1PCRE2CALLOUT(3) Library Functions Manual PCRE2CALLOUT(3)
2
6 PCRE2 - Perl-compatible regular expressions (revised API)
7
9 #include <pcre2.h>
11 int (*pcre2_callout)(pcre2_callout_block *, void *);
13 int pcre2_callout_enumerate(const pcre2_code *code,
15 int (*callback)(pcre2_callout_enumerate_block *, void *),
16 void *user_data);
17
19 PCRE2 provides a feature called "callout", which is a means of tempo‐
21 rarily passing control to the caller of PCRE2 in the middle of pattern
22 matching. The caller of PCRE2 provides an external function by putting
23 its entry point in a match context (see pcre2_set_callout() in the
24 pcre2api documentation).
25 When using the pcre2_substitute() function, an additional callout fea‐
27 ture is available. This does a callout after each change to the subject
28 string and is described in the pcre2api documentation; the rest of this
29 document is concerned with callouts during pattern matching.
30 Within a regular expression, (?C<arg>) indicates a point at which the
32 external function is to be called. Different callout points can be
33 identified by putting a number less than 256 after the letter C. The
34 default value is zero. Alternatively, the argument may be a delimited
35 string. The starting delimiter must be one of ` ' " ^ % # $ { and the
36 ending delimiter is the same as the start, except for {, where the end‐
37 ing delimiter is }. If the ending delimiter is needed within the
38 string, it must be doubled. For example, this pattern has two callout
39 points:
40 (?C1)abc(?C"some ""arbitrary"" text")def
42 If the PCRE2_AUTO_CALLOUT option bit is set when a pattern is compiled,
44 PCRE2 automatically inserts callouts, all with number 255, before each
45 item in the pattern except for immediately before or after an explicit
46 callout. For example, if PCRE2_AUTO_CALLOUT is used with the pattern
47 A(?C3)B
49 it is processed as if it were
51 (?C255)A(?C3)B(?C255)
53 Here is a more complicated example:
55 A(\d{2}|--)
57 With PCRE2_AUTO_CALLOUT, this pattern is processed as if it were
59 (?C255)A(?C255)((?C255)\d{2}(?C255)|(?C255)-(?C255)-(?C255))(?C255)
61 Notice that there is a callout before and after each parenthesis and
63 alternation bar. If the pattern contains a conditional group whose con‐
64 dition is an assertion, an automatic callout is inserted immediately
65 before the condition. Such a callout may also be inserted explicitly,
66 for example:
67 (?(?C9)(?=a)ab|de) (?(?C%text%)(?!=d)ab|de)
69 This applies only to assertion conditions (because they are themselves
71 independent groups).
72 Callouts can be useful for tracking the progress of pattern matching.
74 The pcre2test program has a pattern qualifier (/auto_callout) that sets
75 automatic callouts. When any callouts are present, the output from
76 pcre2test indicates how the pattern is being matched. This is useful
77 information when you are trying to optimize the performance of a par‐
78 ticular pattern.
79
81 You should be aware that, because of optimizations in the way PCRE2
83 compiles and matches patterns, callouts sometimes do not happen exactly
84 as you might expect.
85 Auto-possessification
87 At compile time, PCRE2 "auto-possessifies" repeated items when it knows
89 that what follows cannot be part of the repeat. For example, a+[bc] is
90 compiled as if it were a++[bc]. The pcre2test output when this pattern
91 is compiled with PCRE2_ANCHORED and PCRE2_AUTO_CALLOUT and then applied
92 to the string "aaaa" is:
93 --->aaaa
95 +0 ^ a+
96 +2 ^ ^ [bc]
97 No match
98 This indicates that when matching [bc] fails, there is no backtracking
100 into a+ (because it is being treated as a++) and therefore the callouts
101 that would be taken for the backtracks do not occur. You can disable
102 the auto-possessify feature by passing PCRE2_NO_AUTO_POSSESS to
103 pcre2_compile(), or starting the pattern with (*NO_AUTO_POSSESS). In
104 this case, the output changes to this:
105 --->aaaa
107 +0 ^ a+
108 +2 ^ ^ [bc]
109 +2 ^ ^ [bc]
110 +2 ^ ^ [bc]
111 +2 ^^ [bc]
112 No match
113 This time, when matching [bc] fails, the matcher backtracks into a+ and
115 tries again, repeatedly, until a+ itself fails.
116 Automatic .* anchoring
118 By default, an optimization is applied when .* is the first significant
120 item in a pattern. If PCRE2_DOTALL is set, so that the dot can match
121 any character, the pattern is automatically anchored. If PCRE2_DOTALL
122 is not set, a match can start only after an internal newline or at the
123 beginning of the subject, and pcre2_compile() remembers this. If a pat‐
124 tern has more than one top-level branch, automatic anchoring occurs if
125 all branches are anchorable.
126 This optimization is disabled, however, if .* is in an atomic group or
128 if there is a backreference to the capture group in which it appears.
129 It is also disabled if the pattern contains (*PRUNE) or (*SKIP). How‐
130 ever, the presence of callouts does not affect it.
131 For example, if the pattern .*\d is compiled with PCRE2_AUTO_CALLOUT
133 and applied to the string "aa", the pcre2test output is:
134 --->aa
136 +0 ^ .*
137 +2 ^ ^ \d
138 +2 ^^ \d
139 +2 ^ \d
140 No match
141 This shows that all match attempts start at the beginning of the sub‐
143 ject. In other words, the pattern is anchored. You can disable this
144 optimization by passing PCRE2_NO_DOTSTAR_ANCHOR to pcre2_compile(), or
145 starting the pattern with (*NO_DOTSTAR_ANCHOR). In this case, the out‐
146 put changes to:
147 --->aa
149 +0 ^ .*
150 +2 ^ ^ \d
151 +2 ^^ \d
152 +2 ^ \d
153 +0 ^ .*
154 +2 ^^ \d
155 +2 ^ \d
156 No match
157 This shows more match attempts, starting at the second subject charac‐
159 ter. Another optimization, described in the next section, means that
160 there is no subsequent attempt to match with an empty subject.
161 Other optimizations
163 Other optimizations that provide fast "no match" results also affect
165 callouts. For example, if the pattern is
166 ab(?C4)cd
168 PCRE2 knows that any matching string must contain the letter "d". If
170 the subject string is "abyz", the lack of "d" means that matching
171 doesn't ever start, and the callout is never reached. However, with
172 "abyd", though the result is still no match, the callout is obeyed.
173 For most patterns PCRE2 also knows the minimum length of a matching
175 string, and will immediately give a "no match" return without actually
176 running a match if the subject is not long enough, or, for unanchored
177 patterns, if it has been scanned far enough.
178 You can disable these optimizations by passing the PCRE2_NO_START_OPTI‐
180 MIZE option to pcre2_compile(), or by starting the pattern with
181 (*NO_START_OPT). This slows down the matching process, but does ensure
182 that callouts such as the example above are obeyed.
183
185 During matching, when PCRE2 reaches a callout point, if an external
187 function is provided in the match context, it is called. This applies
188 to both normal, DFA, and JIT matching. The first argument to the call‐
189 out function is a pointer to a pcre2_callout block. The second argument
190 is the void * callout data that was supplied when the callout was set
191 up by calling pcre2_set_callout() (see the pcre2api documentation). The
192 callout block structure contains the following fields, not necessarily
193 in this order:
194 uint32_t version;
196 uint32_t callout_number;
197 uint32_t capture_top;
198 uint32_t capture_last;
199 uint32_t callout_flags;
200 PCRE2_SIZE *offset_vector;
201 PCRE2_SPTR mark;
202 PCRE2_SPTR subject;
203 PCRE2_SIZE subject_length;
204 PCRE2_SIZE start_match;
205 PCRE2_SIZE current_position;
206 PCRE2_SIZE pattern_position;
207 PCRE2_SIZE next_item_length;
208 PCRE2_SIZE callout_string_offset;
209 PCRE2_SIZE callout_string_length;
210 PCRE2_SPTR callout_string;
211 The version field contains the version number of the block format. The
213 current version is 2; the three callout string fields were added for
214 version 1, and the callout_flags field for version 2. If you are writ‐
215 ing an application that might use an earlier release of PCRE2, you
216 should check the version number before accessing any of these fields.
217 The version number will increase in future if more fields are added,
218 but the intention is never to remove any of the existing fields.
219 Fields for numerical callouts
221 For a numerical callout, callout_string is NULL, and callout_number
223 contains the number of the callout, in the range 0-255. This is the
224 number that follows (?C for callouts that part of the pattern; it is
225 255 for automatically generated callouts.
226 Fields for string callouts
228 For callouts with string arguments, callout_number is always zero, and
230 callout_string points to the string that is contained within the com‐
231 piled pattern. Its length is given by callout_string_length. Duplicated
232 ending delimiters that were present in the original pattern string have
233 been turned into single characters, but there is no other processing of
234 the callout string argument. An additional code unit containing binary
235 zero is present after the string, but is not included in the length.
236 The delimiter that was used to start the string is also stored within
237 the pattern, immediately before the string itself. You can access this
238 delimiter as callout_string[-1] if you need it.
239 The callout_string_offset field is the code unit offset to the start of
241 the callout argument string within the original pattern string. This is
242 provided for the benefit of applications such as script languages that
243 might need to report errors in the callout string within the pattern.
244 Fields for all callouts
246 The remaining fields in the callout block are the same for both kinds
248 of callout.
249 The offset_vector field is a pointer to a vector of capturing offsets
251 (the "ovector"). You may read the elements in this vector, but you must
252 not change any of them.
253 For calls to pcre2_match(), the offset_vector field is not (since
255 release 10.30) a pointer to the actual ovector that was passed to the
256 matching function in the match data block. Instead it points to an
257 internal ovector of a size large enough to hold all possible captured
258 substrings in the pattern. Note that whenever a recursion or subroutine
259 call within a pattern completes, the capturing state is reset to what
260 it was before.
261 The capture_last field contains the number of the most recently cap‐
263 tured substring, and the capture_top field contains one more than the
264 number of the highest numbered captured substring so far. If no sub‐
265 strings have yet been captured, the value of capture_last is 0 and the
266 value of capture_top is 1. The values of these fields do not always
267 differ by one; for example, when the callout in the pattern
268 ((a)(b))(?C2) is taken, capture_last is 1 but capture_top is 4.
269 The contents of ovector[2] to ovector[<capture_top>*2-1] can be
271 inspected in order to extract substrings that have been matched so far,
272 in the same way as extracting substrings after a match has completed.
273 The values in ovector[0] and ovector[1] are always PCRE2_UNSET because
274 the match is by definition not complete. Substrings that have not been
275 captured but whose numbers are less than capture_top also have both of
276 their ovector slots set to PCRE2_UNSET.
277 For DFA matching, the offset_vector field points to the ovector that
279 was passed to the matching function in the match data block for call‐
280 outs at the top level, but to an internal ovector during the processing
281 of pattern recursions, lookarounds, and atomic groups. However, these
282 ovectors hold no useful information because pcre2_dfa_match() does not
283 support substring capturing. The value of capture_top is always 1 and
284 the value of capture_last is always 0 for DFA matching.
285 The subject and subject_length fields contain copies of the values that
287 were passed to the matching function.
288 The start_match field normally contains the offset within the subject
290 at which the current match attempt started. However, if the escape
291 sequence \K has been encountered, this value is changed to reflect the
292 modified starting point. If the pattern is not anchored, the callout
293 function may be called several times from the same point in the pattern
294 for different starting points in the subject.
295 The current_position field contains the offset within the subject of
297 the current match pointer.
298 The pattern_position field contains the offset in the pattern string to
300 the next item to be matched.
301 The next_item_length field contains the length of the next item to be
303 processed in the pattern string. When the callout is at the end of the
304 pattern, the length is zero. When the callout precedes an opening
305 parenthesis, the length includes meta characters that follow the paren‐
306 thesis. For example, in a callout before an assertion such as (?=ab)
307 the length is 3. For an an alternation bar or a closing parenthesis,
308 the length is one, unless a closing parenthesis is followed by a quan‐
309 tifier, in which case its length is included. (This changed in release
310 10.23. In earlier releases, before an opening parenthesis the length
311 was that of the entire group, and before an alternation bar or a clos‐
312 ing parenthesis the length was zero.)
313 The pattern_position and next_item_length fields are intended to help
315 in distinguishing between different automatic callouts, which all have
316 the same callout number. However, they are set for all callouts, and
317 are used by pcre2test to show the next item to be matched when display‐
318 ing callout information.
319 In callouts from pcre2_match() the mark field contains a pointer to the
321 zero-terminated name of the most recently passed (*MARK), (*PRUNE), or
322 (*THEN) item in the match, or NULL if no such items have been passed.
323 Instances of (*PRUNE) or (*THEN) without a name do not obliterate a
324 previous (*MARK). In callouts from the DFA matching function this field
325 always contains NULL.
326 The callout_flags field is always zero in callouts from
328 pcre2_dfa_match() or when JIT is being used. When pcre2_match() without
329 JIT is used, the following bits may be set:
330 PCRE2_CALLOUT_STARTMATCH
332 This is set for the first callout after the start of matching for each
334 new starting position in the subject.
335 PCRE2_CALLOUT_BACKTRACK
337 This is set if there has been a matching backtrack since the previous
339 callout, or since the start of matching if this is the first callout
340 from a pcre2_match() run.
341 Both bits are set when a backtrack has caused a "bumpalong" to a new
343 starting position in the subject. Output from pcre2test does not indi‐
344 cate the presence of these bits unless the callout_extra modifier is
345 set.
346 The information in the callout_flags field is provided so that applica‐
348 tions can track and tell their users how matching with backtracking is
349 done. This can be useful when trying to optimize patterns, or just to
350 understand how PCRE2 works. There is no support in pcre2_dfa_match()
351 because there is no backtracking in DFA matching, and there is no sup‐
352 port in JIT because JIT is all about maximimizing matching performance.
353 In both these cases the callout_flags field is always zero.
354
356 The external callout function returns an integer to PCRE2. If the value
358 is zero, matching proceeds as normal. If the value is greater than
359 zero, matching fails at the current point, but the testing of other
360 matching possibilities goes ahead, just as if a lookahead assertion had
361 failed. If the value is less than zero, the match is abandoned, and the
362 matching function returns the negative value.
363 Negative values should normally be chosen from the set of
365 PCRE2_ERROR_xxx values. In particular, PCRE2_ERROR_NOMATCH forces a
366 standard "no match" failure. The error number PCRE2_ERROR_CALLOUT is
367 reserved for use by callout functions; it will never be used by PCRE2
368 itself.
369
371 int pcre2_callout_enumerate(const pcre2_code *code,
373 int (*callback)(pcre2_callout_enumerate_block *, void *),
374 void *user_data);
375 A script language that supports the use of string arguments in callouts
377 might like to scan all the callouts in a pattern before running the
378 match. This can be done by calling pcre2_callout_enumerate(). The first
379 argument is a pointer to a compiled pattern, the second points to a
380 callback function, and the third is arbitrary user data. The callback
381 function is called for every callout in the pattern in the order in
382 which they appear. Its first argument is a pointer to a callout enumer‐
383 ation block, and its second argument is the user_data value that was
384 passed to pcre2_callout_enumerate(). The data block contains the fol‐
385 lowing fields:
386 version Block version number
388 pattern_position Offset to next item in pattern
389 next_item_length Length of next item in pattern
390 callout_number Number for numbered callouts
391 callout_string_offset Offset to string within pattern
392 callout_string_length Length of callout string
393 callout_string Points to callout string or is NULL
394 The version number is currently 0. It will increase if new fields are
396 ever added to the block. The remaining fields are the same as their
397 namesakes in the pcre2_callout block that is used for callouts during
398 matching, as described above.
399 Note that the value of pattern_position is unique for each callout.
401 However, if a callout occurs inside a group that is quantified with a
402 non-zero minimum or a fixed maximum, the group is replicated inside the
403 compiled pattern. For example, a pattern such as /(a){2}/ is compiled
404 as if it were /(a)(a)/. This means that the callout will be enumerated
405 more than once, but with the same value for pattern_position in each
406 case.
407 The callback function should normally return zero. If it returns a non-
409 zero value, scanning the pattern stops, and that value is returned from
410 pcre2_callout_enumerate().
411
413 Philip Hazel
415 University Computing Service
416 Cambridge, England.
417
419 Last updated: 03 February 2019
421 Copyright (c) 1997-2019 University of Cambridge.
422PCRE2 10.33 03 February 2019 PCRE2CALLOUT(3)