must be 2, 3, or 4. Using
202 longer offsets slows down the operation of PCRE because it has to load
203 additional bytes when handling them.
204 .
205 .
207 .rs
208 .sp
209 When matching with the \fBpcre_exec()\fP function, PCRE implements backtracking
210 by making recursive calls to an internal function called \fBmatch()\fP. In
211 environments where the size of the stack is limited, this can severely limit
212 PCRE's operation. (The Unix environment does not usually suffer from this
213 problem, but it may sometimes be necessary to increase the maximum stack size.
214 There is a discussion in the
215 .\" HREF
216 \fBpcrestack\fP
217 .\"
218 documentation.) An alternative approach to recursion that uses memory from the
219 heap to remember data, instead of using recursive function calls, has been
220 implemented to work round the problem of limited stack size. If you want to
221 build a version of PCRE that works this way, add
222 .sp
223 --disable-stack-for-recursion
224 .sp
225 to the \fBconfigure\fP command. With this configuration, PCRE will use the
226 \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP variables to call memory
227 management functions. By default these point to \fBmalloc()\fP and
228 \fBfree()\fP, but you can replace the pointers so that your own functions are
229 used instead.
230 .P
231 Separate functions are provided rather than using \fBpcre_malloc\fP and
232 \fBpcre_free\fP because the usage is very predictable: the block sizes
233 requested are always the same, and the blocks are always freed in reverse
234 order. A calling program might be able to implement optimized functions that
235 perform better than \fBmalloc()\fP and \fBfree()\fP. PCRE runs noticeably more
236 slowly when built in this way. This option affects only the \fBpcre_exec()\fP
237 function; it is not relevant for \fBpcre_dfa_exec()\fP.
238 .
239 .
241 .rs
242 .sp
243 Internally, PCRE has a function called \fBmatch()\fP, which it calls repeatedly
244 (sometimes recursively) when matching a pattern with the \fBpcre_exec()\fP
245 function. By controlling the maximum number of times this function may be
246 called during a single matching operation, a limit can be placed on the
247 resources used by a single call to \fBpcre_exec()\fP. The limit can be changed
248 at run time, as described in the
249 .\" HREF
250 \fBpcreapi\fP
251 .\"
252 documentation. The default is 10 million, but this can be changed by adding a
253 setting such as
254 .sp
255 --with-match-limit=500000
256 .sp
257 to the \fBconfigure\fP command. This setting has no effect on the
258 \fBpcre_dfa_exec()\fP matching function.
259 .P
260 In some environments it is desirable to limit the depth of recursive calls of
261 \fBmatch()\fP more strictly than the total number of calls, in order to
262 restrict the maximum amount of stack (or heap, if --disable-stack-for-recursion
263 is specified) that is used. A second limit controls this; it defaults to the
264 value that is set for --with-match-limit, which imposes no additional
265 constraints. However, you can set a lower limit by adding, for example,
266 .sp
267 --with-match-limit-recursion=10000
268 .sp
269 to the \fBconfigure\fP command. This value can also be overridden at run time.
270 .
271 .
273 .rs
274 .sp
275 PCRE uses fixed tables for processing characters whose code values are less
276 than 256. By default, PCRE is built with a set of tables that are distributed
277 in the file \fIpcre_chartables.c.dist\fP. These tables are for ASCII codes
278 only. If you add
279 .sp
280 --enable-rebuild-chartables
281 .sp
282 to the \fBconfigure\fP command, the distributed tables are no longer used.
283 Instead, a program called \fBdftables\fP is compiled and run. This outputs the
284 source for new set of tables, created in the default locale of your C runtime
285 system. (This method of replacing the tables does not work if you are cross
286 compiling, because \fBdftables\fP is run on the local host. If you need to
287 create alternative tables when cross compiling, you will have to do so "by
288 hand".)
289 .
290 .
292 .rs
293 .sp
294 PCRE assumes by default that it will run in an environment where the character
295 code is ASCII (or Unicode, which is a superset of ASCII). This is the case for
296 most computer operating systems. PCRE can, however, be compiled to run in an
297 EBCDIC environment by adding
298 .sp
299 --enable-ebcdic
300 .sp
301 to the \fBconfigure\fP command. This setting implies
302 --enable-rebuild-chartables. You should only use it if you know that you are in
303 an EBCDIC environment (for example, an IBM mainframe operating system). The
304 --enable-ebcdic option is incompatible with --enable-utf8.
305 .
306 .
308 .rs
309 .sp
310 By default, \fBpcregrep\fP reads all files as plain text. You can build it so
311 that it recognizes files whose names end in \fB.gz\fP or \fB.bz2\fP, and reads
312 them with \fBlibz\fP or \fBlibbz2\fP, respectively, by adding one or both of
313 .sp
314 --enable-pcregrep-libz
315 --enable-pcregrep-libbz2
316 .sp
317 to the \fBconfigure\fP command. These options naturally require that the
318 relevant libraries are installed on your system. Configuration will fail if
319 they are not.
320 .
321 .
323 .rs
324 .sp
325 \fBpcregrep\fP uses an internal buffer to hold a "window" on the file it is
326 scanning, in order to be able to output "before" and "after" lines when it
327 finds a match. The size of the buffer is controlled by a parameter whose
328 default value is 20K. The buffer itself is three times this size, but because
329 of the way it is used for holding "before" lines, the longest line that is
330 guaranteed to be processable is the parameter size. You can change the default
331 parameter value by adding, for example,
332 .sp
333 --with-pcregrep-bufsize=50K
334 .sp
335 to the \fBconfigure\fP command. The caller of \fPpcregrep\fP can, however,
336 override this value by specifying a run-time option.
337 .
338 .
340 .rs
341 .sp
342 If you add
343 .sp
344 --enable-pcretest-libreadline
345 .sp
346 to the \fBconfigure\fP command, \fBpcretest\fP is linked with the
347 \fBlibreadline\fP library, and when its input is from a terminal, it reads it
348 using the \fBreadline()\fP function. This provides line-editing and history
349 facilities. Note that \fBlibreadline\fP is GPL-licensed, so if you distribute a
350 binary of \fBpcretest\fP linked in this way, there may be licensing issues.
351 .P
352 Setting this option causes the \fB-lreadline\fP option to be added to the
353 \fBpcretest\fP build. In many operating environments with a sytem-installed
354 \fBlibreadline\fP this is sufficient. However, in some environments (e.g.
355 if an unmodified distribution version of readline is in use), some extra
356 configuration may be necessary. The INSTALL file for \fBlibreadline\fP says
357 this:
358 .sp
359 "Readline uses the termcap functions, but does not link with the
360 termcap or curses library itself, allowing applications which link
361 with readline the to choose an appropriate library."
362 .sp
363 If your environment has not been set up so that an appropriate library is
364 automatically included, you may need to add something like
365 .sp
366 LIBS="-ncurses"
367 .sp
368 immediately before the \fBconfigure\fP command.
369 .
370 .
371 .SH "SEE ALSO"
372 .rs
373 .sp
374 \fBpcreapi\fP(3), \fBpcre_config\fP(3).
375 .
376 .
378 .rs
379 .sp
380 .nf
381 Philip Hazel
382 University Computing Service
383 Cambridge CB2 3QH, England.
384 .fi
385 .
386 .
388 .rs
389 .sp
390 .nf
391 Last updated: 06 September 2011
392 Copyright (c) 1997-2011 University of Cambridge.
393 .fi


Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

  ViewVC Help
Powered by ViewVC 1.1.5