trunk/doc/pcreperform.3

.TH PCREPERFORM 3
.SH NAME
PCRE - Perl-compatible regular expressions
.SH "PCRE PERFORMANCE"
.rs
.sp
Two aspects of performance are discussed below: memory usage and processing
time. The way you express your pattern as a regular expression can affect both
of them.
.
.SH "MEMORY USAGE"
.rs
.sp
Patterns are compiled by PCRE into a reasonably efficient byte code, so that
most simple patterns do not use much memory. However, there is one case where
memory usage can be unexpectedly large. When a parenthesized subpattern has a
quantifier with a minimum greater than 1 and/or a limited maximum, the whole
subpattern is repeated in the compiled code. For example, the pattern
.sp
  (abc|def){2,4}
.sp
is compiled as if it were
.sp
  (abc|def)(abc|def)((abc|def)(abc|def)?)?
.sp
(Technical aside: It is done this way so that backtrack points within each of
the repetitions can be independently maintained.)
.P
For regular expressions whose quantifiers use only small numbers, this is not
usually a problem. However, if the numbers are large, and particularly if such
repetitions are nested, the memory usage can become an embarrassment. For
example, the very simple pattern
.sp
  ((ab){1,1000}c){1,3}
.sp
uses 51K bytes when compiled. When PCRE is compiled with its default internal
pointer size of two bytes, the size limit on a compiled pattern is 64K, and
this is reached with the above pattern if the outer repetition is increased
from 3 to 4. PCRE can be compiled to use larger internal pointers and thus
handle larger compiled patterns, but it is better to try to rewrite your
pattern to use less memory if you can.
.P
One way of reducing the memory usage for such patterns is to make use of PCRE's
.\" HTML <a href="pcrepattern.html#subpatternsassubroutines">
.\" </a>
"subroutine"
.\"
facility. Re-writing the above pattern as
.sp
  ((ab)(?2){0,999}c)(?1){0,2}
.sp
reduces the memory requirements to 18K, and indeed it remains under 20K even
with the outer repetition increased to 100. However, this pattern is not
exactly equivalent, because the "subroutine" calls are treated as
.\" HTML <a href="pcrepattern.html#atomicgroup">
.\" </a>
atomic groups
.\"
into which there can be no backtracking if there is a subsequent matching
failure. Therefore, PCRE cannot do this kind of rewriting automatically.
Furthermore, there is a noticeable loss of speed when executing the modified
pattern. Nevertheless, if the atomic grouping is not a problem and the loss of
speed is acceptable, this kind of rewriting will allow you to process patterns
that PCRE cannot otherwise handle.
.
.SH "PROCESSING TIME"
.rs
.sp
Certain items in regular expression patterns are processed more efficiently
than others. It is more efficient to use a character class like [aeiou] than a
set of single-character alternatives such as (a|e|i|o|u). In general, the
simplest construction that provides the required behaviour is usually the most
efficient. Jeffrey Friedl's book contains a lot of useful general discussion
about optimizing regular expressions for efficient performance. This document
contains a few observations about PCRE.
.P
Using Unicode character properties (the \ep, \eP, and \eX escapes) is slow,
because PCRE has to scan a structure that contains data for over fifteen
thousand characters whenever it needs a character's property. If you can find
an alternative pattern that does not use character properties, it will probably
be faster.
.P
When a pattern begins with .* not in parentheses, or in parentheses that are
not the subject of a backreference, and the PCRE_DOTALL option is set, the
pattern is implicitly anchored by PCRE, since it can match only at the start of
a subject string. However, if PCRE_DOTALL is not set, PCRE cannot make this
optimization, because the . metacharacter does not then match a newline, and if
the subject string contains newlines, the pattern may match from the character
immediately following one of them instead of from the very start. For example,
the pattern
.sp
  .*second
.sp
matches the subject "first\enand second" (where \en stands for a newline
character), with the match starting at the seventh character. In order to do
this, PCRE has to retry the match starting after every newline in the subject.
.P
If you are using such a pattern with subject strings that do not contain
newlines, the best performance is obtained by setting PCRE_DOTALL, or starting
the pattern with ^.* or ^.*? to indicate explicit anchoring. That saves PCRE
from having to scan along the subject looking for a newline to restart at.
.P
Beware of patterns that contain nested indefinite repeats. These can take a
long time to run when applied to a string that does not match. Consider the
pattern fragment
.sp
  ^(a+)*
.sp
This can match "aaaa" in 16 different ways, and this number increases very
rapidly as the string gets longer. (The * repeat can match 0, 1, 2, 3, or 4
times, and for each of those cases other than 0 or 4, the + repeats can match
different numbers of times.) When the remainder of the pattern is such that the
entire match is going to fail, PCRE has in principle to try every possible
variation, and this can take an extremely long time, even for relatively short
strings.
.P
An optimization catches some of the more simple cases such as
.sp
  (a+)*b
.sp
where a literal character follows. Before embarking on the standard matching
procedure, PCRE checks that there is a "b" later in the subject string, and if
there is not, it fails the match immediately. However, when there is no
following literal this optimization cannot be used. You can see the difference
by comparing the behaviour of
.sp
  (a+)*\ed
.sp
with the pattern above. The former gives a failure almost instantly when
applied to a whole line of "a" characters, whereas the latter takes an
appreciable time with strings longer than about 20 characters.
.P
In many cases, the solution to this kind of performance issue is to use an
atomic group or a possessive quantifier.
.
.
.SH AUTHOR
.rs
.sp
.nf
Philip Hazel
University Computing Service
Cambridge CB2 3QH, England.
.fi
.
.
.SH REVISION
.rs
.sp
.nf
Last updated: 06 March 2007
Copyright (c) 1997-2007 University of Cambridge.
.fi
1	.TH PCREPERFORM 3
2	.SH NAME
3	PCRE - Perl-compatible regular expressions
4	.SH "PCRE PERFORMANCE"
5	.rs
6	.sp
7	Two aspects of performance are discussed below: memory usage and processing
8	time. The way you express your pattern as a regular expression can affect both
9	of them.
10	.
11	.SH "MEMORY USAGE"
12	.rs
13	.sp
14	Patterns are compiled by PCRE into a reasonably efficient byte code, so that
15	most simple patterns do not use much memory. However, there is one case where
16	memory usage can be unexpectedly large. When a parenthesized subpattern has a
17	quantifier with a minimum greater than 1 and/or a limited maximum, the whole
18	subpattern is repeated in the compiled code. For example, the pattern
19	.sp
20	(abc\|def){2,4}
21	.sp
22	is compiled as if it were
23	.sp
24	(abc\|def)(abc\|def)((abc\|def)(abc\|def)?)?
25	.sp
26	(Technical aside: It is done this way so that backtrack points within each of
27	the repetitions can be independently maintained.)
28	.P
29	For regular expressions whose quantifiers use only small numbers, this is not
30	usually a problem. However, if the numbers are large, and particularly if such
31	repetitions are nested, the memory usage can become an embarrassment. For
32	example, the very simple pattern
33	.sp
34	((ab){1,1000}c){1,3}
35	.sp
36	uses 51K bytes when compiled. When PCRE is compiled with its default internal
37	pointer size of two bytes, the size limit on a compiled pattern is 64K, and
38	this is reached with the above pattern if the outer repetition is increased
39	from 3 to 4. PCRE can be compiled to use larger internal pointers and thus
40	handle larger compiled patterns, but it is better to try to rewrite your
41	pattern to use less memory if you can.
42	.P
43	One way of reducing the memory usage for such patterns is to make use of PCRE's
44	.\" HTML <a href="pcrepattern.html#subpatternsassubroutines">
45	.\" </a>
46	"subroutine"
47	.\"
48	facility. Re-writing the above pattern as
49	.sp
50	((ab)(?2){0,999}c)(?1){0,2}
51	.sp
52	reduces the memory requirements to 18K, and indeed it remains under 20K even
53	with the outer repetition increased to 100. However, this pattern is not
54	exactly equivalent, because the "subroutine" calls are treated as
55	.\" HTML <a href="pcrepattern.html#atomicgroup">
56	.\" </a>
57	atomic groups
58	.\"
59	into which there can be no backtracking if there is a subsequent matching
60	failure. Therefore, PCRE cannot do this kind of rewriting automatically.
61	Furthermore, there is a noticeable loss of speed when executing the modified
62	pattern. Nevertheless, if the atomic grouping is not a problem and the loss of
63	speed is acceptable, this kind of rewriting will allow you to process patterns
64	that PCRE cannot otherwise handle.
65	.
66	.SH "PROCESSING TIME"
67	.rs
68	.sp
69	Certain items in regular expression patterns are processed more efficiently
70	than others. It is more efficient to use a character class like [aeiou] than a
71	set of single-character alternatives such as (a\|e\|i\|o\|u). In general, the
72	simplest construction that provides the required behaviour is usually the most
73	efficient. Jeffrey Friedl's book contains a lot of useful general discussion
74	about optimizing regular expressions for efficient performance. This document
75	contains a few observations about PCRE.
76	.P
77	Using Unicode character properties (the \ep, \eP, and \eX escapes) is slow,
78	because PCRE has to scan a structure that contains data for over fifteen
79	thousand characters whenever it needs a character's property. If you can find
80	an alternative pattern that does not use character properties, it will probably
81	be faster.
82	.P
83	When a pattern begins with .* not in parentheses, or in parentheses that are
84	not the subject of a backreference, and the PCRE_DOTALL option is set, the
85	pattern is implicitly anchored by PCRE, since it can match only at the start of
86	a subject string. However, if PCRE_DOTALL is not set, PCRE cannot make this
87	optimization, because the . metacharacter does not then match a newline, and if
88	the subject string contains newlines, the pattern may match from the character
89	immediately following one of them instead of from the very start. For example,
90	the pattern
91	.sp
92	.*second
93	.sp
94	matches the subject "first\enand second" (where \en stands for a newline
95	character), with the match starting at the seventh character. In order to do
96	this, PCRE has to retry the match starting after every newline in the subject.
97	.P
98	If you are using such a pattern with subject strings that do not contain
99	newlines, the best performance is obtained by setting PCRE_DOTALL, or starting
100	the pattern with ^.* or ^.*? to indicate explicit anchoring. That saves PCRE
101	from having to scan along the subject looking for a newline to restart at.
102	.P
103	Beware of patterns that contain nested indefinite repeats. These can take a
104	long time to run when applied to a string that does not match. Consider the
105	pattern fragment
106	.sp
107	^(a+)*
108	.sp
109	This can match "aaaa" in 16 different ways, and this number increases very
110	rapidly as the string gets longer. (The * repeat can match 0, 1, 2, 3, or 4
111	times, and for each of those cases other than 0 or 4, the + repeats can match
112	different numbers of times.) When the remainder of the pattern is such that the
113	entire match is going to fail, PCRE has in principle to try every possible
114	variation, and this can take an extremely long time, even for relatively short
115	strings.
116	.P
117	An optimization catches some of the more simple cases such as
118	.sp
119	(a+)*b
120	.sp
121	where a literal character follows. Before embarking on the standard matching
122	procedure, PCRE checks that there is a "b" later in the subject string, and if
123	there is not, it fails the match immediately. However, when there is no
124	following literal this optimization cannot be used. You can see the difference
125	by comparing the behaviour of
126	.sp
127	(a+)*\ed
128	.sp
129	with the pattern above. The former gives a failure almost instantly when
130	applied to a whole line of "a" characters, whereas the latter takes an
131	appreciable time with strings longer than about 20 characters.
132	.P
133	In many cases, the solution to this kind of performance issue is to use an
134	atomic group or a possessive quantifier.
135	.
136	.
137	.SH AUTHOR
138	.rs
139	.sp
140	.nf
141	Philip Hazel
142	University Computing Service
143	Cambridge CB2 3QH, England.
144	.fi
145	.
146	.
147	.SH REVISION
148	.rs
149	.sp
150	.nf
151	Last updated: 06 March 2007
152	Copyright (c) 1997-2007 University of Cambridge.
153	.fi
Name	Value
svn:eol-style	native
svn:keywords	"Author Date Id Revision Url"