/[pcre]/code/trunk/doc/html/pcrematching.html
ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 172 - (hide annotations) (download) (as text)
Tue Jun 5 10:40:13 2007 UTC (7 years, 3 months ago) by ph10
File MIME type: text/html
File size: 9520 byte(s)
Drastically reduce workspace used for alternatives in groups; also some 
trailing space removals for a test release.

1 nigel 77 <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 ph10 111 <p>
11 nigel 77 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 ph10 111 <br>
15 nigel 77 <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>
19 nigel 93 <li><a name="TOC4" href="#SEC4">THE ALTERNATIVE MATCHING ALGORITHM</a>
20     <li><a name="TOC5" href="#SEC5">ADVANTAGES OF THE ALTERNATIVE ALGORITHM</a>
21     <li><a name="TOC6" href="#SEC6">DISADVANTAGES OF THE ALTERNATIVE ALGORITHM</a>
22 ph10 99 <li><a name="TOC7" href="#SEC7">AUTHOR</a>
23     <li><a name="TOC8" href="#SEC8">REVISION</a>
24 nigel 77 </ul>
25     <br><a name="SEC1" href="#TOC1">PCRE MATCHING ALGORITHMS</a><br>
26     <P>
27     This document describes the two different algorithms that are available in PCRE
28     for matching a compiled regular expression against a given subject string. The
29     "standard" algorithm is the one provided by the <b>pcre_exec()</b> function.
30     This works in the same was as Perl's matching function, and provides a
31     Perl-compatible matching operation.
32     </P>
33     <P>
34     An alternative algorithm is provided by the <b>pcre_dfa_exec()</b> function;
35     this operates in a different way, and is not Perl-compatible. It has advantages
36     and disadvantages compared with the standard algorithm, and these are described
37     below.
38     </P>
39     <P>
40     When there is only one possible way in which a given subject string can match a
41     pattern, the two algorithms give the same answer. A difference arises, however,
42     when there are multiple possibilities. For example, if the pattern
43     <pre>
44     ^&#60;.*&#62;
45     </pre>
46     is matched against the string
47     <pre>
48     &#60;something&#62; &#60;something else&#62; &#60;something further&#62;
49     </pre>
50     there are three possible answers. The standard algorithm finds only one of
51 nigel 93 them, whereas the alternative algorithm finds all three.
52 nigel 77 </P>
53     <br><a name="SEC2" href="#TOC1">REGULAR EXPRESSIONS AS TREES</a><br>
54     <P>
55     The set of strings that are matched by a regular expression can be represented
56     as a tree structure. An unlimited repetition in the pattern makes the tree of
57     infinite size, but it is still a tree. Matching the pattern to a given subject
58     string (from a given starting point) can be thought of as a search of the tree.
59 nigel 91 There are two ways to search a tree: depth-first and breadth-first, and these
60     correspond to the two matching algorithms provided by PCRE.
61 nigel 77 </P>
62     <br><a name="SEC3" href="#TOC1">THE STANDARD MATCHING ALGORITHM</a><br>
63     <P>
64 ph10 148 In the terminology of Jeffrey Friedl's book "Mastering Regular
65     Expressions", the standard algorithm is an "NFA algorithm". It conducts a
66 nigel 77 depth-first search of the pattern tree. That is, it proceeds along a single
67     path through the tree, checking that the subject matches what is required. When
68     there is a mismatch, the algorithm tries any alternatives at the current point,
69     and if they all fail, it backs up to the previous branch point in the tree, and
70     tries the next alternative branch at that level. This often involves backing up
71     (moving to the left) in the subject string as well. The order in which
72     repetition branches are tried is controlled by the greedy or ungreedy nature of
73     the quantifier.
74     </P>
75     <P>
76     If a leaf node is reached, a matching string has been found, and at that point
77     the algorithm stops. Thus, if there is more than one possible match, this
78     algorithm returns the first one that it finds. Whether this is the shortest,
79     the longest, or some intermediate length depends on the way the greedy and
80     ungreedy repetition quantifiers are specified in the pattern.
81     </P>
82     <P>
83     Because it ends up with a single path through the tree, it is relatively
84     straightforward for this algorithm to keep track of the substrings that are
85     matched by portions of the pattern in parentheses. This provides support for
86     capturing parentheses and back references.
87     </P>
88 nigel 93 <br><a name="SEC4" href="#TOC1">THE ALTERNATIVE MATCHING ALGORITHM</a><br>
89 nigel 77 <P>
90 nigel 93 This algorithm conducts a breadth-first search of the tree. Starting from the
91     first matching point in the subject, it scans the subject string from left to
92     right, once, character by character, and as it does this, it remembers all the
93     paths through the tree that represent valid matches. In Friedl's terminology,
94     this is a kind of "DFA algorithm", though it is not implemented as a
95     traditional finite state machine (it keeps multiple states active
96     simultaneously).
97 nigel 77 </P>
98     <P>
99     The scan continues until either the end of the subject is reached, or there are
100     no more unterminated paths. At this point, terminated paths represent the
101     different matching possibilities (if there are none, the match has failed).
102     Thus, if there is more than one possible match, this algorithm finds all of
103     them, and in particular, it finds the longest. In PCRE, there is an option to
104     stop the algorithm after the first match (which is necessarily the shortest)
105     has been found.
106     </P>
107     <P>
108     Note that all the matches that are found start at the same point in the
109     subject. If the pattern
110     <pre>
111     cat(er(pillar)?)
112     </pre>
113     is matched against the string "the caterpillar catchment", the result will be
114     the three strings "cat", "cater", and "caterpillar" that start at the fourth
115     character of the subject. The algorithm does not automatically move on to find
116     matches that start at later positions.
117     </P>
118     <P>
119     There are a number of features of PCRE regular expressions that are not
120 nigel 93 supported by the alternative matching algorithm. They are as follows:
121 nigel 77 </P>
122     <P>
123     1. Because the algorithm finds all possible matches, the greedy or ungreedy
124     nature of repetition quantifiers is not relevant. Greedy and ungreedy
125 nigel 93 quantifiers are treated in exactly the same way. However, possessive
126     quantifiers can make a difference when what follows could also match what is
127     quantified, for example in a pattern like this:
128     <pre>
129     ^a++\w!
130     </pre>
131     This pattern matches "aaab!" but not "aaa!", which would be matched by a
132     non-possessive quantifier. Similarly, if an atomic group is present, it is
133     matched as if it were a standalone pattern at the current point, and the
134     longest match is then "locked in" for the rest of the overall pattern.
135 nigel 77 </P>
136     <P>
137     2. When dealing with multiple paths through the tree simultaneously, it is not
138     straightforward to keep track of captured substrings for the different matching
139     possibilities, and PCRE's implementation of this algorithm does not attempt to
140     do this. This means that no captured substrings are available.
141     </P>
142     <P>
143     3. Because no substrings are captured, back references within the pattern are
144     not supported, and cause errors if encountered.
145     </P>
146     <P>
147     4. For the same reason, conditional expressions that use a backreference as the
148 nigel 93 condition or test for a specific group recursion are not supported.
149 nigel 77 </P>
150     <P>
151 ph10 172 5. Because many paths through the tree may be active, the \K escape sequence,
152     which resets the start of the match when encountered (but may be on some paths
153     and not on others), is not supported. It causes an error if encountered.
154     </P>
155     <P>
156     6. Callouts are supported, but the value of the <i>capture_top</i> field is
157 nigel 77 always 1, and the value of the <i>capture_last</i> field is always -1.
158     </P>
159     <P>
160 ph10 172 7.
161 nigel 77 The \C escape sequence, which (in the standard algorithm) matches a single
162 nigel 93 byte, even in UTF-8 mode, is not supported because the alternative algorithm
163     moves through the subject string one character at a time, for all active paths
164 nigel 77 through the tree.
165     </P>
166 nigel 93 <br><a name="SEC5" href="#TOC1">ADVANTAGES OF THE ALTERNATIVE ALGORITHM</a><br>
167 nigel 77 <P>
168 nigel 93 Using the alternative matching algorithm provides the following advantages:
169 nigel 77 </P>
170     <P>
171     1. All possible matches (at a single point in the subject) are automatically
172     found, and in particular, the longest match is found. To find more than one
173     match using the standard algorithm, you have to do kludgy things with
174     callouts.
175     </P>
176     <P>
177     2. There is much better support for partial matching. The restrictions on the
178     content of the pattern that apply when using the standard algorithm for partial
179 nigel 93 matching do not apply to the alternative algorithm. For non-anchored patterns,
180     the starting position of a partial match is available.
181 nigel 77 </P>
182     <P>
183 nigel 93 3. Because the alternative algorithm scans the subject string just once, and
184     never needs to backtrack, it is possible to pass very long subject strings to
185     the matching function in several pieces, checking for partial matching each
186     time.
187 nigel 77 </P>
188 nigel 93 <br><a name="SEC6" href="#TOC1">DISADVANTAGES OF THE ALTERNATIVE ALGORITHM</a><br>
189 nigel 77 <P>
190 nigel 93 The alternative algorithm suffers from a number of disadvantages:
191 nigel 77 </P>
192     <P>
193     1. It is substantially slower than the standard algorithm. This is partly
194     because it has to search for all possible matches, but is also because it is
195     less susceptible to optimization.
196     </P>
197     <P>
198     2. Capturing parentheses and back references are not supported.
199     </P>
200     <P>
201 nigel 93 3. Although atomic groups are supported, their use does not provide the
202     performance advantage that it does for the standard algorithm.
203 nigel 77 </P>
204 ph10 99 <br><a name="SEC7" href="#TOC1">AUTHOR</a><br>
205 nigel 77 <P>
206 ph10 99 Philip Hazel
207 nigel 77 <br>
208 ph10 99 University Computing Service
209     <br>
210     Cambridge CB2 3QH, England.
211     <br>
212     </P>
213     <br><a name="SEC8" href="#TOC1">REVISION</a><br>
214     <P>
215 ph10 172 Last updated: 29 May 2007
216 ph10 99 <br>
217     Copyright &copy; 1997-2007 University of Cambridge.
218     <br>
219 nigel 77 <p>
220     Return to the <a href="index.html">PCRE index page</a>.
221     </p>

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12