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

Diff of /code/trunk/doc/html/pcrepartial.html

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 111 by ph10, Thu Mar 8 16:53:09 2007 UTC revision 954 by ph10, Sat Mar 31 18:09:26 2012 UTC
# Line 14  man page, in case the conversion went wr Line 14  man page, in case the conversion went wr
14  <br>  <br>
15  <ul>  <ul>
16  <li><a name="TOC1" href="#SEC1">PARTIAL MATCHING IN PCRE</a>  <li><a name="TOC1" href="#SEC1">PARTIAL MATCHING IN PCRE</a>
17  <li><a name="TOC2" href="#SEC2">RESTRICTED PATTERNS FOR PCRE_PARTIAL</a>  <li><a name="TOC2" href="#SEC2">PARTIAL MATCHING USING pcre_exec() OR pcre16_exec()</a>
18  <li><a name="TOC3" href="#SEC3">EXAMPLE OF PARTIAL MATCHING USING PCRETEST</a>  <li><a name="TOC3" href="#SEC3">PARTIAL MATCHING USING pcre_dfa_exec() OR pcre16_dfa_exec()</a>
19  <li><a name="TOC4" href="#SEC4">MULTI-SEGMENT MATCHING WITH pcre_dfa_exec()</a>  <li><a name="TOC4" href="#SEC4">PARTIAL MATCHING AND WORD BOUNDARIES</a>
20  <li><a name="TOC5" href="#SEC5">AUTHOR</a>  <li><a name="TOC5" href="#SEC5">FORMERLY RESTRICTED PATTERNS</a>
21  <li><a name="TOC6" href="#SEC6">REVISION</a>  <li><a name="TOC6" href="#SEC6">EXAMPLE OF PARTIAL MATCHING USING PCRETEST</a>
22    <li><a name="TOC7" href="#SEC7">MULTI-SEGMENT MATCHING WITH pcre_dfa_exec() OR pcre16_dfa_exec()</a>
23    <li><a name="TOC8" href="#SEC8">MULTI-SEGMENT MATCHING WITH pcre_exec() OR pcre16_exec()</a>
24    <li><a name="TOC9" href="#SEC9">ISSUES WITH MULTI-SEGMENT MATCHING</a>
25    <li><a name="TOC10" href="#SEC10">AUTHOR</a>
26    <li><a name="TOC11" href="#SEC11">REVISION</a>
27  </ul>  </ul>
28  <br><a name="SEC1" href="#TOC1">PARTIAL MATCHING IN PCRE</a><br>  <br><a name="SEC1" href="#TOC1">PARTIAL MATCHING IN PCRE</a><br>
29  <P>  <P>
30  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 a matching
31  <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b> matches as far as it goes, but is  function matches as far as it goes, but is too short to match the entire
32  too short to match the entire pattern, PCRE_ERROR_NOMATCH is returned. There  pattern, PCRE_ERROR_NOMATCH is returned. There are circumstances where it might
33  are circumstances where it might be helpful to distinguish this case from other  be helpful to distinguish this case from other cases in which there is no
34  cases in which there is no match.  match.
35  </P>  </P>
36  <P>  <P>
37  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
# Line 37  in the form <i>ddmmmyy</i>, defined by t Line 42  in the form <i>ddmmmyy</i>, defined by t
42  </pre>  </pre>
43  If the application sees the user's keystrokes one by one, and can check that  If the application sees the user's keystrokes one by one, and can check that
44  what has been typed so far is potentially valid, it is able to raise an error  what has been typed so far is potentially valid, it is able to raise an error
45  as soon as a mistake is made, possibly beeping and not reflecting the  as soon as a mistake is made, by beeping and not reflecting the character that
46  character that has been typed. This immediate feedback is likely to be a better  has been typed, for example. This immediate feedback is likely to be a better
47  user interface than a check that is delayed until the entire string has been  user interface than a check that is delayed until the entire string has been
48  entered.  entered. Partial matching can also be useful when the subject string is very
49    long and is not all available at once.
50  </P>  </P>
51  <P>  <P>
52  PCRE supports the concept of partial matching by means of the PCRE_PARTIAL  PCRE supports partial matching by means of the PCRE_PARTIAL_SOFT and
53  option, which can be set when calling <b>pcre_exec()</b> or  PCRE_PARTIAL_HARD options, which can be set when calling any of the matching
54  <b>pcre_dfa_exec()</b>. When this flag is set for <b>pcre_exec()</b>, the return  functions. For backwards compatibility, PCRE_PARTIAL is a synonym for
55  code PCRE_ERROR_NOMATCH is converted into PCRE_ERROR_PARTIAL if at any time  PCRE_PARTIAL_SOFT. The essential difference between the two options is whether
56  during the matching process the last part of the subject string matched part of  or not a partial match is preferred to an alternative complete match, though
57  the pattern. Unfortunately, for non-anchored matching, it is not possible to  the details differ between the two types of matching function. If both options
58  obtain the position of the start of the partial match. No captured data is set  are set, PCRE_PARTIAL_HARD takes precedence.
59  when PCRE_ERROR_PARTIAL is returned.  </P>
60    <P>
61    If you want to use partial matching with just-in-time optimized code, you must
62    call <b>pcre_study()</b> or <b>pcre16_study()</b> with one or both of these
63    options:
64    <pre>
65      PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
66      PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
67    </pre>
68    PCRE_STUDY_JIT_COMPILE should also be set if you are going to run non-partial
69    matches on the same pattern. If the appropriate JIT study mode has not been set
70    for a match, the interpretive matching code is used.
71    </P>
72    <P>
73    Setting a partial matching option disables two of PCRE's standard
74    optimizations. PCRE remembers the last literal data unit in a pattern, and
75    abandons matching immediately if it is not present in the subject string. This
76    optimization cannot be used for a subject string that might match only
77    partially. If the pattern was studied, PCRE knows the minimum length of a
78    matching string, and does not bother to run the matching function on shorter
79    strings. This optimization is also disabled for partial matching.
80    </P>
81    <br><a name="SEC2" href="#TOC1">PARTIAL MATCHING USING pcre_exec() OR pcre16_exec()</a><br>
82    <P>
83    A partial match occurs during a call to <b>pcre_exec()</b> or
84    <b>pcre16_exec()</b> when the end of the subject string is reached successfully,
85    but matching cannot continue because more characters are needed. However, at
86    least one character in the subject must have been inspected. This character
87    need not form part of the final matched string; lookbehind assertions and the
88    \K escape sequence provide ways of inspecting characters before the start of a
89    matched substring. The requirement for inspecting at least one character exists
90    because an empty string can always be matched; without such a restriction there
91    would always be a partial match of an empty string at the end of the subject.
92    </P>
93    <P>
94    If there are at least two slots in the offsets vector when a partial match is
95    returned, the first slot is set to the offset of the earliest character that
96    was inspected. For convenience, the second offset points to the end of the
97    subject so that a substring can easily be identified.
98    </P>
99    <P>
100    For the majority of patterns, the first offset identifies the start of the
101    partially matched string. However, for patterns that contain lookbehind
102    assertions, or \K, or begin with \b or \B, earlier characters have been
103    inspected while carrying out the match. For example:
104    <pre>
105      /(?&#60;=abc)123/
106    </pre>
107    This pattern matches "123", but only if it is preceded by "abc". If the subject
108    string is "xyzabc12", the offsets after a partial match are for the substring
109    "abc12", because all these characters are needed if another match is tried
110    with extra characters added to the subject.
111    </P>
112    <P>
113    What happens when a partial match is identified depends on which of the two
114    partial matching options are set.
115  </P>  </P>
116    <br><b>
117    PCRE_PARTIAL_SOFT WITH pcre_exec() OR pcre16_exec()
118    </b><br>
119  <P>  <P>
120  When PCRE_PARTIAL is set for <b>pcre_dfa_exec()</b>, the return code  If PCRE_PARTIAL_SOFT is set when <b>pcre_exec()</b> or <b>pcre16_exec()</b>
121  PCRE_ERROR_NOMATCH is converted into PCRE_ERROR_PARTIAL if the end of the  identifies a partial match, the partial match is remembered, but matching
122  subject is reached, there have been no complete matches, but there is still at  continues as normal, and other alternatives in the pattern are tried. If no
123  least one matching possibility. The portion of the string that provided the  complete match can be found, PCRE_ERROR_PARTIAL is returned instead of
124  partial match is set as the first matching string.  PCRE_ERROR_NOMATCH.
125  </P>  </P>
126  <P>  <P>
127  Using PCRE_PARTIAL disables one of PCRE's optimizations. PCRE remembers the  This option is "soft" because it prefers a complete match over a partial match.
128  last literal byte in a pattern, and abandons matching immediately if such a  All the various matching items in a pattern behave as if the subject string is
129  byte is not present in the subject string. This optimization cannot be used  potentially complete. For example, \z, \Z, and $ match at the end of the
130  for a subject string that might match only partially.  subject, as normal, and for \b and \B the end of the subject is treated as a
131    non-alphanumeric.
132  </P>  </P>
 <br><a name="SEC2" href="#TOC1">RESTRICTED PATTERNS FOR PCRE_PARTIAL</a><br>  
133  <P>  <P>
134  Because of the way certain internal optimizations are implemented in the  If there is more than one partial match, the first one that was found provides
135  <b>pcre_exec()</b> function, the PCRE_PARTIAL option cannot be used with all  the data that is returned. Consider this pattern:
 patterns. These restrictions do not apply when <b>pcre_dfa_exec()</b> is used.  
 For <b>pcre_exec()</b>, repeated single characters such as  
136  <pre>  <pre>
137    a{2,4}    /123\w+X|dogY/
138  </pre>  </pre>
139  and repeated single metasequences such as  If this is matched against the subject string "abc123dog", both
140    alternatives fail to match, but the end of the subject is reached during
141    matching, so PCRE_ERROR_PARTIAL is returned. The offsets are set to 3 and 9,
142    identifying "123dog" as the first partial match that was found. (In this
143    example, there are two partial matches, because "dog" on its own partially
144    matches the second alternative.)
145    </P>
146    <br><b>
147    PCRE_PARTIAL_HARD WITH pcre_exec() OR pcre16_exec()
148    </b><br>
149    <P>
150    If PCRE_PARTIAL_HARD is set for <b>pcre_exec()</b> or <b>pcre16_exec()</b>,
151    PCRE_ERROR_PARTIAL is returned as soon as a partial match is found, without
152    continuing to search for possible complete matches. This option is "hard"
153    because it prefers an earlier partial match over a later complete match. For
154    this reason, the assumption is made that the end of the supplied subject string
155    may not be the true end of the available data, and so, if \z, \Z, \b, \B,
156    or $ are encountered at the end of the subject, the result is
157    PCRE_ERROR_PARTIAL, provided that at least one character in the subject has
158    been inspected.
159    </P>
160    <P>
161    Setting PCRE_PARTIAL_HARD also affects the way UTF-8 and UTF-16
162    subject strings are checked for validity. Normally, an invalid sequence
163    causes the error PCRE_ERROR_BADUTF8 or PCRE_ERROR_BADUTF16. However, in the
164    special case of a truncated character at the end of the subject,
165    PCRE_ERROR_SHORTUTF8 or PCRE_ERROR_SHORTUTF16 is returned when
166    PCRE_PARTIAL_HARD is set.
167    </P>
168    <br><b>
169    Comparing hard and soft partial matching
170    </b><br>
171    <P>
172    The difference between the two partial matching options can be illustrated by a
173    pattern such as:
174    <pre>
175      /dog(sbody)?/
176    </pre>
177    This matches either "dog" or "dogsbody", greedily (that is, it prefers the
178    longer string if possible). If it is matched against the string "dog" with
179    PCRE_PARTIAL_SOFT, it yields a complete match for "dog". However, if
180    PCRE_PARTIAL_HARD is set, the result is PCRE_ERROR_PARTIAL. On the other hand,
181    if the pattern is made ungreedy the result is different:
182  <pre>  <pre>
183    \d+    /dog(sbody)??/
184  </pre>  </pre>
185  are not permitted if the maximum number of occurrences is greater than one.  In this case the result is always a complete match because that is found first,
186  Optional items such as \d? (where the maximum is one) are permitted.  and matching never continues after finding a complete match. It might be easier
187  Quantifiers with any values are permitted after parentheses, so the invalid  to follow this explanation by thinking of the two patterns like this:
 examples above can be coded thus:  
188  <pre>  <pre>
189    (a){2,4}    /dog(sbody)?/    is the same as  /dogsbody|dog/
190    (\d)+    /dog(sbody)??/   is the same as  /dog|dogsbody/
191  </pre>  </pre>
192  These constructions run more slowly, but for the kinds of application that are  The second pattern will never match "dogsbody", because it will always find the
193  envisaged for this facility, this is not felt to be a major restriction.  shorter match first.
194  </P>  </P>
195    <br><a name="SEC3" href="#TOC1">PARTIAL MATCHING USING pcre_dfa_exec() OR pcre16_dfa_exec()</a><br>
196  <P>  <P>
197  If PCRE_PARTIAL is set for a pattern that does not conform to the restrictions,  The DFA functions move along the subject string character by character, without
198  <b>pcre_exec()</b> returns the error code PCRE_ERROR_BADPARTIAL (-13).  backtracking, searching for all possible matches simultaneously. If the end of
199    the subject is reached before the end of the pattern, there is the possibility
200    of a partial match, again provided that at least one character has been
201    inspected.
202    </P>
203    <P>
204    When PCRE_PARTIAL_SOFT is set, PCRE_ERROR_PARTIAL is returned only if there
205    have been no complete matches. Otherwise, the complete matches are returned.
206    However, if PCRE_PARTIAL_HARD is set, a partial match takes precedence over any
207    complete matches. The portion of the string that was inspected when the longest
208    partial match was found is set as the first matching string, provided there are
209    at least two slots in the offsets vector.
210    </P>
211    <P>
212    Because the DFA functions always search for all possible matches, and there is
213    no difference between greedy and ungreedy repetition, their behaviour is
214    different from the standard functions when PCRE_PARTIAL_HARD is set. Consider
215    the string "dog" matched against the ungreedy pattern shown above:
216    <pre>
217      /dog(sbody)??/
218    </pre>
219    Whereas the standard functions stop as soon as they find the complete match for
220    "dog", the DFA functions also find the partial match for "dogsbody", and so
221    return that when PCRE_PARTIAL_HARD is set.
222    </P>
223    <br><a name="SEC4" href="#TOC1">PARTIAL MATCHING AND WORD BOUNDARIES</a><br>
224    <P>
225    If a pattern ends with one of sequences \b or \B, which test for word
226    boundaries, partial matching with PCRE_PARTIAL_SOFT can give counter-intuitive
227    results. Consider this pattern:
228    <pre>
229      /\bcat\b/
230    </pre>
231    This matches "cat", provided there is a word boundary at either end. If the
232    subject string is "the cat", the comparison of the final "t" with a following
233    character cannot take place, so a partial match is found. However, normal
234    matching carries on, and \b matches at the end of the subject when the last
235    character is a letter, so a complete match is found. The result, therefore, is
236    <i>not</i> PCRE_ERROR_PARTIAL. Using PCRE_PARTIAL_HARD in this case does yield
237    PCRE_ERROR_PARTIAL, because then the partial match takes precedence.
238    </P>
239    <br><a name="SEC5" href="#TOC1">FORMERLY RESTRICTED PATTERNS</a><br>
240    <P>
241    For releases of PCRE prior to 8.00, because of the way certain internal
242    optimizations were implemented in the <b>pcre_exec()</b> function, the
243    PCRE_PARTIAL option (predecessor of PCRE_PARTIAL_SOFT) could not be used with
244    all patterns. From release 8.00 onwards, the restrictions no longer apply, and
245    partial matching with can be requested for any pattern.
246    </P>
247    <P>
248    Items that were formerly restricted were repeated single characters and
249    repeated metasequences. If PCRE_PARTIAL was set for a pattern that did not
250    conform to the restrictions, <b>pcre_exec()</b> returned the error code
251    PCRE_ERROR_BADPARTIAL (-13). This error code is no longer in use. The
252    PCRE_INFO_OKPARTIAL call to <b>pcre_fullinfo()</b> to find out if a compiled
253    pattern can be used for partial matching now always returns 1.
254  </P>  </P>
255  <br><a name="SEC3" href="#TOC1">EXAMPLE OF PARTIAL MATCHING USING PCRETEST</a><br>  <br><a name="SEC6" href="#TOC1">EXAMPLE OF PARTIAL MATCHING USING PCRETEST</a><br>
256  <P>  <P>
257  If the escape sequence \P is present in a <b>pcretest</b> data line, the  If the escape sequence \P is present in a <b>pcretest</b> data line, the
258  PCRE_PARTIAL flag is used for the match. Here is a run of <b>pcretest</b> that  PCRE_PARTIAL_SOFT option is used for the match. Here is a run of <b>pcretest</b>
259  uses the date example quoted above:  that uses the date example quoted above:
260  <pre>  <pre>
261      re&#62; /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/      re&#62; /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/
262    data&#62; 25jun04\P    data&#62; 25jun04\P
263     0: 25jun04     0: 25jun04
264     1: jun     1: jun
265    data&#62; 25dec3\P    data&#62; 25dec3\P
266    Partial match    Partial match: 23dec3
267    data&#62; 3ju\P    data&#62; 3ju\P
268    Partial match    Partial match: 3ju
269    data&#62; 3juj\P    data&#62; 3juj\P
270    No match    No match
271    data&#62; j\P    data&#62; j\P
# Line 114  uses the date example quoted above: Line 273  uses the date example quoted above:
273  </pre>  </pre>
274  The first data string is matched completely, so <b>pcretest</b> shows the  The first data string is matched completely, so <b>pcretest</b> shows the
275  matched substrings. The remaining four strings do not match the complete  matched substrings. The remaining four strings do not match the complete
276  pattern, but the first two are partial matches. The same test, using  pattern, but the first two are partial matches. Similar output is obtained
277  <b>pcre_dfa_exec()</b> matching (by means of the \D escape sequence), produces  if DFA matching is used.
278  the following output:  </P>
279  <pre>  <P>
280      re&#62; /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/  If the escape sequence \P is present more than once in a <b>pcretest</b> data
281    data&#62; 25jun04\P\D  line, the PCRE_PARTIAL_HARD option is set for the match.
    0: 25jun04  
   data&#62; 23dec3\P\D  
   Partial match: 23dec3  
   data&#62; 3ju\P\D  
   Partial match: 3ju  
   data&#62; 3juj\P\D  
   No match  
   data&#62; j\P\D  
   No match  
 </pre>  
 Notice that in this case the portion of the string that was matched is made  
 available.  
282  </P>  </P>
283  <br><a name="SEC4" href="#TOC1">MULTI-SEGMENT MATCHING WITH pcre_dfa_exec()</a><br>  <br><a name="SEC7" href="#TOC1">MULTI-SEGMENT MATCHING WITH pcre_dfa_exec() OR pcre16_dfa_exec()</a><br>
284  <P>  <P>
285  When a partial match has been found using <b>pcre_dfa_exec()</b>, it is possible  When a partial match has been found using a DFA matching function, it is
286  to continue the match by providing additional subject data and calling  possible to continue the match by providing additional subject data and calling
287  <b>pcre_dfa_exec()</b> again with the same compiled regular expression, this  the function again with the same compiled regular expression, this time setting
288  time setting the PCRE_DFA_RESTART option. You must also pass the same working  the PCRE_DFA_RESTART option. You must pass the same working space as before,
289  space as before, because this is where details of the previous partial match  because this is where details of the previous partial match are stored. Here is
290  are stored. Here is an example using <b>pcretest</b>, using the \R escape  an example using <b>pcretest</b>, using the \R escape sequence to set the
291  sequence to set the PCRE_DFA_RESTART option (\P and \D are as above):  PCRE_DFA_RESTART option (\D specifies the use of the DFA matching function):
292  <pre>  <pre>
293      re&#62; /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/      re&#62; /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/
294    data&#62; 23ja\P\D    data&#62; 23ja\P\D
# Line 156  not retain the previously partially-matc Line 303  not retain the previously partially-matc
303  program to do that if it needs to.  program to do that if it needs to.
304  </P>  </P>
305  <P>  <P>
306  You can set PCRE_PARTIAL with PCRE_DFA_RESTART to continue partial matching  You can set the PCRE_PARTIAL_SOFT or PCRE_PARTIAL_HARD options with
307  over multiple segments. This facility can be used to pass very long subject  PCRE_DFA_RESTART to continue partial matching over multiple segments. This
308  strings to <b>pcre_dfa_exec()</b>. However, some care is needed for certain  facility can be used to pass very long subject strings to the DFA matching
309  types of pattern.  functions.
310  </P>  </P>
311  <P>  <br><a name="SEC8" href="#TOC1">MULTI-SEGMENT MATCHING WITH pcre_exec() OR pcre16_exec()</a><br>
312  1. If the pattern contains tests for the beginning or end of a line, you need  <P>
313  to pass the PCRE_NOTBOL or PCRE_NOTEOL options, as appropriate, when the  From release 8.00, the standard matching functions can also be used to do
314  subject string for any call does not contain the beginning or end of a line.  multi-segment matching. Unlike the DFA functions, it is not possible to
315  </P>  restart the previous match with a new segment of data. Instead, new data must
316  <P>  be added to the previous subject string, and the entire match re-run, starting
317  2. If the pattern contains backward assertions (including \b or \B), you need  from the point where the partial match occurred. Earlier data can be discarded.
318  to arrange for some overlap in the subject strings to allow for this. For  </P>
319  example, you could pass the subject in chunks that are 500 bytes long, but in  <P>
320  a buffer of 700 bytes, with the starting offset set to 200 and the previous 200  It is best to use PCRE_PARTIAL_HARD in this situation, because it does not
321  bytes at the start of the buffer.  treat the end of a segment as the end of the subject when matching \z, \Z,
322  </P>  \b, \B, and $. Consider an unanchored pattern that matches dates:
323  <P>  <pre>
324  3. Matching a subject string that is split into multiple segments does not      re&#62; /\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d/
325  always produce exactly the same result as matching over one single long string.    data&#62; The date is 23ja\P\P
326  The difference arises when there are multiple matching possibilities, because a    Partial match: 23ja
327  partial match result is given only when there are no completed matches in a  </pre>
328  call to fBpcre_dfa_exec()\fP. This means that as soon as the shortest match has  At this stage, an application could discard the text preceding "23ja", add on
329  been found, continuation to a new subject segment is no longer possible.  text from the next segment, and call the matching function again. Unlike the
330  Consider this <b>pcretest</b> example:  DFA matching functions, the entire matching string must always be available,
331    and the complete matching process occurs for each call, so more memory and more
332    processing time is needed.
333    </P>
334    <P>
335    <b>Note:</b> If the pattern contains lookbehind assertions, or \K, or starts
336    with \b or \B, the string that is returned for a partial match includes
337    characters that precede the partially matched string itself, because these must
338    be retained when adding on more characters for a subsequent matching attempt.
339    However, in some cases you may need to retain even earlier characters, as
340    discussed in the next section.
341    </P>
342    <br><a name="SEC9" href="#TOC1">ISSUES WITH MULTI-SEGMENT MATCHING</a><br>
343    <P>
344    Certain types of pattern may give problems with multi-segment matching,
345    whichever matching function is used.
346    </P>
347    <P>
348    1. If the pattern contains a test for the beginning of a line, you need to pass
349    the PCRE_NOTBOL option when the subject string for any call does start at the
350    beginning of a line. There is also a PCRE_NOTEOL option, but in practice when
351    doing multi-segment matching you should be using PCRE_PARTIAL_HARD, which
352    includes the effect of PCRE_NOTEOL.
353    </P>
354    <P>
355    2. Lookbehind assertions that have already been obeyed are catered for in the
356    offsets that are returned for a partial match. However a lookbehind assertion
357    later in the pattern could require even earlier characters to be inspected. You
358    can handle this case by using the PCRE_INFO_MAXLOOKBEHIND option of the
359    <b>pcre_fullinfo()</b> or <b>pcre16_fullinfo()</b> functions to obtain the length
360    of the largest lookbehind in the pattern. This length is given in characters,
361    not bytes. If you always retain at least that many characters before the
362    partially matched string, all should be well. (Of course, near the start of the
363    subject, fewer characters may be present; in that case all characters should be
364    retained.)
365    </P>
366    <P>
367    3. Because a partial match must always contain at least one character, what
368    might be considered a partial match of an empty string actually gives a "no
369    match" result. For example:
370    <pre>
371        re&#62; /c(?&#60;=abc)x/
372      data&#62; ab\P
373      No match
374    </pre>
375    If the next segment begins "cx", a match should be found, but this will only
376    happen if characters from the previous segment are retained. For this reason, a
377    "no match" result should be interpreted as "partial match of an empty string"
378    when the pattern contains lookbehinds.
379    </P>
380    <P>
381    4. Matching a subject string that is split into multiple segments may not
382    always produce exactly the same result as matching over one single long string,
383    especially when PCRE_PARTIAL_SOFT is used. The section "Partial Matching and
384    Word Boundaries" above describes an issue that arises if the pattern ends with
385    \b or \B. Another kind of difference may occur when there are multiple
386    matching possibilities, because (for PCRE_PARTIAL_SOFT) a partial match result
387    is given only when there are no completed matches. This means that as soon as
388    the shortest match has been found, continuation to a new subject segment is no
389    longer possible. Consider again this <b>pcretest</b> example:
390  <pre>  <pre>
391      re&#62; /dog(sbody)?/      re&#62; /dog(sbody)?/
392      data&#62; dogsb\P
393       0: dog
394    data&#62; do\P\D    data&#62; do\P\D
395    Partial match: do    Partial match: do
396    data&#62; gsb\R\P\D    data&#62; gsb\R\P\D
# Line 191  Consider this <b>pcretest</b> example: Line 399  Consider this <b>pcretest</b> example:
399     0: dogsbody     0: dogsbody
400     1: dog     1: dog
401  </pre>  </pre>
402  The pattern matches the words "dog" or "dogsbody". When the subject is  The first data line passes the string "dogsb" to a standard matching function,
403  presented in several parts ("do" and "gsb" being the first two) the match stops  setting the PCRE_PARTIAL_SOFT option. Although the string is a partial match
404  when "dog" has been found, and it is not possible to continue. On the other  for "dogsbody", the result is not PCRE_ERROR_PARTIAL, because the shorter
405  hand, if "dogsbody" is presented as a single string, both matches are found.  string "dog" is a complete match. Similarly, when the subject is presented to
406  </P>  a DFA matching function in several parts ("do" and "gsb" being the first two)
407  <P>  the match stops when "dog" has been found, and it is not possible to continue.
408  Because of this phenomenon, it does not usually make sense to end a pattern  On the other hand, if "dogsbody" is presented as a single string, a DFA
409  that is going to be matched in this way with a variable repeat.  matching function finds both matches.
410  </P>  </P>
411  <P>  <P>
412  4. Patterns that contain alternatives at the top level which do not all  Because of these problems, it is best to use PCRE_PARTIAL_HARD when matching
413  start with the same pattern item may not work as expected. For example,  multi-segment data. The example above then behaves differently:
414  consider this pattern:  <pre>
415        re&#62; /dog(sbody)?/
416      data&#62; dogsb\P\P
417      Partial match: dogsb
418      data&#62; do\P\D
419      Partial match: do
420      data&#62; gsb\R\P\P\D
421      Partial match: gsb
422    </pre>
423    5. Patterns that contain alternatives at the top level which do not all start
424    with the same pattern item may not work as expected when PCRE_DFA_RESTART is
425    used. For example, consider this pattern:
426  <pre>  <pre>
427    1234|3789    1234|3789
428  </pre>  </pre>
429  If the first part of the subject is "ABC123", a partial match of the first  If the first part of the subject is "ABC123", a partial match of the first
430  alternative is found at offset 3. There is no partial match for the second  alternative is found at offset 3. There is no partial match for the second
431  alternative, because such a match does not start at the same point in the  alternative, because such a match does not start at the same point in the
432  subject string. Attempting to continue with the string "789" does not yield a  subject string. Attempting to continue with the string "7890" does not yield a
433  match because only those alternatives that match at one point in the subject  match because only those alternatives that match at one point in the subject
434  are remembered. The problem arises because the start of the second alternative  are remembered. The problem arises because the start of the second alternative
435  matches within the first alternative. There is no problem with anchored  matches within the first alternative. There is no problem with anchored
# Line 218  patterns or patterns such as: Line 437  patterns or patterns such as:
437  <pre>  <pre>
438    1234|ABCD    1234|ABCD
439  </pre>  </pre>
440  where no string can be a partial match for both alternatives.  where no string can be a partial match for both alternatives. This is not a
441    problem if a standard matching function is used, because the entire match has
442    to be rerun each time:
443    <pre>
444        re&#62; /1234|3789/
445      data&#62; ABC123\P\P
446      Partial match: 123
447      data&#62; 1237890
448       0: 3789
449    </pre>
450    Of course, instead of using PCRE_DFA_RESTART, the same technique of re-running
451    the entire match can also be used with the DFA matching functions. Another
452    possibility is to work with two buffers. If a partial match at offset <i>n</i>
453    in the first buffer is followed by "no match" when PCRE_DFA_RESTART is used on
454    the second buffer, you can then try a new match starting at offset <i>n+1</i> in
455    the first buffer.
456  </P>  </P>
457  <br><a name="SEC5" href="#TOC1">AUTHOR</a><br>  <br><a name="SEC10" href="#TOC1">AUTHOR</a><br>
458  <P>  <P>
459  Philip Hazel  Philip Hazel
460  <br>  <br>
# Line 229  University Computing Service Line 463  University Computing Service
463  Cambridge CB2 3QH, England.  Cambridge CB2 3QH, England.
464  <br>  <br>
465  </P>  </P>
466  <br><a name="SEC6" href="#TOC1">REVISION</a><br>  <br><a name="SEC11" href="#TOC1">REVISION</a><br>
467  <P>  <P>
468  Last updated: 06 March 2007  Last updated: 24 February 2012
469  <br>  <br>
470  Copyright &copy; 1997-2007 University of Cambridge.  Copyright &copy; 1997-2012 University of Cambridge.
471  <br>  <br>
472  <p>  <p>
473  Return to the <a href="index.html">PCRE index page</a>.  Return to the <a href="index.html">PCRE index page</a>.

Legend:
Removed from v.111  
changed lines
  Added in v.954

  ViewVC Help
Powered by ViewVC 1.1.5