ViewVC logotype

Contents of /code/trunk/doc/html/pcrematching.html

Parent Directory Parent Directory | Revision Log Revision Log

Revision 93 - (show annotations)
Sat Feb 24 21:41:42 2007 UTC (14 years, 1 month ago) by nigel
File MIME type: text/html
File size: 9007 byte(s)
Load pcre-7.0 into code/trunk.
1 <html>
2 <head>
3 <title>pcrematching specification</title>
4 </head>
5 <body bgcolor="#FFFFFF" text="#00005A" link="#0066FF" alink="#3399FF" vlink="#2222BB">
6 <h1>pcrematching man page</h1>
7 <p>
8 Return to the <a href="index.html">PCRE index page</a>.
9 </p>
10 <p>
11 This page is part of the PCRE HTML documentation. It was generated automatically
12 from the original man page. If there is any nonsense in it, please consult the
13 man page, in case the conversion went wrong.
14 <br>
15 <ul>
16 <li><a name="TOC1" href="#SEC1">PCRE MATCHING ALGORITHMS</a>
17 <li><a name="TOC2" href="#SEC2">REGULAR EXPRESSIONS AS TREES</a>
18 <li><a name="TOC3" href="#SEC3">THE STANDARD MATCHING ALGORITHM</a>
22 </ul>
23 <br><a name="SEC1" href="#TOC1">PCRE MATCHING ALGORITHMS</a><br>
24 <P>
25 This document describes the two different algorithms that are available in PCRE
26 for matching a compiled regular expression against a given subject string. The
27 "standard" algorithm is the one provided by the <b>pcre_exec()</b> function.
28 This works in the same was as Perl's matching function, and provides a
29 Perl-compatible matching operation.
30 </P>
31 <P>
32 An alternative algorithm is provided by the <b>pcre_dfa_exec()</b> function;
33 this operates in a different way, and is not Perl-compatible. It has advantages
34 and disadvantages compared with the standard algorithm, and these are described
35 below.
36 </P>
37 <P>
38 When there is only one possible way in which a given subject string can match a
39 pattern, the two algorithms give the same answer. A difference arises, however,
40 when there are multiple possibilities. For example, if the pattern
41 <pre>
42 ^&#60;.*&#62;
43 </pre>
44 is matched against the string
45 <pre>
46 &#60;something&#62; &#60;something else&#62; &#60;something further&#62;
47 </pre>
48 there are three possible answers. The standard algorithm finds only one of
49 them, whereas the alternative algorithm finds all three.
50 </P>
51 <br><a name="SEC2" href="#TOC1">REGULAR EXPRESSIONS AS TREES</a><br>
52 <P>
53 The set of strings that are matched by a regular expression can be represented
54 as a tree structure. An unlimited repetition in the pattern makes the tree of
55 infinite size, but it is still a tree. Matching the pattern to a given subject
56 string (from a given starting point) can be thought of as a search of the tree.
57 There are two ways to search a tree: depth-first and breadth-first, and these
58 correspond to the two matching algorithms provided by PCRE.
59 </P>
60 <br><a name="SEC3" href="#TOC1">THE STANDARD MATCHING ALGORITHM</a><br>
61 <P>
62 In the terminology of Jeffrey Friedl's book \fIMastering Regular
63 Expressions\fP, the standard algorithm is an "NFA algorithm". It conducts a
64 depth-first search of the pattern tree. That is, it proceeds along a single
65 path through the tree, checking that the subject matches what is required. When
66 there is a mismatch, the algorithm tries any alternatives at the current point,
67 and if they all fail, it backs up to the previous branch point in the tree, and
68 tries the next alternative branch at that level. This often involves backing up
69 (moving to the left) in the subject string as well. The order in which
70 repetition branches are tried is controlled by the greedy or ungreedy nature of
71 the quantifier.
72 </P>
73 <P>
74 If a leaf node is reached, a matching string has been found, and at that point
75 the algorithm stops. Thus, if there is more than one possible match, this
76 algorithm returns the first one that it finds. Whether this is the shortest,
77 the longest, or some intermediate length depends on the way the greedy and
78 ungreedy repetition quantifiers are specified in the pattern.
79 </P>
80 <P>
81 Because it ends up with a single path through the tree, it is relatively
82 straightforward for this algorithm to keep track of the substrings that are
83 matched by portions of the pattern in parentheses. This provides support for
84 capturing parentheses and back references.
85 </P>
86 <br><a name="SEC4" href="#TOC1">THE ALTERNATIVE MATCHING ALGORITHM</a><br>
87 <P>
88 This algorithm conducts a breadth-first search of the tree. Starting from the
89 first matching point in the subject, it scans the subject string from left to
90 right, once, character by character, and as it does this, it remembers all the
91 paths through the tree that represent valid matches. In Friedl's terminology,
92 this is a kind of "DFA algorithm", though it is not implemented as a
93 traditional finite state machine (it keeps multiple states active
94 simultaneously).
95 </P>
96 <P>
97 The scan continues until either the end of the subject is reached, or there are
98 no more unterminated paths. At this point, terminated paths represent the
99 different matching possibilities (if there are none, the match has failed).
100 Thus, if there is more than one possible match, this algorithm finds all of
101 them, and in particular, it finds the longest. In PCRE, there is an option to
102 stop the algorithm after the first match (which is necessarily the shortest)
103 has been found.
104 </P>
105 <P>
106 Note that all the matches that are found start at the same point in the
107 subject. If the pattern
108 <pre>
109 cat(er(pillar)?)
110 </pre>
111 is matched against the string "the caterpillar catchment", the result will be
112 the three strings "cat", "cater", and "caterpillar" that start at the fourth
113 character of the subject. The algorithm does not automatically move on to find
114 matches that start at later positions.
115 </P>
116 <P>
117 There are a number of features of PCRE regular expressions that are not
118 supported by the alternative matching algorithm. They are as follows:
119 </P>
120 <P>
121 1. Because the algorithm finds all possible matches, the greedy or ungreedy
122 nature of repetition quantifiers is not relevant. Greedy and ungreedy
123 quantifiers are treated in exactly the same way. However, possessive
124 quantifiers can make a difference when what follows could also match what is
125 quantified, for example in a pattern like this:
126 <pre>
127 ^a++\w!
128 </pre>
129 This pattern matches "aaab!" but not "aaa!", which would be matched by a
130 non-possessive quantifier. Similarly, if an atomic group is present, it is
131 matched as if it were a standalone pattern at the current point, and the
132 longest match is then "locked in" for the rest of the overall pattern.
133 </P>
134 <P>
135 2. When dealing with multiple paths through the tree simultaneously, it is not
136 straightforward to keep track of captured substrings for the different matching
137 possibilities, and PCRE's implementation of this algorithm does not attempt to
138 do this. This means that no captured substrings are available.
139 </P>
140 <P>
141 3. Because no substrings are captured, back references within the pattern are
142 not supported, and cause errors if encountered.
143 </P>
144 <P>
145 4. For the same reason, conditional expressions that use a backreference as the
146 condition or test for a specific group recursion are not supported.
147 </P>
148 <P>
149 5. Callouts are supported, but the value of the <i>capture_top</i> field is
150 always 1, and the value of the <i>capture_last</i> field is always -1.
151 </P>
152 <P>
153 6.
154 The \C escape sequence, which (in the standard algorithm) matches a single
155 byte, even in UTF-8 mode, is not supported because the alternative algorithm
156 moves through the subject string one character at a time, for all active paths
157 through the tree.
158 </P>
159 <br><a name="SEC5" href="#TOC1">ADVANTAGES OF THE ALTERNATIVE ALGORITHM</a><br>
160 <P>
161 Using the alternative matching algorithm provides the following advantages:
162 </P>
163 <P>
164 1. All possible matches (at a single point in the subject) are automatically
165 found, and in particular, the longest match is found. To find more than one
166 match using the standard algorithm, you have to do kludgy things with
167 callouts.
168 </P>
169 <P>
170 2. There is much better support for partial matching. The restrictions on the
171 content of the pattern that apply when using the standard algorithm for partial
172 matching do not apply to the alternative algorithm. For non-anchored patterns,
173 the starting position of a partial match is available.
174 </P>
175 <P>
176 3. Because the alternative algorithm scans the subject string just once, and
177 never needs to backtrack, it is possible to pass very long subject strings to
178 the matching function in several pieces, checking for partial matching each
179 time.
180 </P>
182 <P>
183 The alternative algorithm suffers from a number of disadvantages:
184 </P>
185 <P>
186 1. It is substantially slower than the standard algorithm. This is partly
187 because it has to search for all possible matches, but is also because it is
188 less susceptible to optimization.
189 </P>
190 <P>
191 2. Capturing parentheses and back references are not supported.
192 </P>
193 <P>
194 3. Although atomic groups are supported, their use does not provide the
195 performance advantage that it does for the standard algorithm.
196 </P>
197 <P>
198 Last updated: 24 November 2006
199 <br>
200 Copyright &copy; 1997-2006 University of Cambridge.
201 <p>
202 Return to the <a href="index.html">PCRE index page</a>.
203 </p>

  ViewVC Help
Powered by ViewVC 1.1.5