1 <html> 2 <head> 3 <title>pcre2matching specification</title> 4 </head> 5 <body bgcolor="#FFFFFF" text="#00005A" link="#0066FF" alink="#3399FF" vlink="#2222BB"> 6 <h1>pcre2matching man page</h1> 7 <p> 8 Return to the <a href="index.html">PCRE2 index page</a>. 9 </p> 10 <p> 11 This page is part of the PCRE2 HTML documentation. It was generated 12 automatically from the original man page. If there is any nonsense in it, 13 please consult the man page, in case the conversion went wrong. 14 <br> 15 <ul> 16 <li><a name="TOC1" href="#SEC1">PCRE2 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 <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 <li><a name="TOC7" href="#SEC7">AUTHOR</a> 23 <li><a name="TOC8" href="#SEC8">REVISION</a> 24 </ul> 25 <br><a name="SEC1" href="#TOC1">PCRE2 MATCHING ALGORITHMS</a><br> 26 <P> 27 This document describes the two different algorithms that are available in 28 PCRE2 for matching a compiled regular expression against a given subject 29 string. The "standard" algorithm is the one provided by the <b>pcre2_match()</b> 30 function. This works in the same as as Perl's matching function, and provide a 31 Perl-compatible matching operation. The just-in-time (JIT) optimization that is 32 described in the 33 <a href="pcre2jit.html"><b>pcre2jit</b></a> 34 documentation is compatible with this function. 35 </P> 36 <P> 37 An alternative algorithm is provided by the <b>pcre2_dfa_match()</b> function; 38 it operates in a different way, and is not Perl-compatible. This alternative 39 has advantages and disadvantages compared with the standard algorithm, and 40 these are described below. 41 </P> 42 <P> 43 When there is only one possible way in which a given subject string can match a 44 pattern, the two algorithms give the same answer. A difference arises, however, 45 when there are multiple possibilities. For example, if the pattern 46 <pre> 47 ^<.*> 48 </pre> 49 is matched against the string 50 <pre> 51 <something> <something else> <something further> 52 </pre> 53 there are three possible answers. The standard algorithm finds only one of 54 them, whereas the alternative algorithm finds all three. 55 </P> 56 <br><a name="SEC2" href="#TOC1">REGULAR EXPRESSIONS AS TREES</a><br> 57 <P> 58 The set of strings that are matched by a regular expression can be represented 59 as a tree structure. An unlimited repetition in the pattern makes the tree of 60 infinite size, but it is still a tree. Matching the pattern to a given subject 61 string (from a given starting point) can be thought of as a search of the tree. 62 There are two ways to search a tree: depth-first and breadth-first, and these 63 correspond to the two matching algorithms provided by PCRE2. 64 </P> 65 <br><a name="SEC3" href="#TOC1">THE STANDARD MATCHING ALGORITHM</a><br> 66 <P> 67 In the terminology of Jeffrey Friedl's book "Mastering Regular Expressions", 68 the standard algorithm is an "NFA algorithm". It conducts a depth-first search 69 of the pattern tree. That is, it proceeds along a single path through the tree, 70 checking that the subject matches what is required. When there is a mismatch, 71 the algorithm tries any alternatives at the current point, and if they all 72 fail, it backs up to the previous branch point in the tree, and tries the next 73 alternative branch at that level. This often involves backing up (moving to the 74 left) in the subject string as well. The order in which repetition branches are 75 tried is controlled by the greedy or ungreedy nature of the quantifier. 76 </P> 77 <P> 78 If a leaf node is reached, a matching string has been found, and at that point 79 the algorithm stops. Thus, if there is more than one possible match, this 80 algorithm returns the first one that it finds. Whether this is the shortest, 81 the longest, or some intermediate length depends on the way the greedy and 82 ungreedy repetition quantifiers are specified in the pattern. 83 </P> 84 <P> 85 Because it ends up with a single path through the tree, it is relatively 86 straightforward for this algorithm to keep track of the substrings that are 87 matched by portions of the pattern in parentheses. This provides support for 88 capturing parentheses and back references. 89 </P> 90 <br><a name="SEC4" href="#TOC1">THE ALTERNATIVE MATCHING ALGORITHM</a><br> 91 <P> 92 This algorithm conducts a breadth-first search of the tree. Starting from the 93 first matching point in the subject, it scans the subject string from left to 94 right, once, character by character, and as it does this, it remembers all the 95 paths through the tree that represent valid matches. In Friedl's terminology, 96 this is a kind of "DFA algorithm", though it is not implemented as a 97 traditional finite state machine (it keeps multiple states active 98 simultaneously). 99 </P> 100 <P> 101 Although the general principle of this matching algorithm is that it scans the 102 subject string only once, without backtracking, there is one exception: when a 103 lookaround assertion is encountered, the characters following or preceding the 104 current point have to be independently inspected. 105 </P> 106 <P> 107 The scan continues until either the end of the subject is reached, or there are 108 no more unterminated paths. At this point, terminated paths represent the 109 different matching possibilities (if there are none, the match has failed). 110 Thus, if there is more than one possible match, this algorithm finds all of 111 them, and in particular, it finds the longest. The matches are returned in 112 decreasing order of length. There is an option to stop the algorithm after the 113 first match (which is necessarily the shortest) is found. 114 </P> 115 <P> 116 Note that all the matches that are found start at the same point in the 117 subject. If the pattern 118 <pre> 119 cat(er(pillar)?)? 120 </pre> 121 is matched against the string "the caterpillar catchment", the result is the 122 three strings "caterpillar", "cater", and "cat" that start at the fifth 123 character of the subject. The algorithm does not automatically move on to find 124 matches that start at later positions. 125 </P> 126 <P> 127 PCRE2's "auto-possessification" optimization usually applies to character 128 repeats at the end of a pattern (as well as internally). For example, the 129 pattern "a\d+" is compiled as if it were "a\d++" because there is no point 130 even considering the possibility of backtracking into the repeated digits. For 131 DFA matching, this means that only one possible match is found. If you really 132 do want multiple matches in such cases, either use an ungreedy repeat 133 ("a\d+?") or set the PCRE2_NO_AUTO_POSSESS option when compiling. 134 </P> 135 <P> 136 There are a number of features of PCRE2 regular expressions that are not 137 supported by the alternative matching algorithm. They are as follows: 138 </P> 139 <P> 140 1. Because the algorithm finds all possible matches, the greedy or ungreedy 141 nature of repetition quantifiers is not relevant (though it may affect 142 auto-possessification, as just described). During matching, greedy and ungreedy 143 quantifiers are treated in exactly the same way. However, possessive 144 quantifiers can make a difference when what follows could also match what is 145 quantified, for example in a pattern like this: 146 <pre> 147 ^a++\w! 148 </pre> 149 This pattern matches "aaab!" but not "aaa!", which would be matched by a 150 non-possessive quantifier. Similarly, if an atomic group is present, it is 151 matched as if it were a standalone pattern at the current point, and the 152 longest match is then "locked in" for the rest of the overall pattern. 153 </P> 154 <P> 155 2. When dealing with multiple paths through the tree simultaneously, it is not 156 straightforward to keep track of captured substrings for the different matching 157 possibilities, and PCRE2's implementation of this algorithm does not attempt to 158 do this. This means that no captured substrings are available. 159 </P> 160 <P> 161 3. Because no substrings are captured, back references within the pattern are 162 not supported, and cause errors if encountered. 163 </P> 164 <P> 165 4. For the same reason, conditional expressions that use a backreference as the 166 condition or test for a specific group recursion are not supported. 167 </P> 168 <P> 169 5. Because many paths through the tree may be active, the \K escape sequence, 170 which resets the start of the match when encountered (but may be on some paths 171 and not on others), is not supported. It causes an error if encountered. 172 </P> 173 <P> 174 6. Callouts are supported, but the value of the <i>capture_top</i> field is 175 always 1, and the value of the <i>capture_last</i> field is always 0. 176 </P> 177 <P> 178 7. The \C escape sequence, which (in the standard algorithm) always matches a 179 single code unit, even in a UTF mode, is not supported in these modes, because 180 the alternative algorithm moves through the subject string one character (not 181 code unit) at a time, for all active paths through the tree. 182 </P> 183 <P> 184 8. Except for (*FAIL), the backtracking control verbs such as (*PRUNE) are not 185 supported. (*FAIL) is supported, and behaves like a failing negative assertion. 186 </P> 187 <br><a name="SEC5" href="#TOC1">ADVANTAGES OF THE ALTERNATIVE ALGORITHM</a><br> 188 <P> 189 Using the alternative matching algorithm provides the following advantages: 190 </P> 191 <P> 192 1. All possible matches (at a single point in the subject) are automatically 193 found, and in particular, the longest match is found. To find more than one 194 match using the standard algorithm, you have to do kludgy things with 195 callouts. 196 </P> 197 <P> 198 2. Because the alternative algorithm scans the subject string just once, and 199 never needs to backtrack (except for lookbehinds), it is possible to pass very 200 long subject strings to the matching function in several pieces, checking for 201 partial matching each time. Although it is also possible to do multi-segment 202 matching using the standard algorithm, by retaining partially matched 203 substrings, it is more complicated. The 204 <a href="pcre2partial.html"><b>pcre2partial</b></a> 205 documentation gives details of partial matching and discusses multi-segment 206 matching. 207 </P> 208 <br><a name="SEC6" href="#TOC1">DISADVANTAGES OF THE ALTERNATIVE ALGORITHM</a><br> 209 <P> 210 The alternative algorithm suffers from a number of disadvantages: 211 </P> 212 <P> 213 1. It is substantially slower than the standard algorithm. This is partly 214 because it has to search for all possible matches, but is also because it is 215 less susceptible to optimization. 216 </P> 217 <P> 218 2. Capturing parentheses and back references are not supported. 219 </P> 220 <P> 221 3. Although atomic groups are supported, their use does not provide the 222 performance advantage that it does for the standard algorithm. 223 </P> 224 <br><a name="SEC7" href="#TOC1">AUTHOR</a><br> 225 <P> 226 Philip Hazel 227 <br> 228 University Computing Service 229 <br> 230 Cambridge, England. 231 <br> 232 </P> 233 <br><a name="SEC8" href="#TOC1">REVISION</a><br> 234 <P> 235 Last updated: 29 September 2014 236 <br> 237 Copyright © 1997-2014 University of Cambridge. 238 <br> 239 <p> 240 Return to the <a href="index.html">PCRE2 index page</a>. 241 </p> 242
