1 |
.TH PCRE 3 |
.TH PCREPARTIAL 3 |
2 |
.SH NAME |
.SH NAME |
3 |
PCRE - Perl-compatible regular expressions |
PCRE - Perl-compatible regular expressions |
4 |
.SH "PARTIAL MATCHING IN PCRE" |
.SH "PARTIAL MATCHING IN PCRE" |
5 |
.rs |
.rs |
6 |
.sp |
.sp |
7 |
In normal use of PCRE, if the subject string that is passed to |
In normal use of PCRE, if the subject string that is passed to |
8 |
\fBpcre_exec()\fP matches as far as it goes, but is too short to match the |
\fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP matches as far as it goes, but is |
9 |
entire pattern, PCRE_ERROR_NOMATCH is returned. There are circumstances where |
too short to match the entire pattern, PCRE_ERROR_NOMATCH is returned. There |
10 |
it might be helpful to distinguish this case from other cases in which there is |
are circumstances where it might be helpful to distinguish this case from other |
11 |
no match. |
cases in which there is no match. |
12 |
.P |
.P |
13 |
Consider, for example, an application where a human is required to type in data |
Consider, for example, an application where a human is required to type in data |
14 |
for a field with specific formatting requirements. An example might be a date |
for a field with specific formatting requirements. An example might be a date |
24 |
entered. |
entered. |
25 |
.P |
.P |
26 |
PCRE supports the concept of partial matching by means of the PCRE_PARTIAL |
PCRE supports the concept of partial matching by means of the PCRE_PARTIAL |
27 |
option, which can be set when calling \fBpcre_exec()\fP. When this is done, the |
option, which can be set when calling \fBpcre_exec()\fP or |
28 |
return code PCRE_ERROR_NOMATCH is converted into PCRE_ERROR_PARTIAL if at any |
\fBpcre_dfa_exec()\fP. When this flag is set for \fBpcre_exec()\fP, the return |
29 |
time during the matching process the entire subject string matched part of the |
code PCRE_ERROR_NOMATCH is converted into PCRE_ERROR_PARTIAL if at any time |
30 |
pattern. No captured data is set when this occurs. |
during the matching process the last part of the subject string matched part of |
31 |
|
the pattern. Unfortunately, for non-anchored matching, it is not possible to |
32 |
|
obtain the position of the start of the partial match. No captured data is set |
33 |
|
when PCRE_ERROR_PARTIAL is returned. |
34 |
|
.P |
35 |
|
When PCRE_PARTIAL is set for \fBpcre_dfa_exec()\fP, the return code |
36 |
|
PCRE_ERROR_NOMATCH is converted into PCRE_ERROR_PARTIAL if the end of the |
37 |
|
subject is reached, there have been no complete matches, but there is still at |
38 |
|
least one matching possibility. The portion of the string that provided the |
39 |
|
partial match is set as the first matching string. |
40 |
.P |
.P |
41 |
Using PCRE_PARTIAL disables one of PCRE's optimizations. PCRE remembers the |
Using PCRE_PARTIAL disables one of PCRE's optimizations. PCRE remembers the |
42 |
last literal byte in a pattern, and abandons matching immediately if such a |
last literal byte in a pattern, and abandons matching immediately if such a |
47 |
.SH "RESTRICTED PATTERNS FOR PCRE_PARTIAL" |
.SH "RESTRICTED PATTERNS FOR PCRE_PARTIAL" |
48 |
.rs |
.rs |
49 |
.sp |
.sp |
50 |
Because of the way certain internal optimizations are implemented in PCRE, the |
Because of the way certain internal optimizations are implemented in the |
51 |
PCRE_PARTIAL option cannot be used with all patterns. Repeated single |
\fBpcre_exec()\fP function, the PCRE_PARTIAL option cannot be used with all |
52 |
characters such as |
patterns. These restrictions do not apply when \fBpcre_dfa_exec()\fP is used. |
53 |
|
For \fBpcre_exec()\fP, repeated single characters such as |
54 |
.sp |
.sp |
55 |
a{2,4} |
a{2,4} |
56 |
.sp |
.sp |
81 |
uses the date example quoted above: |
uses the date example quoted above: |
82 |
.sp |
.sp |
83 |
re> /^\ed?\ed(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\ed\ed$/ |
re> /^\ed?\ed(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\ed\ed$/ |
84 |
data> 25jun04\P |
data> 25jun04\eP |
85 |
0: 25jun04 |
0: 25jun04 |
86 |
1: jun |
1: jun |
87 |
data> 25dec3\P |
data> 25dec3\eP |
88 |
Partial match |
Partial match |
89 |
data> 3ju\P |
data> 3ju\eP |
90 |
Partial match |
Partial match |
91 |
data> 3juj\P |
data> 3juj\eP |
92 |
No match |
No match |
93 |
data> j\P |
data> j\eP |
94 |
No match |
No match |
95 |
.sp |
.sp |
96 |
The first data string is matched completely, so \fBpcretest\fP shows the |
The first data string is matched completely, so \fBpcretest\fP shows the |
97 |
matched substrings. The remaining four strings do not match the complete |
matched substrings. The remaining four strings do not match the complete |
98 |
pattern, but the first two are partial matches. |
pattern, but the first two are partial matches. The same test, using |
99 |
|
\fBpcre_dfa_exec()\fP matching (by means of the \eD escape sequence), produces |
100 |
|
the following output: |
101 |
|
.sp |
102 |
|
re> /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/ |
103 |
|
data> 25jun04\eP\eD |
104 |
|
0: 25jun04 |
105 |
|
data> 23dec3\eP\eD |
106 |
|
Partial match: 23dec3 |
107 |
|
data> 3ju\eP\eD |
108 |
|
Partial match: 3ju |
109 |
|
data> 3juj\eP\eD |
110 |
|
No match |
111 |
|
data> j\eP\eD |
112 |
|
No match |
113 |
|
.sp |
114 |
|
Notice that in this case the portion of the string that was matched is made |
115 |
|
available. |
116 |
|
. |
117 |
|
. |
118 |
|
.SH "MULTI-SEGMENT MATCHING WITH pcre_dfa_exec()" |
119 |
|
.rs |
120 |
|
.sp |
121 |
|
When a partial match has been found using \fBpcre_dfa_exec()\fP, it is possible |
122 |
|
to continue the match by providing additional subject data and calling |
123 |
|
\fBpcre_dfa_exec()\fP again with the same compiled regular expression, this |
124 |
|
time setting the PCRE_DFA_RESTART option. You must also pass the same working |
125 |
|
space as before, because this is where details of the previous partial match |
126 |
|
are stored. Here is an example using \fBpcretest\fP, using the \eR escape |
127 |
|
sequence to set the PCRE_DFA_RESTART option (\eP and \eD are as above): |
128 |
|
.sp |
129 |
|
re> /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/ |
130 |
|
data> 23ja\eP\eD |
131 |
|
Partial match: 23ja |
132 |
|
data> n05\eR\eD |
133 |
|
0: n05 |
134 |
|
.sp |
135 |
|
The first call has "23ja" as the subject, and requests partial matching; the |
136 |
|
second call has "n05" as the subject for the continued (restarted) match. |
137 |
|
Notice that when the match is complete, only the last part is shown; PCRE does |
138 |
|
not retain the previously partially-matched string. It is up to the calling |
139 |
|
program to do that if it needs to. |
140 |
|
.P |
141 |
|
You can set PCRE_PARTIAL with PCRE_DFA_RESTART to continue partial matching |
142 |
|
over multiple segments. This facility can be used to pass very long subject |
143 |
|
strings to \fBpcre_dfa_exec()\fP. However, some care is needed for certain |
144 |
|
types of pattern. |
145 |
|
.P |
146 |
|
1. If the pattern contains tests for the beginning or end of a line, you need |
147 |
|
to pass the PCRE_NOTBOL or PCRE_NOTEOL options, as appropriate, when the |
148 |
|
subject string for any call does not contain the beginning or end of a line. |
149 |
|
.P |
150 |
|
2. If the pattern contains backward assertions (including \eb or \eB), you need |
151 |
|
to arrange for some overlap in the subject strings to allow for this. For |
152 |
|
example, you could pass the subject in chunks that are 500 bytes long, but in |
153 |
|
a buffer of 700 bytes, with the starting offset set to 200 and the previous 200 |
154 |
|
bytes at the start of the buffer. |
155 |
|
.P |
156 |
|
3. Matching a subject string that is split into multiple segments does not |
157 |
|
always produce exactly the same result as matching over one single long string. |
158 |
|
The difference arises when there are multiple matching possibilities, because a |
159 |
|
partial match result is given only when there are no completed matches in a |
160 |
|
call to fBpcre_dfa_exec()\fP. This means that as soon as the shortest match has |
161 |
|
been found, continuation to a new subject segment is no longer possible. |
162 |
|
Consider this \fBpcretest\fP example: |
163 |
|
.sp |
164 |
|
re> /dog(sbody)?/ |
165 |
|
data> do\eP\eD |
166 |
|
Partial match: do |
167 |
|
data> gsb\eR\eP\eD |
168 |
|
0: g |
169 |
|
data> dogsbody\eD |
170 |
|
0: dogsbody |
171 |
|
1: dog |
172 |
|
.sp |
173 |
|
The pattern matches the words "dog" or "dogsbody". When the subject is |
174 |
|
presented in several parts ("do" and "gsb" being the first two) the match stops |
175 |
|
when "dog" has been found, and it is not possible to continue. On the other |
176 |
|
hand, if "dogsbody" is presented as a single string, both matches are found. |
177 |
|
.P |
178 |
|
Because of this phenomenon, it does not usually make sense to end a pattern |
179 |
|
that is going to be matched in this way with a variable repeat. |
180 |
|
.P |
181 |
|
4. Patterns that contain alternatives at the top level which do not all |
182 |
|
start with the same pattern item may not work as expected. For example, |
183 |
|
consider this pattern: |
184 |
|
.sp |
185 |
|
1234|3789 |
186 |
|
.sp |
187 |
|
If the first part of the subject is "ABC123", a partial match of the first |
188 |
|
alternative is found at offset 3. There is no partial match for the second |
189 |
|
alternative, because such a match does not start at the same point in the |
190 |
|
subject string. Attempting to continue with the string "789" does not yield a |
191 |
|
match because only those alternatives that match at one point in the subject |
192 |
|
are remembered. The problem arises because the start of the second alternative |
193 |
|
matches within the first alternative. There is no problem with anchored |
194 |
|
patterns or patterns such as: |
195 |
|
.sp |
196 |
|
1234|ABCD |
197 |
|
.sp |
198 |
|
where no string can be a partial match for both alternatives. |
199 |
. |
. |
200 |
. |
. |
201 |
.P |
.P |
202 |
.in 0 |
.in 0 |
203 |
Last updated: 08 September 2004 |
Last updated: 30 November 2006 |
204 |
.br |
.br |
205 |
Copyright (c) 1997-2004 University of Cambridge. |
Copyright (c) 1997-2006 University of Cambridge. |