/[pcre]/code/trunk/ChangeLog
ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 967 - (show annotations)
Mon May 14 07:36:39 2012 UTC (3 years, 2 months ago) by zherczeg
File size: 230861 byte(s)
Error occurred while calculating annotation data.
cache flush for Apple devices
1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.31
5 -----------------------------
6
7 1. Fixing a wrong JIT test case and some compiler warnings.
8
9 2. Removed a bashism from the RunTest script.
10
11 3. Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
12 to unsigned type, result still unsigned" that was given by an MS compiler
13 on encountering the code "-sizeof(xxx)".
14
15 4. Partial matching support is added to the JIT compiler.
16
17 5. Fixed several bugs concerned with partial matching of items that consist
18 of more than one character:
19
20 (a) /^(..)\1/ did not partially match "aba" because checking references was
21 done on an "all or nothing" basis. This also applied to repeated
22 references.
23
24 (b) \R did not give a hard partial match if \r was found at the end of the
25 subject.
26
27 (c) \X did not give a hard partial match after matching one or more
28 characters at the end of the subject.
29
30 (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
31 a partial match for the string "\r".
32
33 (e) When newline was set to CRLF, the metacharacter "." did not recognize
34 a partial match for a CR character at the end of the subject string.
35
36 6. If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
37 running pcretest, the text "(JIT)" added to the output whenever JIT is
38 actually used to run the match.
39
40 7. Individual JIT compile options can be set in pcretest by following -s+[+]
41 or /S+[+] with a digit between 1 and 7.
42
43 8. OP_NOT now supports any UTF character not just single-byte ones.
44
45 9. (*MARK) control verb is now supported by the JIT compiler.
46
47 10. The command "./RunTest list" lists the available tests without actually
48 running any of them. (Because I keep forgetting what they all are.)
49
50 11. Add PCRE_INFO_MAXLOOKBEHIND.
51
52 12. Applied a (slightly modified) user-supplied patch that improves performance
53 when the heap is used for recursion (compiled with --disable-stack-for-
54 recursion). Instead of malloc and free for each heap frame each time a
55 logical recursion happens, frames are retained on a chain and re-used where
56 possible. This sometimes gives as much as 30% improvement.
57
58 13. As documented, (*COMMIT) is now confined to within a recursive subpattern
59 call.
60
61 14. As documented, (*COMMIT) is now confined to within a positive assertion.
62
63 15. It is now possible to link pcretest with libedit as an alternative to
64 libreadline.
65
66 16. (*COMMIT) control verb is now supported by the JIT compiler.
67
68 17. The Unicode data tables have been updated to Unicode 6.1.0.
69
70 18. Added --file-list option to pcregrep.
71
72 19. Added binary file support to pcregrep, including the -a, --binary-files,
73 -I, and --text options.
74
75 20. The madvise function is renamed for posix_madvise for QNX compatibility
76 reasons. Fixed by Giuseppe D'Angelo.
77
78 21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
79
80 22. Changed the option for creating symbolic links for 16-bit man pages from
81 -s to -sf so that re-installing does not cause issues.
82
83 23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
84
85 24. Fixed a very old bug in pcretest that caused errors with restarted DFA
86 matches in certain environments (the workspace was not being correctly
87 retained). Also added to pcre_dfa_exec() a simple plausibility check on
88 some of the workspace data at the beginning of a restart.
89
90 25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
91 was not doing so when it should - probably a typo introduced by SVN 528
92 (change 8.10/14).
93
94 26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
95 \w+ when the character tables indicated that \x{c4} was a word character.
96 There were several related cases, all because the tests for doing a table
97 lookup were testing for characters less than 127 instead of 255.
98
99 27. If a pattern contains capturing parentheses that are not used in a match,
100 their slots in the ovector are set to -1. For those that are higher than
101 any matched groups, this happens at the end of processing. In the case when
102 there were back references that the ovector was too small to contain
103 (causing temporary malloc'd memory to be used during matching), and the
104 highest capturing number was not used, memory off the end of the ovector
105 was incorrectly being set to -1. (It was using the size of the temporary
106 memory instead of the true size.)
107
108 28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
109 ovector size, it uses memory at the end of the block that it has got.
110
111 29. Check for an overlong MARK name and give an error at compile time. The
112 limit is 255 for the 8-bit library and 65535 for the 16-bit library.
113
114 30. JIT compiler update.
115
116 31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
117 Rill for the patch.
118
119
120 Version 8.30 04-February-2012
121 -----------------------------
122
123 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
124 name is defined in ctype.h.
125
126 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
127 only in quite long subpatterns.
128
129 3. Removed the function pcre_info(), which has been obsolete and deprecated
130 since it was replaced by pcre_fullinfo() in February 2000.
131
132 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
133 match a (*MARK), and the match failed at the start of the subject, a
134 reference to memory before the start of the subject could occur. This bug
135 was introduced by fix 17 of release 8.21.
136
137 5. A reference to an unset group with zero minimum repetition was giving
138 totally wrong answers (in non-JavaScript-compatibility mode). For example,
139 /(another)?(\1?)test/ matched against "hello world test". This bug was
140 introduced in release 8.13.
141
142 6. Add support for 16-bit character strings (a large amount of work involving
143 many changes and refactorings).
144
145 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
146 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
147 from a file.
148
149 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
150 rounding is not applied in this particular case).
151
152 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
153 if they appear, or are escaped, in patterns.
154
155 10. Get rid of a number of -Wunused-but-set-variable warnings.
156
157 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
158 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
159 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
160 also returns the mark "x". This bug applied to capturing parentheses,
161 non-capturing parentheses, and atomic parentheses. It also applied to some
162 assertions.
163
164 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
165 information out of configure.ac instead of relying on pcre.h.generic, which
166 is not stored in the repository.
167
168 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
169 -lreadline.
170
171 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
172
173 15. Applied Graycode's patch to put the top-level frame on the stack rather
174 than the heap when not using the stack for recursion. This gives a
175 performance improvement in many cases when recursion is not deep.
176
177 16. Experimental code added to "pcretest -C" to output the stack frame size.
178
179
180 Version 8.21 12-Dec-2011
181 ------------------------
182
183 1. Updating the JIT compiler.
184
185 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
186 are added as well.
187
188 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
189 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
190 calling _pcre_jit_exec. Some extra comments are added.
191
192 4. (*MARK) settings inside atomic groups that do not contain any capturing
193 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
194 was introduced by change 18 for 8.20.
195
196 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
197 ECMA-262 standard.
198
199 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
200 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
201 This bug was probably introduced by change 9 of 8.13.
202
203 7. While fixing 6 above, I noticed that a number of other items were being
204 incorrectly rejected as "not fixed length". This arose partly because newer
205 opcodes had not been added to the fixed-length checking code. I have (a)
206 corrected the bug and added tests for these items, and (b) arranged for an
207 error to occur if an unknown opcode is encountered while checking for fixed
208 length instead of just assuming "not fixed length". The items that were
209 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
210 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
211 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
212
213 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
214 being incorrectly compiled and would have given unpredicatble results.
215
216 9. A possessively repeated subpattern with minimum repeat count greater than
217 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
218 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
219 the first (A) could occur when it should not.
220
221 10. Add a cast and remove a redundant test from the code.
222
223 11. JIT should use pcre_malloc/pcre_free for allocation.
224
225 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
226 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
227 is anything other than /usr, -L is still shown).
228
229 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
230
231 14. Perl does not support \N without a following name in a [] class; PCRE now
232 also gives an error.
233
234 15. If a forward reference was repeated with an upper limit of around 2000,
235 it caused the error "internal error: overran compiling workspace". The
236 maximum number of forward references (including repeats) was limited by the
237 internal workspace, and dependent on the LINK_SIZE. The code has been
238 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
239 the default depends on LINK_SIZE. There is a new upper limit (for safety)
240 of around 200,000 forward references. While doing this, I also speeded up
241 the filling in of repeated forward references.
242
243 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
244 incorrectly expecting the subject to contain another "a" after the start.
245
246 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
247 in the match, the SKIP should be ignored. This was not happening; instead
248 the SKIP was being treated as NOMATCH. For patterns such as
249 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
250 tested.
251
252 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
253 now much more compatible with Perl, in particular in cases where the result
254 is a non-match for a non-anchored pattern. For example, if
255 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
256 "m", where previously it did not return a name. A side effect of this
257 change is that for partial matches, the last encountered mark name is
258 returned, as for non matches. A number of tests that were previously not
259 Perl-compatible have been moved into the Perl-compatible test files. The
260 refactoring has had the pleasing side effect of removing one argument from
261 the match() function, thus reducing its stack requirements.
262
263 19. If the /S+ option was used in pcretest to study a pattern using JIT,
264 subsequent uses of /S (without +) incorrectly behaved like /S+.
265
266 21. Retrieve executable code size support for the JIT compiler and fixing
267 some warnings.
268
269 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
270 not work when the shorter character appeared right at the end of the
271 subject string.
272
273 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
274 systems.
275
276 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
277 output it when the /M option is used in pcretest.
278
279 25. The CheckMan script was not being included in the distribution. Also, added
280 an explicit "perl" to run Perl scripts from the PrepareRelease script
281 because this is reportedly needed in Windows.
282
283 26. If study data was being save in a file and studying had not found a set of
284 "starts with" bytes for the pattern, the data written to the file (though
285 never used) was taken from uninitialized memory and so caused valgrind to
286 complain.
287
288 27. Updated RunTest.bat as provided by Sheri Pierce.
289
290 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
291
292 29. Computation of memory usage for the table of capturing group names was
293 giving an unnecessarily large value.
294
295
296 Version 8.20 21-Oct-2011
297 ------------------------
298
299 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
300 a POSIX class. After further experiments with Perl, which convinced me that
301 Perl has bugs and confusions, a closing square bracket is no longer allowed
302 in a POSIX name. This bug also affected patterns with classes that started
303 with full stops.
304
305 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
306 captured substring, but while checking the failing first alternative,
307 substring 1 is temporarily captured. If the output vector supplied to
308 pcre_exec() was not big enough for this capture, the yield of the function
309 was still zero ("insufficient space for captured substrings"). This cannot
310 be totally fixed without adding another stack variable, which seems a lot
311 of expense for a edge case. However, I have improved the situation in cases
312 such as /(a)(b)x|abc/ matched against "abc", where the return code
313 indicates that fewer than the maximum number of slots in the ovector have
314 been set.
315
316 3. Related to (2) above: when there are more back references in a pattern than
317 slots in the output vector, pcre_exec() uses temporary memory during
318 matching, and copies in the captures as far as possible afterwards. It was
319 using the entire output vector, but this conflicts with the specification
320 that only 2/3 is used for passing back captured substrings. Now it uses
321 only the first 2/3, for compatibility. This is, of course, another edge
322 case.
323
324 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
325 main code base, and can be used by building with --enable-jit. When this is
326 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
327 runtime --no-jit option is given.
328
329 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
330 ovector, the return from the function was zero, implying that there were
331 other matches that did not fit. The correct "exactly full" value is now
332 returned.
333
334 6. If a subpattern that was called recursively or as a subroutine contained
335 (*PRUNE) or any other control that caused it to give a non-standard return,
336 invalid errors such as "Error -26 (nested recursion at the same subject
337 position)" or even infinite loops could occur.
338
339 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
340 computing the minimum length on reaching *ACCEPT, and so ended up with the
341 wrong value of 1 rather than 0. Further investigation indicates that
342 computing a minimum subject length in the presence of *ACCEPT is difficult
343 (think back references, subroutine calls), and so I have changed the code
344 so that no minimum is registered for a pattern that contains *ACCEPT.
345
346 8. If (*THEN) was present in the first (true) branch of a conditional group,
347 it was not handled as intended. [But see 16 below.]
348
349 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
350 Sheri Pierce.
351
352 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
353 the first byte in a match must be "a".
354
355 11. Change 17 for 8.13 increased the recursion depth for patterns like
356 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
357 pattern contains any instances of (*THEN). If it does not, the old
358 optimizations are restored. It would be nice to do this on a per-group
359 basis, but at the moment that is not feasible.
360
361 12. In some environments, the output of pcretest -C is CRLF terminated. This
362 broke RunTest's code that checks for the link size. A single white space
363 character after the value is now allowed for.
364
365 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
366 For "fr", it uses the Windows-specific input and output files.
367
368 14. If (*THEN) appeared in a group that was called recursively or as a
369 subroutine, it did not work as intended. [But see next item.]
370
371 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
372 pattern fragments (but not containing any | characters). If A and B are
373 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
374 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
375 D. In other words, Perl considers parentheses that do not contain any |
376 characters to be part of a surrounding alternative, whereas PCRE was
377 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
378 differently. PCRE now behaves in the same way as Perl, except in the case
379 of subroutine/recursion calls such as (?1) which have in any case always
380 been different (but PCRE had them first :-).
381
382 16. Related to 15 above: Perl does not treat the | in a conditional group as
383 creating alternatives. Such a group is treated in the same way as an
384 ordinary group without any | characters when processing (*THEN). PCRE has
385 been changed to match Perl's behaviour.
386
387 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
388 RunGrepTest script failed.
389
390 18. Change 22 for version 13 caused atomic groups to use more stack. This is
391 inevitable for groups that contain captures, but it can lead to a lot of
392 stack use in large patterns. The old behaviour has been restored for atomic
393 groups that do not contain any capturing parentheses.
394
395 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
396 suppress the check for a minimum subject length at run time. (If it was
397 given to pcre_exec() or pcre_dfa_exec() it did work.)
398
399 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
400 fail to work when decoding hex characters in data strings in EBCDIC
401 environments.
402
403 21. It appears that in at least one Mac OS environment, the isxdigit() function
404 is implemented as a macro that evaluates to its argument more than once,
405 contravening the C 90 Standard (I haven't checked a later standard). There
406 was an instance in pcretest which caused it to go wrong when processing
407 \x{...} escapes in subject strings. The has been rewritten to avoid using
408 things like p++ in the argument of isxdigit().
409
410
411 Version 8.13 16-Aug-2011
412 ------------------------
413
414 1. The Unicode data tables have been updated to Unicode 6.0.0.
415
416 2. Two minor typos in pcre_internal.h have been fixed.
417
418 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
419 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
420 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
421
422 4. There were a number of related bugs in the code for matching backrefences
423 caselessly in UTF-8 mode when codes for the characters concerned were
424 different numbers of bytes. For example, U+023A and U+2C65 are an upper
425 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
426 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
427 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
428 2-byte code at the end of the subject (it thought there wasn't enough data
429 left).
430
431 5. Comprehensive information about what went wrong is now returned by
432 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
433 as the output vector has at least 2 elements. The offset of the start of
434 the failing character and a reason code are placed in the vector.
435
436 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
437 now returned is for the first byte of the failing character, instead of the
438 last byte inspected. This is an incompatible change, but I hope it is small
439 enough not to be a problem. It makes the returned offset consistent with
440 pcre_exec() and pcre_dfa_exec().
441
442 7. pcretest now gives a text phrase as well as the error number when
443 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
444 failure, the offset and reason code are output.
445
446 8. When \R was used with a maximizing quantifier it failed to skip backwards
447 over a \r\n pair if the subsequent match failed. Instead, it just skipped
448 back over a single character (\n). This seems wrong (because it treated the
449 two characters as a single entity when going forwards), conflicts with the
450 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
451 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
452 has been changed.
453
454 9. Some internal refactoring has changed the processing so that the handling
455 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
456 time (the PCRE_DOTALL option was changed this way some time ago: version
457 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
458 which was always a bit of a fudge. It also means that there is one less
459 argument for the match() function, which reduces its stack requirements
460 slightly. This change also fixes an incompatibility with Perl: the pattern
461 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
462
463 10. More internal refactoring has drastically reduced the number of recursive
464 calls to match() for possessively repeated groups such as (abc)++ when
465 using pcre_exec().
466
467 11. While implementing 10, a number of bugs in the handling of groups were
468 discovered and fixed:
469
470 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
471 (a|)*(?1) gave a compile-time internal error.
472 ((a|)+)+ did not notice that the outer group could match an empty string.
473 (^a|^)+ was not marked as anchored.
474 (.*a|.*)+ was not marked as matching at start or after a newline.
475
476 12. Yet more internal refactoring has removed another argument from the match()
477 function. Special calls to this function are now indicated by setting a
478 value in a variable in the "match data" data block.
479
480 13. Be more explicit in pcre_study() instead of relying on "default" for
481 opcodes that mean there is no starting character; this means that when new
482 ones are added and accidentally left out of pcre_study(), testing should
483 pick them up.
484
485 14. The -s option of pcretest has been documented for ages as being an old
486 synonym of -m (show memory usage). I have changed it to mean "force study
487 for every regex", that is, assume /S for every regex. This is similar to -i
488 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
489 using it. It makes it easier to run collections of tests with and without
490 study enabled, and thereby test pcre_study() more easily. All the standard
491 tests are now run with and without -s (but some patterns can be marked as
492 "never study" - see 20 below).
493
494 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
495 restoration of the capturing data to the outer values was not happening
496 correctly.
497
498 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
499 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
500 pattern had matched an empty string, and so incorrectly returned a no
501 match.
502
503 17. There was optimizing code for the last branch of non-capturing parentheses,
504 and also for the obeyed branch of a conditional subexpression, which used
505 tail recursion to cut down on stack usage. Unfortunately, now that there is
506 the possibility of (*THEN) occurring in these branches, tail recursion is
507 no longer possible because the return has to be checked for (*THEN). These
508 two optimizations have therefore been removed. [But see 8.20/11 above.]
509
510 18. If a pattern containing \R was studied, it was assumed that \R always
511 matched two bytes, thus causing the minimum subject length to be
512 incorrectly computed because \R can also match just one byte.
513
514 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
515 was incorrectly computed.
516
517 20. If /S is present twice on a test pattern in pcretest input, it now
518 *disables* studying, thereby overriding the use of -s on the command line
519 (see 14 above). This is necessary for one or two tests to keep the output
520 identical in both cases.
521
522 21. When (*ACCEPT) was used in an assertion that matched an empty string and
523 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
524
525 22. When an atomic group that contained a capturing parenthesis was
526 successfully matched, but the branch in which it appeared failed, the
527 capturing was not being forgotten if a higher numbered group was later
528 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
529 group 1 to "a", when in fact it should be unset. This applied to multi-
530 branched capturing and non-capturing groups, repeated or not, and also to
531 positive assertions (capturing in negative assertions does not happen
532 in PCRE) and also to nested atomic groups.
533
534 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
535 subject after a captured substring, to make it easier to tell which of a
536 number of identical substrings has been captured.
537
538 24. The way atomic groups are processed by pcre_exec() has been changed so that
539 if they are repeated, backtracking one repetition now resets captured
540 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
541 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
542 "aaa". Previously, it would have been "a". As part of this code
543 refactoring, the way recursive calls are handled has also been changed.
544
545 25. If an assertion condition captured any substrings, they were not passed
546 back unless some other capturing happened later. For example, if
547 (?(?=(a))a) was matched against "a", no capturing was returned.
548
549 26. When studying a pattern that contained subroutine calls or assertions,
550 the code for finding the minimum length of a possible match was handling
551 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
552 group 1 called group 2 while simultaneously a separate group 2 called group
553 1). A stack overflow occurred in this case. I have fixed this by limiting
554 the recursion depth to 10.
555
556 27. Updated RunTest.bat in the distribution to the version supplied by Tom
557 Fortmann. This supports explicit test numbers on the command line, and has
558 argument validation and error reporting.
559
560 28. An instance of \X with an unlimited repeat could fail if at any point the
561 first character it looked at was a mark character.
562
563 29. Some minor code refactoring concerning Unicode properties and scripts
564 should reduce the stack requirement of match() slightly.
565
566 30. Added the '=' option to pcretest to check the setting of unused capturing
567 slots at the end of the pattern, which are documented as being -1, but are
568 not included in the return count.
569
570 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
571 compiled something random. Now it gives a compile-time error (as does
572 Perl).
573
574 32. A *MARK encountered during the processing of a positive assertion is now
575 recorded and passed back (compatible with Perl).
576
577 33. If --only-matching or --colour was set on a pcregrep call whose pattern
578 had alternative anchored branches, the search for a second match in a line
579 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
580 matched the line "0102" twice. The same bug affected patterns that started
581 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
582 twice.
583
584 34. Previously, PCRE did not allow quantification of assertions. However, Perl
585 does, and because of capturing effects, quantifying parenthesized
586 assertions may at times be useful. Quantifiers are now allowed for
587 parenthesized assertions.
588
589 35. A minor code tidy in pcre_compile() when checking options for \R usage.
590
591 36. \g was being checked for fancy things in a character class, when it should
592 just be a literal "g".
593
594 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
595 appearance of a nested POSIX class supersedes an apparent external class.
596 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
597 unescaped square brackets may also appear as part of class names. For
598 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
599 more like Perl. (But see 8.20/1 above.)
600
601 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
602 was because it thought it was \N{name}, which is not supported).
603
604 39. Add minix to OS list not supporting the -S option in pcretest.
605
606 40. PCRE tries to detect cases of infinite recursion at compile time, but it
607 cannot analyze patterns in sufficient detail to catch mutual recursions
608 such as ((?1))((?2)). There is now a runtime test that gives an error if a
609 subgroup is called recursively as a subpattern for a second time at the
610 same position in the subject string. In previous releases this might have
611 been caught by the recursion limit, or it might have run out of stack.
612
613 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
614 happen only once. PCRE was, however incorrectly giving a compile time error
615 "recursive call could loop indefinitely" because it cannot analyze the
616 pattern in sufficient detail. The compile time test no longer happens when
617 PCRE is compiling a conditional subpattern, but actual runaway loops are
618 now caught at runtime (see 40 above).
619
620 42. It seems that Perl allows any characters other than a closing parenthesis
621 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
622 has been changed to be the same.
623
624 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
625 as not to get warnings when autogen.sh is called. Also changed
626 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
627
628 44. To help people who use pcregrep to scan files containing exceedingly long
629 lines, the following changes have been made:
630
631 (a) The default value of the buffer size parameter has been increased from
632 8K to 20K. (The actual buffer used is three times this size.)
633
634 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
635 PCRE is built.
636
637 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
638 to be set at run time.
639
640 (d) Numerical values in pcregrep options can be followed by K or M, for
641 example --buffer-size=50K.
642
643 (e) If a line being scanned overflows pcregrep's buffer, an error is now
644 given and the return code is set to 2.
645
646 45. Add a pointer to the latest mark to the callout data block.
647
648 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
649 partial match of an empty string instead of no match. This was specific to
650 the use of ".".
651
652 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
653 complete match instead of a partial match. This bug was dependent on both
654 the PCRE_UTF8 and PCRE_DOTALL options being set.
655
656 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
657 starting byte set, because \b was not being ignored.
658
659
660 Version 8.12 15-Jan-2011
661 ------------------------
662
663 1. Fixed some typos in the markup of the man pages, and wrote a script that
664 checks for such things as part of the documentation building process.
665
666 2. On a big-endian 64-bit system, pcregrep did not correctly process the
667 --match-limit and --recursion-limit options (added for 8.11). In
668 particular, this made one of the standard tests fail. (The integer value
669 went into the wrong half of a long int.)
670
671 3. If the --colour option was given to pcregrep with -v (invert match), it
672 did strange things, either producing crazy output, or crashing. It should,
673 of course, ignore a request for colour when reporting lines that do not
674 match.
675
676 4. Another pcregrep bug caused similar problems if --colour was specified with
677 -M (multiline) and the pattern match finished with a line ending.
678
679 5. In pcregrep, when a pattern that ended with a literal newline sequence was
680 matched in multiline mode, the following line was shown as part of the
681 match. This seems wrong, so I have changed it.
682
683 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
684 the check for further matches in the same line (so they could be coloured)
685 to overrun the end of the current line. If another match was found, it was
686 incorrectly shown (and then shown again when found in the next line).
687
688 7. If pcregrep was compiled under Windows, there was a reference to the
689 function pcregrep_exit() before it was defined. I am assuming this was
690 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
691 reported by a user. I've moved the definition above the reference.
692
693
694 Version 8.11 10-Dec-2010
695 ------------------------
696
697 1. (*THEN) was not working properly if there were untried alternatives prior
698 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
699 backtracked to try for "b" instead of moving to the next alternative branch
700 at the same level (in this case, to look for "c"). The Perl documentation
701 is clear that when (*THEN) is backtracked onto, it goes to the "next
702 alternative in the innermost enclosing group".
703
704 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
705 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
706 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
707 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
708 (*THEN).
709
710 3. If \s appeared in a character class, it removed the VT character from
711 the class, even if it had been included by some previous item, for example
712 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
713 of \s, but is part of the POSIX "space" class.)
714
715 4. A partial match never returns an empty string (because you can always
716 match an empty string at the end of the subject); however the checking for
717 an empty string was starting at the "start of match" point. This has been
718 changed to the "earliest inspected character" point, because the returned
719 data for a partial match starts at this character. This means that, for
720 example, /(?<=abc)def/ gives a partial match for the subject "abc"
721 (previously it gave "no match").
722
723 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
724 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
725 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
726 has an implication that the given string is incomplete (because a partial
727 match is preferred over a full match). For this reason, these items now
728 give a partial match in this situation. [Aside: previously, the one case
729 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
730 match rather than a full match, which was wrong by the old rules, but is
731 now correct.]
732
733 6. There was a bug in the handling of #-introduced comments, recognized when
734 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
735 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
736 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
737 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
738 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
739 places in pcre_compile().
740
741 7. Related to (6) above, when pcre_compile() was skipping #-introduced
742 comments when looking ahead for named forward references to subpatterns,
743 the only newline sequence it recognized was NL. It now handles newlines
744 according to the set newline convention.
745
746 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
747 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
748 cater for a lack of strerror(). These oversights have been fixed.
749
750 9. Added --match-limit and --recursion-limit to pcregrep.
751
752 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
753
754 11. When the -o option was used, pcregrep was setting a return code of 1, even
755 when matches were found, and --line-buffered was not being honoured.
756
757 12. Added an optional parentheses number to the -o and --only-matching options
758 of pcregrep.
759
760 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
761 can match an empty string. The code to do it in pcretest and pcredemo
762 needed fixing:
763
764 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
765 only one byte after an empty string match just before CRLF (this case
766 just got forgotten; "any" and "anycrlf" were OK).
767
768 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
769 mode when an empty string match preceded an ASCII character followed by
770 a non-ASCII character. (The code for advancing by one character rather
771 than one byte was nonsense.)
772
773 (c) The pcredemo.c sample program did not have any code at all to handle
774 the cases when CRLF is a valid newline sequence.
775
776 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
777 as a starting offset was within the subject string. There is now a new
778 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
779 negative or greater than the length of the string. In order to test this,
780 pcretest is extended to allow the setting of negative starting offsets.
781
782 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
783 starting offset points to the beginning of a UTF-8 character was
784 unnecessarily clumsy. I tidied it up.
785
786 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
787 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
788
789 17. Nobody had reported that the --include_dir option, which was added in
790 release 7.7 should have been called --include-dir (hyphen, not underscore)
791 for compatibility with GNU grep. I have changed it to --include-dir, but
792 left --include_dir as an undocumented synonym, and the same for
793 --exclude-dir, though that is not available in GNU grep, at least as of
794 release 2.5.4.
795
796 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
797 characters from a string of bytes) have been redefined so as not to use
798 loops, in order to improve performance in some environments. At the same
799 time, I abstracted some of the common code into auxiliary macros to save
800 repetition (this should not affect the compiled code).
801
802 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
803 compile-time error is now given if \c is not followed by an ASCII
804 character, that is, a byte less than 128. (In EBCDIC mode, the code is
805 different, and any byte value is allowed.)
806
807 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
808 START_OPTIMIZE option, which is now allowed at compile time - but just
809 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
810 to pcregrep and other applications that have no direct access to PCRE
811 options. The new /Y option in pcretest sets this option when calling
812 pcre_compile().
813
814 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
815 back references. Groups containing recursive back references were forced to
816 be atomic by that change, but in the case of named groups, the amount of
817 memory required was incorrectly computed, leading to "Failed: internal
818 error: code overflow". This has been fixed.
819
820 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
821 pcretest.c, to avoid build problems in some Borland environments.
822
823
824 Version 8.10 25-Jun-2010
825 ------------------------
826
827 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
828 THEN.
829
830 2. (*ACCEPT) was not working when inside an atomic group.
831
832 3. Inside a character class, \B is treated as a literal by default, but
833 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
834 causes the error). The code is unchanged, but I tidied the documentation.
835
836 4. Inside a character class, PCRE always treated \R and \X as literals,
837 whereas Perl faults them if its -w option is set. I have changed PCRE so
838 that it faults them when PCRE_EXTRA is set.
839
840 5. Added support for \N, which always matches any character other than
841 newline. (It is the same as "." when PCRE_DOTALL is not set.)
842
843 6. When compiling pcregrep with newer versions of gcc which may have
844 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
845 declared with attribute warn_unused_result" were given. Just casting the
846 result to (void) does not stop the warnings; a more elaborate fudge is
847 needed. I've used a macro to implement this.
848
849 7. Minor change to pcretest.c to avoid a compiler warning.
850
851 8. Added four artifical Unicode properties to help with an option to make
852 \s etc use properties (see next item). The new properties are: Xan
853 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
854
855 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
856 use Unicode properties. (*UCP) at the start of a pattern can be used to set
857 this option. Modified pcretest to add /W to test this facility. Added
858 REG_UCP to make it available via the POSIX interface.
859
860 10. Added --line-buffered to pcregrep.
861
862 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
863 studied, and the match started with a letter with a code point greater than
864 127 whose first byte was different to the first byte of the other case of
865 the letter, the other case of this starting letter was not recognized
866 (#976).
867
868 12. If a pattern that was studied started with a repeated Unicode property
869 test, for example, \p{Nd}+, there was the theoretical possibility of
870 setting up an incorrect bitmap of starting bytes, but fortunately it could
871 not have actually happened in practice until change 8 above was made (it
872 added property types that matched character-matching opcodes).
873
874 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
875 possible starting bytes for non-anchored patterns.
876
877 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
878 \R, and also a number of cases that involve Unicode properties, both
879 explicit and implicit when PCRE_UCP is set.
880
881 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
882 input, it could crash or give wrong results if characters with values
883 greater than 0xc0 were present in the subject string. (Detail: it assumed
884 UTF-8 input when processing these items.)
885
886 16. Added a lot of (int) casts to avoid compiler warnings in systems where
887 size_t is 64-bit (#991).
888
889 17. Added a check for running out of memory when PCRE is compiled with
890 --disable-stack-for-recursion (#990).
891
892 18. If the last data line in a file for pcretest does not have a newline on
893 the end, a newline was missing in the output.
894
895 19. The default pcre_chartables.c file recognizes only ASCII characters (values
896 less than 128) in its various bitmaps. However, there is a facility for
897 generating tables according to the current locale when PCRE is compiled. It
898 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
899 characters, are recognized by isspace() and therefore were getting set in
900 these tables, and indeed these tables seem to approximate to ISO 8859. This
901 caused a problem in UTF-8 mode when pcre_study() was used to create a list
902 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
903 which of course cannot start UTF-8 characters. I have changed the code so
904 that only real ASCII characters (less than 128) and the correct starting
905 bytes for UTF-8 encodings are set for characters greater than 127 when in
906 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
907 altogether.)
908
909 20. Added the /T option to pcretest so as to be able to run tests with non-
910 standard character tables, thus making it possible to include the tests
911 used for 19 above in the standard set of tests.
912
913 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
914 reference to a subpattern the other side of a comment that contains an
915 opening parenthesis caused either an internal compiling error, or a
916 reference to the wrong subpattern.
917
918
919 Version 8.02 19-Mar-2010
920 ------------------------
921
922 1. The Unicode data tables have been updated to Unicode 5.2.0.
923
924 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
925 configured.
926
927 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
928 original author of that file, following a query about its status.
929
930 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
931 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
932
933 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
934 quantifier applied to a forward-referencing subroutine call, could compile
935 incorrect code or give the error "internal error: previously-checked
936 referenced subpattern not found".
937
938 6. Both MS Visual Studio and Symbian OS have problems with initializing
939 variables to point to external functions. For these systems, therefore,
940 pcre_malloc etc. are now initialized to local functions that call the
941 relevant global functions.
942
943 7. There were two entries missing in the vectors called coptable and poptable
944 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
945 I've fixed the data, and added a kludgy way of testing at compile time that
946 the lengths are correct (equal to the number of opcodes).
947
948 8. Following on from 7, I added a similar kludge to check the length of the
949 eint vector in pcreposix.c.
950
951 9. Error texts for pcre_compile() are held as one long string to avoid too
952 much relocation at load time. To find a text, the string is searched,
953 counting zeros. There was no check for running off the end of the string,
954 which could happen if a new error number was added without updating the
955 string.
956
957 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
958
959 11. \K was not working if it appeared in an atomic group or in a group that
960 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
961 \K is "not well defined" if used in an assertion. PCRE now accepts it if
962 the assertion is positive, but not if it is negative.
963
964 12. Change 11 fortuitously reduced the size of the stack frame used in the
965 "match()" function of pcre_exec.c by one pointer. Forthcoming
966 implementation of support for (*MARK) will need an extra pointer on the
967 stack; I have reserved it now, so that the stack frame size does not
968 decrease.
969
970 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
971 item in branch that calls a recursion is a subroutine call - as in the
972 second branch in the above example - was incorrectly given the compile-
973 time error "recursive call could loop indefinitely" because pcre_compile()
974 was not correctly checking the subroutine for matching a non-empty string.
975
976 14. The checks for overrunning compiling workspace could trigger after an
977 overrun had occurred. This is a "should never occur" error, but it can be
978 triggered by pathological patterns such as hundreds of nested parentheses.
979 The checks now trigger 100 bytes before the end of the workspace.
980
981 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
982
983
984 Version 8.01 19-Jan-2010
985 ------------------------
986
987 1. If a pattern contained a conditional subpattern with only one branch (in
988 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
989 computed the wrong minimum data length (which is of course zero for such
990 subpatterns). This could cause incorrect "no match" results.
991
992 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
993 the pattern is reset in the first branch, pcre_compile() failed with
994 "internal error: code overflow at offset...". This happened only when
995 the reset was to the original external option setting. (An optimization
996 abstracts leading options settings into an external setting, which was the
997 cause of this.)
998
999 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1000 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1001 assertion pattern did not match (meaning that the assertion was true), it
1002 was incorrectly treated as false if the SKIP had been reached during the
1003 matching. This also applied to assertions used as conditions.
1004
1005 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
1006 assertion subpattern, including such a pattern used as a condition,
1007 unpredictable results occurred, instead of the error return
1008 PCRE_ERROR_DFA_UITEM.
1009
1010 5. The C++ GlobalReplace function was not working like Perl for the special
1011 situation when an empty string is matched. It now does the fancy magic
1012 stuff that is necessary.
1013
1014 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1015 removed. (These were left over from very, very early versions of PCRE.)
1016
1017 7. Some cosmetic changes to the code to make life easier when compiling it
1018 as part of something else:
1019
1020 (a) Change DEBUG to PCRE_DEBUG.
1021
1022 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1023 called "current" as "current_branch", to prevent a collision with the
1024 Linux macro when compiled as a kernel module.
1025
1026 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1027 prevent a collision with the Linux macro when compiled as a kernel
1028 module.
1029
1030 8. In pcre_compile() there are some checks for integer overflows that used to
1031 cast potentially large values to (double). This has been changed to that
1032 when building, a check for int64_t is made, and if it is found, it is used
1033 instead, thus avoiding the use of floating point arithmetic. (There is no
1034 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1035 double.
1036
1037 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1038 2005 (difference between two addresses compared to an unsigned value).
1039
1040 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1041 custom one, because of the following reported problem in Windows:
1042
1043 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1044 under Win32.
1045 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1046 therefore missing the function definition.
1047 - The compiler thus generates a "C" signature for the test function.
1048 - The linker fails to find the "C" function.
1049 - PCRE fails to configure if asked to do so against libbz2.
1050
1051 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1052 messages were output:
1053
1054 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1055 rerunning libtoolize, to keep the correct libtool macros in-tree.
1056 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1057
1058 I have done both of these things.
1059
1060 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1061 most of the time, it *can* run out if it is given a pattern that contains a
1062 runaway infinite recursion. I updated the discussion in the pcrestack man
1063 page.
1064
1065 13. Now that we have gone to the x.xx style of version numbers, the minor
1066 version may start with zero. Using 08 or 09 is a bad idea because users
1067 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1068 interpreted as invalid octal numbers. I've updated the previous comment in
1069 configure.ac, and also added a check that gives an error if 08 or 09 are
1070 used.
1071
1072 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1073 causing partial matching to fail when the end of the subject matched \W
1074 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1075
1076 15. There were some discrepancies between the declarations in pcre_internal.h
1077 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1078 their definitions. The declarations used "const uschar *" and the
1079 definitions used USPTR. Even though USPTR is normally defined as "const
1080 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1081 reported that: "This difference in casting confuses some C++ compilers, for
1082 example, SunCC recognizes above declarations as different functions and
1083 generates broken code for hbpcre." I have changed the declarations to use
1084 USPTR.
1085
1086 16. GNU libtool is named differently on some systems. The autogen.sh script now
1087 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1088 (FreeBSD).
1089
1090 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1091 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1092 comment: "Figure out how to create a longlong from a string: strtoll and
1093 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1094 instance, but it only takes 2 args instead of 3!"
1095
1096 18. A subtle bug concerned with back references has been fixed by a change of
1097 specification, with a corresponding code fix. A pattern such as
1098 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1099 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1100 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1101 same bug. Such groups have to be quantified to be useful, or contained
1102 inside another quantified group. (If there's no repetition, the reference
1103 can never match.) The problem arises because, having left the group and
1104 moved on to the rest of the pattern, a later failure that backtracks into
1105 the group uses the captured value from the final iteration of the group
1106 rather than the correct earlier one. I have fixed this in PCRE by forcing
1107 any group that contains a reference to itself to be an atomic group; that
1108 is, there cannot be any backtracking into it once it has completed. This is
1109 similar to recursive and subroutine calls.
1110
1111
1112 Version 8.00 19-Oct-09
1113 ----------------------
1114
1115 1. The table for translating pcre_compile() error codes into POSIX error codes
1116 was out-of-date, and there was no check on the pcre_compile() error code
1117 being within the table. This could lead to an OK return being given in
1118 error.
1119
1120 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1121 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1122 in a Windows environment.
1123
1124 3. The pcregrep --count option prints the count for each file even when it is
1125 zero, as does GNU grep. However, pcregrep was also printing all files when
1126 --files-with-matches was added. Now, when both options are given, it prints
1127 counts only for those files that have at least one match. (GNU grep just
1128 prints the file name in this circumstance, but including the count seems
1129 more useful - otherwise, why use --count?) Also ensured that the
1130 combination -clh just lists non-zero counts, with no names.
1131
1132 4. The long form of the pcregrep -F option was incorrectly implemented as
1133 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1134 but it seems right to fix it, and I didn't think it was worth preserving
1135 the old behaviour.
1136
1137 5. The command line items --regex=pattern and --regexp=pattern were not
1138 recognized by pcregrep, which required --regex pattern or --regexp pattern
1139 (with a space rather than an '='). The man page documented the '=' forms,
1140 which are compatible with GNU grep; these now work.
1141
1142 6. No libpcreposix.pc file was created for pkg-config; there was just
1143 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1144
1145 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1146 when UCP support is not needed, by modifying the Python script that
1147 generates it from Unicode data files. This should not matter if the module
1148 is correctly used as a library, but I received one complaint about 50K of
1149 unwanted data. My guess is that the person linked everything into his
1150 program rather than using a library. Anyway, it does no harm.
1151
1152 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1153 was a minimum greater than 1 for a wide character in a possessive
1154 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1155 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1156 character. Chaos in the form of incorrect output or a compiling loop could
1157 result.
1158
1159 9. The restrictions on what a pattern can contain when partial matching is
1160 requested for pcre_exec() have been removed. All patterns can now be
1161 partially matched by this function. In addition, if there are at least two
1162 slots in the offset vector, the offset of the earliest inspected character
1163 for the match and the offset of the end of the subject are set in them when
1164 PCRE_ERROR_PARTIAL is returned.
1165
1166 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1167 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1168 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1169 and may be more useful for multi-segment matching.
1170
1171 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1172 used to be given if ever the end of the subject was reached; now it is
1173 given only if matching could not proceed because another character was
1174 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1175 string "Z", which now yields "no match" instead of "partial match". In the
1176 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1177 final character ended with (*FAIL).
1178
1179 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1180 if the pattern had a "must contain" character that was already found in the
1181 earlier partial match, unless partial matching was again requested. For
1182 example, with the pattern /dog.(body)?/, the "must contain" character is
1183 "g". If the first part-match was for the string "dog", restarting with
1184 "sbody" failed. This bug has been fixed.
1185
1186 13. The string returned by pcre_dfa_exec() after a partial match has been
1187 changed so that it starts at the first inspected character rather than the
1188 first character of the match. This makes a difference only if the pattern
1189 starts with a lookbehind assertion or \b or \B (\K is not supported by
1190 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1191 matching functions compatible, and I think it's the right thing to do.
1192
1193 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1194 so that the demonstration program is easily available in environments where
1195 PCRE has not been installed from source.
1196
1197 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1198 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1199 library.
1200
1201 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1202 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1203 is not the first non-POSIX option to be added. Clearly some people find
1204 these options useful.
1205
1206 17. If a caller to the POSIX matching function regexec() passes a non-zero
1207 value for nmatch with a NULL value for pmatch, the value of
1208 nmatch is forced to zero.
1209
1210 18. RunGrepTest did not have a test for the availability of the -u option of
1211 the diff command, as RunTest does. It now checks in the same way as
1212 RunTest, and also checks for the -b option.
1213
1214 19. If an odd number of negated classes containing just a single character
1215 interposed, within parentheses, between a forward reference to a named
1216 subpattern and the definition of the subpattern, compilation crashed with
1217 an internal error, complaining that it could not find the referenced
1218 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1219 [The bug was that it was starting one character too far in when skipping
1220 over the character class, thus treating the ] as data rather than
1221 terminating the class. This meant it could skip too much.]
1222
1223 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1224 /g option in pcretest when the pattern contains \K, which makes it possible
1225 to have an empty string match not at the start, even when the pattern is
1226 anchored. Updated pcretest and pcredemo to use this option.
1227
1228 21. If the maximum number of capturing subpatterns in a recursion was greater
1229 than the maximum at the outer level, the higher number was returned, but
1230 with unset values at the outer level. The correct (outer level) value is
1231 now given.
1232
1233 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1234 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1235 make it do so. The string so far is captured, making this feature
1236 compatible with Perl.
1237
1238 23. The tests have been re-organized, adding tests 11 and 12, to make it
1239 possible to check the Perl 5.10 features against Perl 5.10.
1240
1241 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1242 pattern matches a fixed length string. PCRE did not allow this; now it
1243 does. Neither allows recursion.
1244
1245 25. I finally figured out how to implement a request to provide the minimum
1246 length of subject string that was needed in order to match a given pattern.
1247 (It was back references and recursion that I had previously got hung up
1248 on.) This code has now been added to pcre_study(); it finds a lower bound
1249 to the length of subject needed. It is not necessarily the greatest lower
1250 bound, but using it to avoid searching strings that are too short does give
1251 some useful speed-ups. The value is available to calling programs via
1252 pcre_fullinfo().
1253
1254 26. While implementing 25, I discovered to my embarrassment that pcretest had
1255 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1256 study optimizations had never been tested with that matching function.
1257 Oops. What is worse, even when it was passed study data, there was a bug in
1258 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1259 were also very few tests of studied patterns with pcre_dfa_exec().
1260
1261 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1262 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1263 on the other side of the coin, they are no longer allowed to have different
1264 names, because these cannot be distinguished in PCRE, and this has caused
1265 confusion. (This is a difference from Perl.)
1266
1267 28. When duplicate subpattern names are present (necessarily with different
1268 numbers, as required by 27 above), and a test is made by name in a
1269 conditional pattern, either for a subpattern having been matched, or for
1270 recursion in such a pattern, all the associated numbered subpatterns are
1271 tested, and the overall condition is true if the condition is true for any
1272 one of them. This is the way Perl works, and is also more like the way
1273 testing by number works.
1274
1275
1276 Version 7.9 11-Apr-09
1277 ---------------------
1278
1279 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1280 (pcretest), all targets were linked against these libraries. This included
1281 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1282 libraries. This caused unwanted dependencies to be created. This problem
1283 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1284 pcretest is linked with readline.
1285
1286 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1287 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1288 moved outside it again, because FALSE and TRUE are already defined in AIX,
1289 but BOOL is not.
1290
1291 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1292 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1293
1294 4. The pcregrep documentation said spaces were inserted as well as colons (or
1295 hyphens) following file names and line numbers when outputting matching
1296 lines. This is not true; no spaces are inserted. I have also clarified the
1297 wording for the --colour (or --color) option.
1298
1299 5. In pcregrep, when --colour was used with -o, the list of matching strings
1300 was not coloured; this is different to GNU grep, so I have changed it to be
1301 the same.
1302
1303 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1304 each matching line was coloured. Now it goes on to look for further matches
1305 of any of the test patterns, which is the same behaviour as GNU grep.
1306
1307 7. A pattern that could match an empty string could cause pcregrep to loop; it
1308 doesn't make sense to accept an empty string match in pcregrep, so I have
1309 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1310 seems to be how GNU grep behaves.
1311
1312 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1313 start or after a newline", because the conditional assertion was not being
1314 correctly handled. The rule now is that both the assertion and what follows
1315 in the first alternative must satisfy the test.
1316
1317 9. If auto-callout was enabled in a pattern with a conditional group whose
1318 condition was an assertion, PCRE could crash during matching, both with
1319 pcre_exec() and pcre_dfa_exec().
1320
1321 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1322 used for matching.
1323
1324 11. Unicode property support in character classes was not working for
1325 characters (bytes) greater than 127 when not in UTF-8 mode.
1326
1327 12. Added the -M command line option to pcretest.
1328
1329 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1330
1331 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1332
1333 16. Added comments and documentation about mis-use of no_arg in the C++
1334 wrapper.
1335
1336 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1337 from Martin Jerabek that uses macro names for all relevant character and
1338 string constants.
1339
1340 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1341 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1342 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1343 these, but not everybody uses configure.
1344
1345 19. A conditional group that had only one branch was not being correctly
1346 recognized as an item that could match an empty string. This meant that an
1347 enclosing group might also not be so recognized, causing infinite looping
1348 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1349 with the subject "ab", where knowledge that the repeated group can match
1350 nothing is needed in order to break the loop.
1351
1352 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1353 exec(), but without supplying a callout function, matching went wrong.
1354
1355 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1356 leak if the size of the offset vector was greater than 30. When the vector
1357 is smaller, the saved offsets during recursion go onto a local stack
1358 vector, but for larger vectors malloc() is used. It was failing to free
1359 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1360 error, in fact).
1361
1362 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1363 heapframe that is used only when UTF-8 support is enabled. This caused no
1364 problem, but was untidy.
1365
1366 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1367 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1368 included within another project.
1369
1370 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1371 slightly modified by me:
1372
1373 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1374 not building pcregrep.
1375
1376 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1377 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1378
1379 25. Forward references, both numeric and by name, in patterns that made use of
1380 duplicate group numbers, could behave incorrectly or give incorrect errors,
1381 because when scanning forward to find the reference group, PCRE was not
1382 taking into account the duplicate group numbers. A pattern such as
1383 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1384
1385 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1386 the feature of a custom pointer more persuasive (as requested by a user).
1387
1388 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1389 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1390 pre-defined.
1391
1392 28. Added support for (*UTF8) at the start of a pattern.
1393
1394 29. Arrange for flags added by the "release type" setting in CMake to be shown
1395 in the configuration summary.
1396
1397
1398 Version 7.8 05-Sep-08
1399 ---------------------
1400
1401 1. Replaced UCP searching code with optimized version as implemented for Ad
1402 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1403 stage table and inline lookup instead of a function, giving speed ups of 2
1404 to 5 times on some simple patterns that I tested. Permission was given to
1405 distribute the MultiStage2.py script that generates the tables (it's not in
1406 the tarball, but is in the Subversion repository).
1407
1408 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1409 scripts.
1410
1411 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1412 a group with a zero qualifier. The result of the study could be incorrect,
1413 or the function might crash, depending on the pattern.
1414
1415 4. Caseless matching was not working for non-ASCII characters in back
1416 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1417 It now works when Unicode Property Support is available.
1418
1419 5. In pcretest, an escape such as \x{de} in the data was always generating
1420 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1421 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1422 truncation.
1423
1424 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1425
1426 7. Added two (int) casts to pcregrep when printing the difference of two
1427 pointers, in case they are 64-bit values.
1428
1429 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1430 test 2 if it fails.
1431
1432 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1433 and a #define of that name to empty if it is not externally set. This is to
1434 allow users of MSVC to set it if necessary.
1435
1436 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1437 the convenience functions in the pcre_get.c source file.
1438
1439 11. An option change at the start of a pattern that had top-level alternatives
1440 could cause overwriting and/or a crash. This command provoked a crash in
1441 some environments:
1442
1443 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1444
1445 This potential security problem was recorded as CVE-2008-2371.
1446
1447 12. For a pattern where the match had to start at the beginning or immediately
1448 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1449 pcre_dfa_exec() could read past the end of the passed subject if there was
1450 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1451 pcretest so that it places the subject at the end of its malloc-ed buffer.
1452
1453 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1454 exec() might read past the end of the data buffer in UTF-8 mode.
1455
1456 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1457 the data contained the byte 0x85 as part of a UTF-8 character within its
1458 first line. This applied both to normal and DFA matching.
1459
1460 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1461 /^[^d]*?$/8 failed to match "abc".
1462
1463 16. Added a missing copyright notice to pcrecpp_internal.h.
1464
1465 17. Make it more clear in the documentation that values returned from
1466 pcre_exec() in ovector are byte offsets, not character counts.
1467
1468 18. Tidied a few places to stop certain compilers from issuing warnings.
1469
1470 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1471 supplied by Stefan Weber. I made a further small update for 7.8 because
1472 there is a change of source arrangements: the pcre_searchfuncs.c module is
1473 replaced by pcre_ucd.c.
1474
1475
1476 Version 7.7 07-May-08
1477 ---------------------
1478
1479 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1480 a string to a long long, pretend we don't even have a long long." This is
1481 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1482
1483 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1484 pre-7.6 versions, which defined a global no_arg variable instead of putting
1485 it in the RE class. (See also #8 below.)
1486
1487 3. Remove a line of dead code, identified by coverity and reported by Nuno
1488 Lopes.
1489
1490 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1491
1492 (1) The include/exclude patterns were being applied to the whole pathnames
1493 of files, instead of just to the final components.
1494
1495 (2) If there was more than one level of directory, the subdirectories were
1496 skipped unless they satisfied the include/exclude conditions. This is
1497 inconsistent with GNU grep (and could even be seen as contrary to the
1498 pcregrep specification - which I improved to make it absolutely clear).
1499 The action now is always to scan all levels of directory, and just
1500 apply the include/exclude patterns to regular files.
1501
1502 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1503 --exclude_dir in the tests to avoid scanning .svn directories.
1504
1505 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1506 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1507 doesn't support NULs in patterns.
1508
1509 7. Added some missing "const"s to declarations of static tables in
1510 pcre_compile.c and pcre_dfa_exec.c.
1511
1512 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1513 caused by fix #2 above. (Subsequently also a second patch to fix the
1514 first patch. And a third patch - this was a messy problem.)
1515
1516 9. Applied Craig's patch to remove the use of push_back().
1517
1518 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1519 matching function regexec().
1520
1521 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1522 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1523 references. PCRE supports relative numbers with this syntax (I don't think
1524 Oniguruma does).
1525
1526 12. Previously, a group with a zero repeat such as (...){0} was completely
1527 omitted from the compiled regex. However, this means that if the group
1528 was called as a subroutine from elsewhere in the pattern, things went wrong
1529 (an internal error was given). Such groups are now left in the compiled
1530 pattern, with a new opcode that causes them to be skipped at execution
1531 time.
1532
1533 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1534 to the way PCRE behaves:
1535
1536 (a) A lone ] character is dis-allowed (Perl treats it as data).
1537
1538 (b) A back reference to an unmatched subpattern matches an empty string
1539 (Perl fails the current match path).
1540
1541 (c) A data ] in a character class must be notated as \] because if the
1542 first data character in a class is ], it defines an empty class. (In
1543 Perl it is not possible to have an empty class.) The empty class []
1544 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1545 The negative empty class [^] matches any one character, independently
1546 of the DOTALL setting.
1547
1548 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1549 non-existent subpattern following a character class starting with ']' and
1550 containing () gave an internal compiling error instead of "reference to
1551 non-existent subpattern". Fortunately, when the pattern did exist, the
1552 compiled code was correct. (When scanning forwards to check for the
1553 existencd of the subpattern, it was treating the data ']' as terminating
1554 the class, so got the count wrong. When actually compiling, the reference
1555 was subsequently set up correctly.)
1556
1557 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1558 it was being rejected as not supported by pcre_dfa_exec(), even though
1559 other assertions are supported. I have made pcre_dfa_exec() support
1560 (*FAIL).
1561
1562 16. The implementation of 13c above involved the invention of a new opcode,
1563 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1564 cannot be changed at match time, I realized I could make a small
1565 improvement to matching performance by compiling OP_ALLANY instead of
1566 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1567 on the OP_ANY path.
1568
1569 17. Compiling pcretest on Windows with readline support failed without the
1570 following two fixes: (1) Make the unistd.h include conditional on
1571 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1572
1573 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1574 ncurses library to be included for pcretest when ReadLine support is
1575 requested, but also to allow for it to be overridden. This patch came from
1576 Daniel Bergström.
1577
1578 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1579 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1580 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1581 spotting this.
1582
1583
1584 Version 7.6 28-Jan-08
1585 ---------------------
1586
1587 1. A character class containing a very large number of characters with
1588 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1589 overflow.
1590
1591 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1592 HAVE_LONG_LONG is not defined.
1593
1594 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1595 bring them up to date and include new features. This patch includes:
1596
1597 - Fixed PH's badly added libz and libbz2 support.
1598 - Fixed a problem with static linking.
1599 - Added pcredemo. [But later removed - see 7 below.]
1600 - Fixed dftables problem and added an option.
1601 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1602 HAVE_LONG_LONG.
1603 - Added readline support for pcretest.
1604 - Added an listing of the option settings after cmake has run.
1605
1606 4. A user submitted a patch to Makefile that makes it easy to create
1607 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1608 Makefile.am that cause it to include this special target, without
1609 affecting anything else. Note that the same mingw target plus all
1610 the other distribution libraries and programs are now supported
1611 when configuring with CMake (see 6 below) instead of with
1612 Configure/Make.
1613
1614 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1615 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1616 exported in the Windows port". It has not yet been confirmed that the patch
1617 solves the problem, but it does no harm.
1618
1619 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1620 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1621 with CMake, and also correct the comment about stack recursion.
1622
1623 7. Remove the automatic building of pcredemo from the ./configure system and
1624 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1625 of a program that users should build themselves after PCRE is installed, so
1626 building it automatically is not really right. What is more, it gave
1627 trouble in some build environments.
1628
1629 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1630
1631
1632 Version 7.5 10-Jan-08
1633 ---------------------
1634
1635 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1636 values in parens when parsing an RE using the C++ wrapper."
1637
1638 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1639 Characters greater than 255 were excluded from the class instead of being
1640 included.
1641
1642 3. The same bug as (2) above applied to negated POSIX classes such as
1643 [:^space:].
1644
1645 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1646 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1647 I have changed it.
1648
1649 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1650 first named subpattern) and a construct such as (?&a) would reference the
1651 first named subpattern whose name started with "a" (in other words, the
1652 length check was missing). Both these problems are fixed. "Subpattern name
1653 expected" is now given for (?&) (a zero-length name), and this patch also
1654 makes it give the same error for \k'' (previously it complained that that
1655 was a reference to a non-existent subpattern).
1656
1657 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1658 this is right because (?- can be followed by option settings as well as by
1659 digits. I have, however, made the messages clearer.
1660
1661 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1662 than the number used in the conditional) now cause a compile-time error.
1663 This is actually not compatible with Perl, which accepts such patterns, but
1664 treats the conditional as always being FALSE (as PCRE used to), but it
1665 seems to me that giving a diagnostic is better.
1666
1667 8. Change "alphameric" to the more common word "alphanumeric" in comments
1668 and messages.
1669
1670 9. Fix two occurrences of "backslash" in comments that should have been
1671 "backspace".
1672
1673 10. Remove two redundant lines of code that can never be obeyed (their function
1674 was moved elsewhere).
1675
1676 11. The program that makes PCRE's Unicode character property table had a bug
1677 which caused it to generate incorrect table entries for sequences of
1678 characters that have the same character type, but are in different scripts.
1679 It amalgamated them into a single range, with the script of the first of
1680 them. In other words, some characters were in the wrong script. There were
1681 thirteen such cases, affecting characters in the following ranges:
1682
1683 U+002b0 - U+002c1
1684 U+0060c - U+0060d
1685 U+0061e - U+00612
1686 U+0064b - U+0065e
1687 U+0074d - U+0076d
1688 U+01800 - U+01805
1689 U+01d00 - U+01d77
1690 U+01d9b - U+01dbf
1691 U+0200b - U+0200f
1692 U+030fc - U+030fe
1693 U+03260 - U+0327f
1694 U+0fb46 - U+0fbb1
1695 U+10450 - U+1049d
1696
1697 12. The -o option (show only the matching part of a line) for pcregrep was not
1698 compatible with GNU grep in that, if there was more than one match in a
1699 line, it showed only the first of them. It now behaves in the same way as
1700 GNU grep.
1701
1702 13. If the -o and -v options were combined for pcregrep, it printed a blank
1703 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1704 does the same. The return code can be used to tell if there were any
1705 non-matching lines.
1706
1707 14. Added --file-offsets and --line-offsets to pcregrep.
1708
1709 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1710 infinitely looping recursion. The bug was that positive lookaheads were not
1711 being skipped when checking for a possible empty match (negative lookaheads
1712 and both kinds of lookbehind were skipped).
1713
1714 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1715 inclusion of <windows.h> to before rather than after the definition of
1716 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1717
1718 17. Specifying a possessive quantifier with a specific limit for a Unicode
1719 character property caused pcre_compile() to compile bad code, which led at
1720 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1721 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1722 caused the error; without that there was no problem.
1723
1724 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1725
1726 19. Added --enable-pcretest-libreadline.
1727
1728 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1729 RE::GlobalReplace(). As a result, the number of replacements returned was
1730 double what it should be. I removed one of the increments, but Craig sent a
1731 later patch that removed the other one (the right fix) and added unit tests
1732 that check the return values (which was not done before).
1733
1734 21. Several CMake things:
1735
1736 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1737 the names libpcre and libpcreposix, not just pcre and pcreposix.
1738
1739 (2) The above change means that pcretest and pcregrep are now correctly
1740 linked with the newly-built libraries, not previously installed ones.
1741
1742 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1743
1744 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1745 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1746 UTF-8 newline character). The key issue is that the pattern starts .*;
1747 this means that the match must be either at the beginning, or after a
1748 newline. The bug was in the code for advancing after a failed match and
1749 checking that the new position followed a newline. It was not taking
1750 account of UTF-8 characters correctly.
1751
1752 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1753 character classes. PCRE was not treating the sequence [:...:] as a
1754 character class unless the ... were all letters. Perl, however, seems to
1755 allow any characters between [: and :], though of course it rejects as
1756 unknown any "names" that contain non-letters, because all the known class
1757 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1758 for example, whereas PCRE did not - it did not recognize a POSIX character
1759 class. This seemed a bit dangerous, so the code has been changed to be
1760 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1761 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1762 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1763 Perl does, and where it didn't before.
1764
1765 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1766 Windows environments %n is disabled by default.
1767
1768
1769 Version 7.4 21-Sep-07
1770 ---------------------
1771
1772 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1773 means that a class such as [\s] counted as "explicit reference to CR or
1774 LF". That isn't really right - the whole point of the change was to try to
1775 help when there was an actual mention of one of the two characters. So now
1776 the change happens only if \r or \n (or a literal CR or LF) character is
1777 encountered.
1778
1779 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1780 of both had grown to the point where there were only 3 bits left.
1781 Fortunately, there was spare space in the data structure, and so I have
1782 moved the internal flags into a new 16-bit field to free up more option
1783 bits.
1784
1785 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1786 but did not set the internal JCHANGED flag - either of these is enough to
1787 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1788 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1789 start sets both bits.
1790
1791 4. Added options (at build time, compile time, exec time) to change \R from
1792 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1793
1794 5. doc/pcresyntax.html was missing from the distribution.
1795
1796 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1797 compatibility, even though it is no longer used.
1798
1799 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1800 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1801 windows.h file is present (where different names are used). [This was
1802 reversed later after testing - see 16 below.]
1803
1804 8. Changed all #include <config.h> to #include "config.h". There were also
1805 some further <pcre.h> cases that I changed to "pcre.h".
1806
1807 9. When pcregrep was used with the --colour option, it missed the line ending
1808 sequence off the lines that it output.
1809
1810 10. It was pointed out to me that arrays of string pointers cause lots of
1811 relocations when a shared library is dynamically loaded. A technique of
1812 using a single long string with a table of offsets can drastically reduce
1813 these. I have refactored PCRE in four places to do this. The result is
1814 dramatic:
1815
1816 Originally: 290
1817 After changing UCP table: 187
1818 After changing error message table: 43
1819 After changing table of "verbs" 36
1820 After changing table of Posix names 22
1821
1822 Thanks to the folks working on Gregex for glib for this insight.
1823
1824 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1825 unicode-properties was also set.
1826
1827 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1828
1829 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1830 checked only for CRLF.
1831
1832 14. Added casts to pcretest.c to avoid compiler warnings.
1833
1834 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1835
1836 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1837 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1838 entirely. This removes changes made in 7 above.
1839
1840 17. The CMake files have been updated, and there is now more information about
1841 building with CMake in the NON-UNIX-USE document.
1842
1843
1844 Version 7.3 28-Aug-07
1845 ---------------------
1846
1847 1. In the rejigging of the build system that eventually resulted in 7.1, the
1848 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1849 brackets there is not right, since it causes compilers to look for an
1850 installed pcre.h, not the version that is in the source that is being
1851 compiled (which of course may be different). I have changed it back to:
1852
1853 #include "pcre.h"
1854
1855 I have a vague recollection that the change was concerned with compiling in
1856 different directories, but in the new build system, that is taken care of
1857 by the VPATH setting the Makefile.
1858
1859 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1860 when the subject happened to end in the byte 0x85 (e.g. if the last
1861 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1862 characters but of course it shouldn't be taken as a newline when it is part
1863 of another character. The bug was that, for an unlimited repeat of . in
1864 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1865 characters when looking for a newline.
1866
1867 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1868
1869 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1870 in debug output.
1871
1872 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1873 long printing in the pcrecpp unittest when running under MinGW.
1874
1875 6. ESC_K was left out of the EBCDIC table.
1876
1877 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1878 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1879 limit also applies to "virtual nesting" when a pattern is recursive, and in
1880 this case 1000 isn't so big. I have been able to remove this limit at the
1881 expense of backing off one optimization in certain circumstances. Normally,
1882 when pcre_exec() would call its internal match() function recursively and
1883 immediately return the result unconditionally, it uses a "tail recursion"
1884 feature to save stack. However, when a subpattern that can match an empty
1885 string has an unlimited repetition quantifier, it no longer makes this
1886 optimization. That gives it a stack frame in which to save the data for
1887 checking that an empty string has been matched. Previously this was taken
1888 from the 1000-entry workspace that had been reserved. So now there is no
1889 explicit limit, but more stack is used.
1890
1891 8. Applied Daniel's patches to solve problems with the import/export magic
1892 syntax that is required for Windows, and which was going wrong for the
1893 pcreposix and pcrecpp parts of the library. These were overlooked when this
1894 problem was solved for the main library.
1895
1896 9. There were some crude static tests to avoid integer overflow when computing
1897 the size of patterns that contain repeated groups with explicit upper
1898 limits. As the maximum quantifier is 65535, the maximum group length was
1899 set at 30,000 so that the product of these two numbers did not overflow a
1900 32-bit integer. However, it turns out that people want to use groups that
1901 are longer than 30,000 bytes (though not repeat them that many times).
1902 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1903 made it possible to implement the integer overflow checks in a much more
1904 dynamic way, which I have now done. The artificial limitation on group
1905 length has been removed - we now have only the limit on the total length of
1906 the compiled pattern, which depends on the LINK_SIZE setting.
1907
1908 10. Fixed a bug in the documentation for get/copy named substring when
1909 duplicate names are permitted. If none of the named substrings are set, the
1910 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1911 empty string.
1912
1913 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1914 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1915 because the ] is interpreted as the first data character and the
1916 terminating ] is not found. PCRE has been made compatible with Perl in this
1917 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1918 cause memory overwriting.
1919
1920 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1921 string has been matched (to stop an infinite loop). It was not recognizing
1922 a conditional subpattern that could match an empty string if that
1923 subpattern was within another subpattern. For example, it looped when
1924 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1925 condition was not nested. This bug has been fixed.
1926
1927 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1928 past the start of the subject in the presence of bytes with the top bit
1929 set, for example "\x8aBCD".
1930
1931 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1932 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1933
1934 14. Optimized (?!) to (*FAIL).
1935
1936 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1937 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1938 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1939 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1940 does: it's just the validity check that is more restrictive.
1941
1942 16. Inserted checks for integer overflows during escape sequence (backslash)
1943 processing, and also fixed erroneous offset values for syntax errors during
1944 backslash processing.
1945
1946 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1947 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1948
1949 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1950 caused an overrun.
1951
1952 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1953 something other than just ASCII characters) inside a group that had an
1954 unlimited repeat caused a loop at compile time (while checking to see
1955 whether the group could match an empty string).
1956
1957 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1958 [\P{Any}] did so. (Error in the code for printing property names.)
1959
1960 21. An orphan \E inside a character class could cause a crash.
1961
1962 22. A repeated capturing bracket such as (A)? could cause a wild memory
1963 reference during compilation.
1964
1965 23. There are several functions in pcre_compile() that scan along a compiled
1966 expression for various reasons (e.g. to see if it's fixed length for look
1967 behind). There were bugs in these functions when a repeated \p or \P was
1968 present in the pattern. These operators have additional parameters compared
1969 with \d, etc, and these were not being taken into account when moving along
1970 the compiled data. Specifically:
1971
1972 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1973 length.
1974
1975 (b) An item such as \pL+ within a repeated group could cause crashes or
1976 loops.
1977
1978 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1979 "reference to non-existent subpattern" error.
1980
1981 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1982
1983 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1984 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1985
1986 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1987
1988 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1989 character were causing crashes (broken optimization).
1990
1991 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1992 \p or \P) caused a compile-time loop.
1993
1994 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1995 break. For example, the unstudied pattern [\r\n]A does not match the string
1996 "\r\nA" because change 7.0/46 below moves the current point on by two
1997 characters after failing to match at the start. However, the pattern \nA
1998 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1999 the same is true. There doesn't seem any very clean way out of this, but
2000 what I have chosen to do makes the common cases work: PCRE now takes note
2001 of whether there can be an explicit match for \r or \n anywhere in the
2002 pattern, and if so, 7.0/46 no longer applies. As part of this change,
2003 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2004 pattern has explicit CR or LF references.
2005
2006 29. Added (*CR) etc for changing newline setting at start of pattern.
2007
2008
2009 Version 7.2 19-Jun-07
2010 ---------------------
2011
2012 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2013 which is apparently normally available under Windows.
2014
2015 2. Re-jig the pcregrep tests with different newline settings in an attempt
2016 to make them independent of the local environment's newline setting.
2017
2018 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2019
2020 4. Some of the "internals" tests were previously cut out when the link size
2021 was not 2, because the output contained actual offsets. The recent new
2022 "Z" feature of pcretest means that these can be cut out, making the tests
2023 usable with all link sizes.
2024
2025 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2026 stack recursion. This gives a massive performance boost under BSD, but just
2027 a small improvement under Linux. However, it saves one field in the frame
2028 in all cases.
2029
2030 6. Added more features from the forthcoming Perl 5.10:
2031
2032 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2033 recursion call. It refers to the nth most recently opened parentheses.
2034
2035 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2036 to be opened parentheses.
2037
2038 (c) Conditions that refer to capturing parentheses can be specified
2039 relatively, for example, (?(-2)... or (?(+3)...
2040
2041 (d) \K resets the start of the current match so that everything before
2042 is not part of it.
2043
2044 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2045
2046 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2047 reference syntax.
2048
2049 (g) (?| introduces a group in which the numbering of parentheses in each
2050 alternative starts with the same number.
2051
2052 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2053
2054 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2055 PCRE_INFO_JCHANGED.
2056
2057 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2058 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2059 for detecting groups that can match an empty string.
2060
2061 9. A pattern with a very large number of alternatives (more than several
2062 hundred) was running out of internal workspace during the pre-compile
2063 phase, where pcre_compile() figures out how much memory will be needed. A
2064 bit of new cunning has reduced the workspace needed for groups with
2065 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2066 workspace instead of running out of the 4096 that are available.
2067
2068 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2069
2070 11. Applied patch from Google to remove an optimization that didn't quite work.
2071 The report of the bug said:
2072
2073 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2074 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2075 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2076
2077 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2078 it matched the wrong number of bytes.
2079
2080
2081 Version 7.1 24-Apr-07
2082 ---------------------
2083
2084 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2085 that is more "standard", making use of automake and other Autotools. There
2086 is some re-arrangement of the files and adjustment of comments consequent
2087 on this.
2088
2089 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2090 for recursive directory scanning broke on some systems because the files
2091 are not scanned in any specific order and on different systems the order
2092 was different. A call to "sort" has been inserted into RunGrepTest for the
2093 approprate test as a short-term fix. In the longer term there may be an
2094 alternative.
2095
2096 3. I had an email from Eric Raymond about problems translating some of PCRE's
2097 man pages to HTML (despite the fact that I distribute HTML pages, some
2098 people do their own conversions for various reasons). The problems
2099 concerned the use of low-level troff macros .br and .in. I have therefore
2100 removed all such uses from the man pages (some were redundant, some could
2101 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2102 HTML has been updated to handle .nf/.fi and to complain if it encounters
2103 .br or .in.
2104
2105 4. Updated comments in configure.ac that get placed in config.h.in and also
2106 arranged for config.h to be included in the distribution, with the name
2107 config.h.generic, for the benefit of those who have to compile without
2108 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2109
2110 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2111 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2112 makevp.bat for the current PCRE, using the additional files
2113 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2114
2115 6. A Windows user reported a minor discrepancy with test 2, which turned out
2116 to be caused by a trailing space on an input line that had got lost in his
2117 copy. The trailing space was an accident, so I've just removed it.
2118
2119 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2120 that is needed.
2121
2122 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2123 as "const" (a) because they are and (b) because it helps the PHP
2124 maintainers who have recently made a script to detect big data structures
2125 in the php code that should be moved to the .rodata section. I remembered
2126 to update Builducptable as well, so it won't revert if ucptable.h is ever
2127 re-created.
2128
2129 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2130 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2131 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2132 support is not required. This saves 1.5-2K of code, which is important in
2133 some applications.
2134
2135 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2136 so as not to refer to the tables, even though these functions will never be
2137 called when UTF-8 support is disabled. Otherwise there are problems with a
2138 shared library.
2139
2140 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2141
2142 (a) It was defining its arguments as char * instead of void *.
2143
2144 (b) It was assuming that all moves were upwards in memory; this was true
2145 a long time ago when I wrote it, but is no longer the case.
2146
2147 The emulated memove() is provided for those environments that have neither
2148 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2149 is clearly not the case, as these two bugs were recently reported.
2150
2151 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2152 and Detrail to create the HTML documentation, the .txt form of the man
2153 pages, and it removes trailing spaces from listed files. It also creates
2154 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2155 case, it wraps all the #defines with #ifndefs. This script should be run
2156 before "make dist".
2157
2158 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2159 with Unicode property support.
2160
2161 (a) For a maximizing quantifier, if the two different cases of the
2162 character were of different lengths in their UTF-8 codings (there are
2163 some cases like this - I found 11), and the matching function had to
2164 back up over a mixture of the two cases, it incorrectly assumed they
2165 were both the same length.
2166
2167 (b) When PCRE was configured to use the heap rather than the stack for
2168 recursion during matching, it was not correctly preserving the data for
2169 the other case of a UTF-8 character when checking ahead for a match
2170 while processing a minimizing repeat. If the check also involved
2171 matching a wide character, but failed, corruption could cause an
2172 erroneous result when trying to check for a repeat of the original
2173 character.
2174
2175 13. Some tidying changes to the testing mechanism:
2176
2177 (a) The RunTest script now detects the internal link size and whether there
2178 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2179 values substituted by "configure". (The RunGrepTest script already did
2180 this for UTF-8.) The configure.ac script no longer substitutes the
2181 relevant variables.
2182
2183 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2184 with length and offset values. This means that the output is different
2185 for different internal link sizes. Test 2 is skipped for link sizes
2186 other than 2 because of this, bypassing the problem. Unfortunately,
2187 there was also a test in test 3 (the locale tests) that used /B and
2188 failed for link sizes other than 2. Rather than cut the whole test out,
2189 I have added a new /Z option to pcretest that replaces the length and
2190 offset values with spaces. This is now used to make test 3 independent
2191 of link size. (Test 2 will be tidied up later.)
2192
2193 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2194 segmentation fault instead of returning the appropriate error message.
2195
2196 15. In multiline mode when the newline sequence was set to "any", the pattern
2197 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2198 This doesn't seem right; it now treats the CRLF combination as the line
2199 ending, and so does not match in that case. It's only a pattern such as ^$
2200 that would hit this one: something like ^ABC$ would have failed after \r
2201 and then tried again after \r\n.
2202
2203 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2204 in an attempt to make files that differ only in their line terminators
2205 compare equal. This works on Linux.
2206
2207 17. Under certain error circumstances pcregrep might try to free random memory
2208 as it exited. This is now fixed, thanks to valgrind.
2209
2210 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2211 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2212 was because its rules for how to advance for /g after matching an empty
2213 string at the end of a line did not allow for this case. They now check for
2214 it specially.
2215
2216 20. pcretest is supposed to handle patterns and data of any length, by
2217 extending its buffers when necessary. It was getting this wrong when the
2218 buffer for a data line had to be extended.
2219
2220 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2221 CRLF as a newline sequence.
2222
2223 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2224 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2225 I have nevertheless tidied it up.
2226
2227 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2228
2229 24. Added a man page for pcre-config.
2230
2231
2232 Version 7.0 19-Dec-06
2233 ---------------------
2234
2235 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2236 moving to gcc 4.1.1.
2237
2238 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2239 sys/time.h, which is documented as needed for this function. It doesn't
2240 seem to matter on Linux, but it showed up on some releases of OS X.
2241
2242 3. It seems that there are systems where bytes whose values are greater than
2243 127 match isprint() in the "C" locale. The "C" locale should be the
2244 default when a C program starts up. In most systems, only ASCII printing
2245 characters match isprint(). This difference caused the output from pcretest
2246 to vary, making some of the tests fail. I have changed pcretest so that:
2247
2248 (a) When it is outputting text in the compiled version of a pattern, bytes
2249 other than 32-126 are always shown as hex escapes.
2250
2251 (b) When it is outputting text that is a matched part of a subject string,
2252 it does the same, unless a different locale has been set for the match
2253 (using the /L modifier). In this case, it uses isprint() to decide.
2254
2255 4. Fixed a major bug that caused incorrect computation of the amount of memory
2256 required for a compiled pattern when options that changed within the
2257 pattern affected the logic of the preliminary scan that determines the
2258 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2259 that the computed length was too small. The symptoms of this bug were
2260 either the PCRE error "internal error: code overflow" from pcre_compile(),
2261 or a glibc crash with a message such as "pcretest: free(): invalid next
2262 size (fast)". Examples of patterns that provoked this bug (shown in
2263 pcretest format) are:
2264
2265 /(?-x: )/x
2266 /(?x)(?-x: \s*#\s*)/
2267 /((?i)[\x{c0}])/8
2268 /(?i:[\x{c0}])/8
2269
2270 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2271 is now done differently.
2272
2273 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2274 wrapper classes; (b) implement a new function in the C++ scanner that is
2275 more efficient than the old way of doing things because it avoids levels of
2276 recursion in the regex matching; (c) add a paragraph to the documentation
2277 for the FullMatch() function.
2278
2279 6. The escape sequence \n was being treated as whatever was defined as
2280 "newline". Not only was this contrary to the documentation, which states
2281 that \n is character 10 (hex 0A), but it also went horribly wrong when
2282 "newline" was defined as CRLF. This has been fixed.
2283
2284 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2285 was being set to -1 for the "end of line" case (supposedly a value that no
2286 character can have). Though this value is never used (the check for end of
2287 line is "zero bytes in current character"), it caused compiler complaints.
2288 I've changed it to 0xffffffff.
2289
2290 8. In pcre_version.c, the version string was being built by a sequence of
2291 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2292 string (as it is for production releases) called a macro with an empty
2293 argument. The C standard says the result of this is undefined. The gcc
2294 compiler treats it as an empty string (which was what was wanted) but it is
2295 reported that Visual C gives an error. The source has been hacked around to
2296 avoid this problem.
2297
2298 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2299 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2300 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2301 of them did).
2302
2303 10. Originally, pcretest opened its input and output without "b"; then I was
2304 told that "b" was needed in some environments, so it was added for release
2305 5.0 to both the input and output. (It makes no difference on Unix-like
2306 systems.) Later I was told that it is wrong for the input on Windows. I've
2307 now abstracted the modes into two macros, to make it easier to fiddle with
2308 them, and removed "b" from the input mode under Windows.
2309
2310 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2311
2312 12. Added -help and --help to pcretest as an official way of being reminded
2313 of the options.
2314
2315 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2316 and pcrecpp.cc because they annoy compilers at high warning levels.
2317
2318 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2319
2320 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2321 scripts are not C programs :-) and which was not noticed because it works
2322 on Linux.
2323
2324 16. pcretest is supposed to handle any length of pattern and data line (as one
2325 line or as a continued sequence of lines) by extending its input buffer if
2326 necessary. This feature was broken for very long pattern lines, leading to
2327 a string of junk being passed to pcre_compile() if the pattern was longer
2328 than about 50K.
2329
2330 17. I have done a major re-factoring of the way pcre_compile() computes the
2331 amount of memory needed for a compiled pattern. Previously, there was code
2332 that made a preliminary scan of the pattern in order to do this. That was
2333 OK when PCRE was new, but as the facilities have expanded, it has become
2334 harder and harder to keep it in step with the real compile phase, and there
2335 have been a number of bugs (see for example, 4 above). I have now found a
2336 cunning way of running the real compile function in a "fake" mode that
2337 enables it to compute how much memory it would need, while actually only
2338 ever using a few hundred bytes of working memory and without too many
2339 tests of the mode. This should make future maintenance and development
2340 easier. A side effect of this work is that the limit of 200 on the nesting
2341 depth of parentheses has been removed (though this was never a serious
2342 limitation, I suspect). However, there is a downside: pcre_compile() now
2343 runs more slowly than before (30% or more, depending on the pattern). I
2344 hope this isn't a big issue. There is no effect on runtime performance.
2345
2346 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2347 newline (only possible for the last line of a file) and it was a
2348 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2349
2350 19. Added additional timing features to pcretest. (1) The -tm option now times
2351 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2352 separate command line item, by a number that specifies the number of
2353 repeats to use when timing. The default is 50000; this gives better
2354 precision, but takes uncomfortably long for very large patterns.
2355
2356 20. Extended pcre_study() to be more clever in cases where a branch of a
2357 subpattern has no definite first character. For example, (a*|b*)[cd] would
2358 previously give no result from pcre_study(). Now it recognizes that the
2359 first character must be a, b, c, or d.
2360
2361 21. There was an incorrect error "recursive call could loop indefinitely" if
2362 a subpattern (or the entire pattern) that was being tested for matching an
2363 empty string contained only one non-empty item after a nested subpattern.
2364 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2365 incorrectly, because the \d was being skipped in the check.
2366
2367 22. The pcretest program now has a new pattern option /B and a command line
2368 option -b, which is equivalent to adding /B to every pattern. This causes
2369 it to show the compiled bytecode, without the additional information that
2370 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2371 is the same as /B/I).
2372
2373 23. A new optimization is now able automatically to treat some sequences such
2374 as a*b as a*+b. More specifically, if something simple (such as a character
2375 or a simple class like \d) has an unlimited quantifier, and is followed by
2376 something that cannot possibly match the quantified thing, the quantifier
2377 is automatically "possessified".
2378
2379 24. A recursive reference to a subpattern whose number was greater than 39
2380 went wrong under certain circumstances in UTF-8 mode. This bug could also
2381 have affected the operation of pcre_study().
2382
2383 25. Realized that a little bit of performance could be had by replacing
2384 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2385
2386 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2387
2388 27. Possessive quantifiers such as a++ were previously implemented by turning
2389 them into atomic groups such as ($>a+). Now they have their own opcodes,
2390 which improves performance. This includes the automatically created ones
2391 from 23 above.
2392
2393 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2394 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2395 the first matched character to be a colon. This applied both to named and
2396 numbered groups.
2397
2398 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2399
2400 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2401 building PCRE on Windows easier, so I have included it in the distribution.
2402
2403 31. There is now a check in pcretest against a ridiculously large number being
2404 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2405 loop, the loop is abandoned.
2406
2407 32. Forward references to subpatterns in conditions such as (?(2)...) where
2408 subpattern 2 is defined later cause pcre_compile() to search forwards in
2409 the pattern for the relevant set of parentheses. This search went wrong
2410 when there were unescaped parentheses in a character class, parentheses
2411 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2412
2413 33. "Subroutine" calls and backreferences were previously restricted to
2414 referencing subpatterns earlier in the regex. This restriction has now
2415 been removed.
2416
2417 34. Added a number of extra features that are going to be in Perl 5.10. On the
2418 whole, these are just syntactic alternatives for features that PCRE had
2419 previously implemented using the Python syntax or my own invention. The
2420 other formats are all retained for compatibility.
2421
2422 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2423 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2424 also .NET compatible.
2425
2426 (b) A recursion or subroutine call to a named group can now be defined as
2427 (?&name) as well as (?P>name).
2428
2429 (c) A backreference to a named group can now be defined as \k<name> or
2430 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2431 5.10, are also .NET compatible.
2432
2433 (d) A conditional reference to a named group can now use the syntax
2434 (?(<name>) or (?('name') as well as (?(name).
2435
2436 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2437 groups (named and numbered) that are never evaluated inline, but can be
2438 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2439 is always false. There may be only one alternative in such a group.
2440
2441 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2442 as the simple (?(R). The condition is true only if the most recent
2443 recursion is that of the given number or name. It does not search out
2444 through the entire recursion stack.
2445
2446 (g) The escape \gN or \g{N} has been added, where N is a positive or
2447 negative number, specifying an absolute or relative reference.
2448
2449 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2450 some "unreachable code" warnings.
2451
2452 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2453 things, this adds five new scripts.
2454
2455 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2456 There were also incompatibilities regarding the handling of \Q..\E inside
2457 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2458 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2459
2460 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2461 matches an empty string, and forcibly breaks the loop. There were bugs in
2462 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2463 against aaaa the result was just "a" rather than "aaaa", for example. Two
2464 separate and independent bugs (that affected different cases) have been
2465 fixed.
2466
2467 39. Refactored the code to abolish the use of different opcodes for small
2468 capturing bracket numbers. This is a tidy that I avoided doing when I
2469 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2470 The new approach is not only tidier, it makes it possible to reduce the
2471 memory needed to fix the previous bug (38).
2472
2473 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2474 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2475 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2476 mode.
2477
2478 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2479 report.
2480
2481 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2482 copy construction and assignment in the C++ wrapper.
2483
2484 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2485 couple of bugs that could have given wrong results in the "--newline=crlf"
2486 case.
2487
2488 44. Added a number of casts and did some reorganization of signed/unsigned int
2489 variables following suggestions from Dair Grant. Also renamed the variable
2490 "this" as "item" because it is a C++ keyword.
2491
2492 45. Arranged for dftables to add
2493
2494 #include "pcre_internal.h"
2495
2496 to pcre_chartables.c because without it, gcc 4.x may remove the array
2497 definition from the final binary if PCRE is built into a static library and
2498 dead code stripping is activated.
2499
2500 46. For an unanchored pattern, if a match attempt fails at the start of a
2501 newline sequence, and the newline setting is CRLF or ANY, and the next two
2502 characters are CRLF, advance by two characters instead of one.
2503
2504
2505 Version 6.7 04-Jul-06
2506 ---------------------
2507
2508 1. In order to handle tests when input lines are enormously long, pcretest has
2509 been re-factored so that it automatically extends its buffers when
2510 necessary. The code is crude, but this _is_ just a test program. The
2511 default size has been increased from 32K to 50K.
2512
2513 2. The code in pcre_study() was using the value of the re argument before
2514 testing it for NULL. (Of course, in any sensible call of the function, it
2515 won't be NULL.)
2516
2517 3. The memmove() emulation function in pcre_internal.h, which is used on
2518 systems that lack both memmove() and bcopy() - that is, hardly ever -
2519 was missing a "static" storage class specifier.
2520
2521 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2522 containing an extended class (one that cannot be represented by a bitmap
2523 because it contains high-valued characters or Unicode property items, e.g.
2524 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2525 pattern, but PCRE should not loop if you do not (it no longer does).
2526 [Detail: two cases were found: (a) a repeated subpattern containing an
2527 extended class; (b) a recursive reference to a subpattern that followed a
2528 previous extended class. It wasn't skipping over the extended class
2529 correctly when UTF-8 mode was not set.]
2530
2531 5. A negated single-character class was not being recognized as fixed-length
2532 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2533 compile error "lookbehind assertion is not fixed length".
2534
2535 6. The RunPerlTest auxiliary script was showing an unexpected difference
2536 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2537 write a Perl script that can interpret lines of an input file either as
2538 byte characters or as UTF-8, which is what "perltest" was being required to
2539 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2540 can't do is switch easily at run time between having the "use utf8;" pragma
2541 or not. In the end, I fudged it by using the RunPerlTest script to insert
2542 "use utf8;" explicitly for the UTF-8 tests.
2543
2544 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2545 the end of the subject string, contrary to the documentation and to what
2546 Perl does. This was true of both matching functions. Now it matches only at
2547 the start of the subject and immediately after *internal* newlines.
2548
2549 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2550 a pointer to an int instead of a pointer to an unsigned long int. This
2551 caused problems on 64-bit systems.
2552
2553 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2554 instance of the 'standard' template library not being so standard".
2555
2556 10. There was no check on the number of named subpatterns nor the maximum
2557 length of a subpattern name. The product of these values is used to compute
2558 the size of the memory block for a compiled pattern. By supplying a very
2559 long subpattern name and a large number of named subpatterns, the size
2560 computation could be caused to overflow. This is now prevented by limiting
2561 the length of names to 32 characters, and the number of named subpatterns
2562 to 10,000.
2563
2564 11. Subpatterns that are repeated with specific counts have to be replicated in
2565 the compiled pattern. The size of memory for this was computed from the
2566 length of the subpattern and the repeat count. The latter is limited to
2567 65535, but there was no limit on the former, meaning that integer overflow
2568 could in principle occur. The compiled length of a repeated subpattern is
2569 now limited to 30,000 bytes in order to prevent this.
2570
2571 12. Added the optional facility to have named substrings with the same name.
2572
2573 13. Added the ability to use a named substring as a condition, using the
2574 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2575 are numbers (not recommended). Forward references are permitted.
2576
2577 14. Added forward references in named backreferences (if you see what I mean).
2578
2579 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2580 pattern could run off the end of the subject. For example, the pattern
2581 "(?s)(.{1,5})"8 did this with the subject "ab".
2582
2583 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2584 PCRE_CASELESS was set when matching characters that were quantified with ?
2585 or *.
2586
2587 17. A character class other than a single negated character that had a minimum
2588 but no maximum quantifier - for example [ab]{6,} - was not handled
2589 correctly by pce_dfa_exec(). It would match only one character.
2590
2591 18. A valid (though odd) pattern that looked like a POSIX character
2592 class but used an invalid character after [ (for example [[,abc,]]) caused
2593 pcre_compile() to give the error "Failed: internal error: code overflow" or
2594 in some cases to crash with a glibc free() error. This could even happen if
2595 the pattern terminated after [[ but there just happened to be a sequence of
2596 letters, a binary zero, and a closing ] in the memory that followed.
2597
2598 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2599 over the years. Originally (before any Unicode support), just the bottom 8
2600 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2601 output from "man perlunicode" includes this:
2602
2603 The regular expression compiler produces polymorphic opcodes. That
2604 is, the pattern adapts to the data and automatically switches to
2605 the Unicode character scheme when presented with Unicode data--or
2606 instead uses a traditional byte scheme when presented with byte
2607 data.
2608
2609 Sadly, a wide octal escape does not cause a switch, and in a string with
2610 no other multibyte characters, these octal escapes are treated as before.
2611 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2612 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2613 Unicode string.
2614
2615 I have not perpetrated such confusion in PCRE. Up till now, it took just
2616 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2617 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2618 translate to the appropriate multibyte character.
2619
2620 29. Applied some refactoring to reduce the number of warnings from Microsoft
2621 and Borland compilers. This has included removing the fudge introduced
2622 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2623 a warning about an unused variable.
2624
2625 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2626 characters since release 4.0, because Perl (from release 5.004) does not.
2627 [Or at least, is documented not to: some releases seem to be in conflict
2628 with the documentation.] However, when a pattern was studied with
2629 pcre_study() and all its branches started with \s, PCRE still included VT
2630 as a possible starting character. Of course, this did no harm; it just
2631 caused an unnecessary match attempt.
2632
2633 22. Removed a now-redundant internal flag bit that recorded the fact that case
2634 dependency changed within the pattern. This was once needed for "required
2635 byte" processing, but is no longer used. This recovers a now-scarce options
2636 bit. Also moved the least significant internal flag bit to the most-
2637 significant bit of the word, which was not previously used (hangover from
2638 the days when it was an int rather than a uint) to free up another bit for
2639 the future.
2640
2641 23. Added support for CRLF line endings as well as CR and LF. As well as the
2642 default being selectable at build time, it can now be changed at runtime
2643 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2644 specify that it is scanning data with non-default line endings.
2645
2646 24. Changed the definition of CXXLINK to make it agree with the definition of
2647 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2648
2649 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2650 recursions. This makes a big different to stack usage for some patterns.
2651
2652 26. If a subpattern containing a named recursion or subroutine reference such
2653 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2654 the space required for the compiled pattern went wrong and gave too small a
2655 value. Depending on the environment, this could lead to "Failed: internal
2656 error: code overflow at offset 49" or "glibc detected double free or
2657 corruption" errors.
2658
2659 27. Applied patches from Google (a) to support the new newline modes and (b) to
2660 advance over multibyte UTF-8 characters in GlobalReplace.
2661
2662 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2663 difference for some implementation of PCRE in some Windows version.
2664
2665 29. Added some extra testing facilities to pcretest:
2666
2667 \q<number> in a data line sets the "match limit" value
2668 \Q<number> in a data line sets the "match recursion limt" value
2669 -S <number> sets the stack size, where <number> is in megabytes
2670
2671 The -S option isn't available for Windows.
2672
2673
2674 Version 6.6 06-Feb-06
2675 ---------------------
2676
2677 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2678 in pcreposix.h. I have copied the definition from pcre.h.
2679
2680 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2681 because pcre.h is no longer a built file.
2682
2683 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2684 not normally included in the compiled code.
2685
2686
2687 Version 6.5 01-Feb-06
2688 ---------------------
2689
2690 1. When using the partial match feature with pcre_dfa_exec(), it was not
2691 anchoring the second and subsequent partial matches at the new starting
2692 point. This could lead to incorrect results. For example, with the pattern
2693 /1234/, partially matching against "123" and then "a4" gave a match.
2694
2695 2. Changes to pcregrep:
2696
2697 (a) All non-match returns from pcre_exec() were being treated as failures
2698 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2699 error message is output. Some extra information is given for the
2700 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2701 probably the only errors that are likely to be caused by users (by
2702 specifying a regex that has nested indefinite repeats, for instance).
2703 If there are more than 20 of these errors, pcregrep is abandoned.
2704
2705 (b) A binary zero was treated as data while matching, but terminated the
2706 output line if it was written out. This has been fixed: binary zeroes
2707 are now no different to any other data bytes.
2708
2709 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2710 used to set a locale for matching. The --locale=xxxx long option has
2711 been added (no short equivalent) to specify a locale explicitly on the
2712 pcregrep command, overriding the environment variables.
2713
2714 (d) When -B was used with -n, some line numbers in the output were one less
2715 than they should have been.
2716
2717 (e) Added the -o (--only-matching) option.
2718
2719 (f) If -A or -C was used with -c (count only), some lines of context were
2720 accidentally printed for the final match.
2721
2722 (g) Added the -H (--with-filename) option.
2723
2724 (h) The combination of options -rh failed to suppress file names for files
2725 that were found from directory arguments.
2726
2727 (i) Added the -D (--devices) and -d (--directories) options.
2728
2729 (j) Added the -F (--fixed-strings) option.
2730
2731 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2732
2733 (l) Added the --colo(u)r option.
2734
2735 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2736 is not present by default.
2737
2738 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2739 items such as (?R) or (?1), when the recursion could match a number of
2740 alternatives. If it matched one of the alternatives, but subsequently,
2741 outside the recursion, there was a failure, the code tried to back up into
2742 the recursion. However, because of the way PCRE is implemented, this is not
2743 possible, and the result was an incorrect result from the match.
2744
2745 In order to prevent this happening, the specification of recursion has
2746 been changed so that all such subpatterns are automatically treated as
2747 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2748
2749 4. I had overlooked the fact that, in some locales, there are characters for
2750 which isalpha() is true but neither isupper() nor islower() are true. In
2751 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2752 and ordfeminine) are like this. This affected the treatment of \w and \W
2753 when they appeared in character classes, but not when they appeared outside
2754 a character class. The bit map for "word" characters is now created
2755 separately from the results of isalnum() instead of just taking it from the
2756 upper, lower, and digit maps. (Plus the underscore character, of course.)
2757
2758 5. The above bug also affected the handling of POSIX character classes such as
2759 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2760 permanent tables. Instead, the bit maps for such a class were previously
2761 created as the appropriate unions of the upper, lower, and digit bitmaps.
2762 Now they are created by subtraction from the [[:word:]] class, which has
2763 its own bitmap.
2764
2765 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2766 It is created by subtracting the vertical space characters (\x09, \x0a,
2767 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2768 subtraction was done in the overall bitmap for a character class, meaning
2769 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2770 be recognized. This bug has been fixed.
2771
2772 7. Patches from the folks at Google:
2773
2774 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2775 real life, but is still worth protecting against".
2776
2777 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2778 regular expressions".
2779
2780 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2781 have it.
2782
2783 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2784 "configure" and the latter not, in order to fix a problem somebody had
2785 with compiling the Arg class on HP-UX.
2786
2787 (e) Improve the error-handling of the C++ wrapper a little bit.
2788
2789 (f) New tests for checking recursion limiting.
2790
2791 8. The pcre_memmove() function, which is used only if the environment does not
2792 have a standard memmove() function (and is therefore rarely compiled),
2793 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2794 returning a result (though PCRE never actually uses the result).
2795
2796 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2797 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2798 returned instead of calling malloc() with an overflowing number that would
2799 most likely cause subsequent chaos.
2800
2801 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2802
2803 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2804 with this option is matched, the nmatch and pmatch options of regexec() are
2805 ignored.
2806
2807 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2808 provided in case anyone wants to the the POSIX interface with UTF-8
2809 strings.
2810
2811 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2812 C++ linking (needed for some HP-UX environments).
2813
2814 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2815 (unused parameter) and in the pcre_printint() function (omitted "default"
2816 switch label when the default is to do nothing).
2817
2818 15. Added some code to make it possible, when PCRE is compiled as a C++
2819 library, to replace subject pointers for pcre_exec() with a smart pointer
2820 class, thus making it possible to process discontinuous strings.
2821
2822 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2823 much the same function. They were added by different people who were trying
2824 to make PCRE easy to compile on non-Unix systems. It has been suggested
2825 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2826 for compiling on Windows systems. I have therefore replaced it with
2827 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2828 defaults to "extern" for C or "extern C" for C++, which works fine on
2829 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2830 SCOPE with something explicit in config.h. In addition:
2831
2832 (a) pcreposix.h still had just "extern" instead of either of these macros;
2833 I have replaced it with PCRE_DATA_SCOPE.
2834
2835 (b) Functions such as _pcre_xclass(), which are internal to the library,
2836 but external in the C sense, all had PCRE_EXPORT in their definitions.
2837 This is apparently wrong for the Windows case, so I have removed it.
2838 (It makes no difference on Unix-like systems.)
2839
2840 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2841 of recursive calls to match(). This is different to MATCH_LIMIT because
2842 that limits the total number of calls to match(), not all of which increase
2843 the depth of recursion. Limiting the recursion depth limits the amount of
2844 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2845 when PCRE is compiled, and changed at run time. A patch from Google adds
2846 this functionality to the C++ interface.
2847
2848 18. Changes to the handling of Unicode character properties:
2849
2850 (a) Updated the table to Unicode 4.1.0.
2851
2852 (b) Recognize characters that are not in the table as "Cn" (undefined).
2853
2854 (c) I revised the way the table is implemented to a much improved format
2855 which includes recognition of ranges. It now supports the ranges that
2856 are defined in UnicodeData.txt, and it also amalgamates other
2857 characters into ranges. This has reduced the number of entries in the
2858 table from around 16,000 to around 3,000, thus reducing its size
2859 considerably. I realized I did not need to use a tree structure after
2860 all - a binary chop search is just as efficient. Having reduced the
2861 number of entries, I extended their size from 6 bytes to 8 bytes to
2862 allow for more data.
2863
2864 (d) Added support for Unicode script names via properties such as \p{Han}.
2865
2866 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2867 matching that character.
2868
2869 20. When matching a repeated Unicode property with a minimum greater than zero,
2870 (for example \pL{2,}), PCRE could look past the end of the subject if it
2871 reached it while seeking the minimum number of characters. This could
2872 happen only if some of the characters were more than one byte long, because
2873 there is a check for at least the minimum number of bytes.
2874
2875 21. Refactored the implementation of \p and \P so as to be more general, to
2876 allow for more different types of property in future. This has changed the
2877 compiled form incompatibly. Anybody with saved compiled patterns that use
2878 \p or \P will have to recompile them.
2879
2880 22. Added "Any" and "L&" to the supported property types.
2881
2882 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2883 but give a compile time error if the value is greater than 0xff.
2884
2885 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2886 accidentally not being installed or uninstalled.
2887
2888 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2889 made were to insert the current release number. This seemed silly, because
2890 it made things harder for people building PCRE on systems that don't run
2891 "configure". I have turned pcre.h into a distributed file, no longer built
2892 by "configure", with the version identification directly included. There is
2893 no longer a pcre.h.in file.
2894
2895 However, this change necessitated a change to the pcre-config script as
2896 well. It is built from pcre-config.in, and one of the substitutions was the
2897 release number. I have updated configure.ac so that ./configure now finds
2898 the release number by grepping pcre.h.
2899
2900 26. Added the ability to run the tests under valgrind.
2901
2902
2903 Version 6.4 05-Sep-05
2904 ---------------------
2905
2906 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2907 "--" to be printed when multiple files were scanned, even when none of the
2908 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2909 consider it to be a bug, and have restored the previous behaviour.
2910
2911 2. A couple of code tidies to get rid of compiler warnings.
2912
2913 3. The pcretest program used to cheat by referring to symbols in the library
2914 whose names begin with _pcre_. These are internal symbols that are not
2915 really supposed to be visible externally, and in some environments it is
2916 possible to suppress them. The cheating is now confined to including
2917 certain files from the library's source, which is a bit cleaner.
2918
2919 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2920 file's purpose clearer.
2921
2922 5. Reorganized pcre_ucp_findchar().
2923
2924
2925 Version 6.3 15-Aug-05
2926 ---------------------
2927
2928 1. The file libpcre.pc.in did not have general read permission in the tarball.
2929
2930 2. There were some problems when building without C++ support:
2931
2932 (a) If C++ support was not built, "make install" and "make test" still
2933 tried to test it.
2934
2935 (b) There were problems when the value of CXX was explicitly set. Some
2936 changes have been made to try to fix these, and ...
2937
2938 (c) --disable-cpp can now be used to explicitly disable C++ support.
2939
2940 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2941 backslash in a target when C++ was disabled. This confuses some
2942 versions of "make", apparently. Using an intermediate variable solves
2943 this. (Same for CPP_LOBJ.)
2944
2945 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2946 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2947 necessary on certain architectures.
2948
2949 4. Added a setting of -export-symbols-regex to the link command to remove
2950 those symbols that are exported in the C sense, but actually are local
2951 within the library, and not documented. Their names all begin with
2952 "_pcre_". This is not a perfect job, because (a) we have to except some
2953 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2954 available (and never for static libraries). I have made a note to try to
2955 find a way round (a) in the future.
2956
2957
2958 Version 6.2 01-Aug-05
2959 ---------------------
2960
2961 1. There was no test for integer overflow of quantifier values. A construction
2962 such as {1111111111111111} would give undefined results. What is worse, if
2963 a minimum quantifier for a parenthesized subpattern overflowed and became
2964 negative, the calculation of the memory size went wrong. This could have
2965 led to memory overwriting.
2966
2967 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2968
2969 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2970 operating environments where this matters.
2971
2972 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2973 PCRE options from within the C++ wrapper.
2974
2975 5. Named capturing subpatterns were not being correctly counted when a pattern
2976 was compiled. This caused two problems: (a) If there were more than 100
2977 such subpatterns, the calculation of the memory needed for the whole
2978 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2979 back references of the form \12, where the number was greater than 9, were
2980 not recognized as back references, even though there were sufficient
2981 previous subpatterns.
2982
2983 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2984 versions of gcc, e.g. 2.95.4.
2985
2986
2987 Version 6.1 21-Jun-05
2988 ---------------------
2989
2990 1. There was one reference to the variable "posix" in pcretest.c that was not
2991 surrounded by "#if !defined NOPOSIX".
2992
2993 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2994 the cross-check on the old pcre_info() function, for the benefit of the
2995 cut-down version of PCRE that is currently imported into Exim.
2996
2997 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2998 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2999 patterns that start (?i) but I don't think that matters. The use of (?i) is
3000 just an example; this all applies to the other options as well.
3001
3002 4. Since libtool seems to echo the compile commands it is issuing, the output
3003 from "make" can be reduced a bit by putting "@" in front of each libtool
3004 compile command.
3005
3006 5. Patch from the folks at Google for configure.in to be a bit more thorough
3007 in checking for a suitable C++ installation before trying to compile the
3008 C++ stuff. This should fix a reported problem when a compiler was present,
3009 but no suitable headers.
3010
3011 6. The man pages all had just "PCRE" as their title. I have changed them to
3012 be the relevant file name. I have also arranged that these names are
3013 retained in the file doc/pcre.txt, which is a concatenation in text format
3014 of all the man pages except the little individual ones for each function.
3015
3016 7. The NON-UNIX-USE file had not been updated for the different set of source
3017 files that come with release 6. I also added a few comments about the C++
3018 wrapper.
3019
3020
3021 Version 6.0 07-Jun-05
3022 ---------------------
3023
3024 1. Some minor internal re-organization to help with my DFA experiments.
3025
3026 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3027 didn't matter for the library itself when fully configured, but did matter
3028 when compiling without UCP support, or within Exim, where the ucp files are
3029 not imported.
3030
3031 3. Refactoring of the library code to split up the various functions into
3032 different source modules. The addition of the new DFA matching code (see
3033 below) to a single monolithic source would have made it really too
3034 unwieldy, quite apart from causing all the code to be include in a
3035 statically linked application, when only some functions are used. This is
3036 relevant even without the DFA addition now that patterns can be compiled in
3037 one application and matched in another.
3038
3039 The downside of splitting up is that there have to be some external
3040 functions and data tables that are used internally in different modules of
3041 the library but which are not part of the API. These have all had their
3042 names changed to start with "_pcre_" so that they are unlikely to clash
3043 with other external names.
3044
3045 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3046 a different (DFA) algorithm. Although it is slower than the original
3047 function, it does have some advantages for certain types of matching
3048 problem.
3049
3050 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3051 including restarting after a partial match.
3052
3053 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3054 defined when compiling for Windows was sent to me. I have put it into the
3055 code, though I have no means of testing or verifying it.
3056
3057 7. Added the pcre_refcount() auxiliary function.
3058
3059 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3060 match before or at the first newline in the subject string. In pcretest,
3061 the /f option on a pattern can be used to set this.
3062
3063 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3064 would behave wrongly. This has been present in PCRE since release 4.0.
3065
3066 10. A number of changes to the pcregrep command:
3067
3068 (a) Refactored how -x works; insert ^(...)$ instead of setting
3069 PCRE_ANCHORED and checking the length, in preparation for adding
3070 something similar for -w.
3071
3072 (b) Added the -w (match as a word) option.
3073
3074 (c) Refactored the way lines are read and buffered so as to have more
3075 than one at a time available.
3076
3077 (d) Implemented a pcregrep test script.
3078
3079 (e) Added the -M (multiline match) option. This allows patterns to match
3080 over several lines of the subject. The buffering ensures that at least
3081 8K, or the rest of the document (whichever is the shorter) is available
3082 for matching (and similarly the previous 8K for lookbehind assertions).
3083
3084 (f) Changed the --help output so that it now says
3085
3086 -w, --word-regex(p)
3087
3088 instead of two lines, one with "regex" and the other with "regexp"
3089 because that confused at least one person since the short forms are the
3090 same. (This required a bit of code, as the output is generated
3091 automatically from a table. It wasn't just a text change.)
3092
3093 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3094 option but starts with a hyphen. Could be a pattern or a path name
3095 starting with a hyphen, for instance.
3096
3097 (h) "-" can be given as a file name to represent stdin.
3098
3099 (i) When file names are being printed, "(standard input)" is used for
3100 the standard input, for compatibility with GNU grep. Previously
3101 "<stdin>" was used.
3102
3103 (j) The option --label=xxx can be used to supply a name to be used for
3104 stdin when file names are being printed. There is no short form.
3105
3106 (k) Re-factored the options decoding logic because we are going to add
3107 two more options that take data. Such options can now be given in four
3108 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3109
3110 (l) Added the -A, -B, and -C options for requesting that lines of context
3111 around matches be printed.
3112
3113 (m) Added the -L option to print the names of files that do not contain
3114 any matching lines, that is, the complement of -l.
3115
3116 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3117 continue to scan other files.
3118
3119 (o) The -s option was incorrectly implemented. For compatibility with other
3120 greps, it now suppresses the error message for a non-existent or non-
3121 accessible file (but not the return code). There is a new option called
3122 -q that suppresses the output of matching lines, which was what -s was
3123 previously doing.
3124
3125 (p) Added --include and --exclude options to specify files for inclusion
3126 and exclusion when recursing.
3127
3128 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3129 Hopefully, it now does.
3130
3131 12. Missing cast in pcre_study().
3132
3133 13. Added an "uninstall" target to the makefile.
3134
3135 14. Replaced "extern" in the function prototypes in Makefile.in with
3136 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3137 world, but is set differently for Windows.
3138
3139 15. Added a second compiling function called pcre_compile2(). The only
3140 difference is that it has an extra argument, which is a pointer to an
3141 integer error code. When there is a compile-time failure, this is set
3142 non-zero, in addition to the error test pointer being set to point to an
3143 error message. The new argument may be NULL if no error number is required
3144 (but then you may as well call pcre_compile(), which is now just a
3145 wrapper). This facility is provided because some applications need a
3146 numeric error indication, but it has also enabled me to tidy up the way
3147 compile-time errors are handled in the POSIX wrapper.
3148
3149 16. Added VPATH=.libs to the makefile; this should help when building with one
3150 prefix path and installing with another. (Or so I'm told by someone who
3151 knows more about this stuff than I do.)
3152
3153 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3154 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3155 match everything, including newlines. This is not POSIX-compatible, but
3156 somebody wanted the feature. From pcretest it can be activated by using
3157 both the P and the s flags.
3158
3159 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3160
3161 19. libpcre.pc was being incorrectly installed as executable.
3162
3163 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3164 it now also looks for '\r' so that it will work unmodified on Windows.
3165
3166 21. Added Google's contributed C++ wrapper to the distribution.
3167
3168 22. Added some untidy missing memory free() calls in pcretest, to keep
3169 Electric Fence happy when testing.
3170
3171
3172
3173 Version 5.0 13-Sep-04
3174 ---------------------
3175
3176 1. Internal change: literal characters are no longer packed up into items
3177 containing multiple characters in a single byte-string. Each character
3178 is now matched using a separate opcode. However, there may be more than one
3179 byte in the character in UTF-8 mode.
3180
3181 2. The pcre_callout_block structure has two new fields: pattern_position and
3182 next_item_length. These contain the offset in the pattern to the next match
3183 item, and its length, respectively.
3184
3185 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3186 insertion of callouts before each pattern item. Added the /C option to
3187 pcretest to make use of this.
3188
3189 4. On the advice of a Windows user, the lines
3190
3191 #if defined(_WIN32) || defined(WIN32)
3192 _setmode( _fileno( stdout ), 0x8000 );
3193 #endif /* defined(_WIN32) || defined(WIN32) */
3194
3195 have been added to the source of pcretest. This apparently does useful
3196 magic in relation to line terminators.
3197
3198 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3199 for the benefit of those environments where the "b" makes a difference.
3200
3201 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3202 to know about it. I have put a hack into configure.in that adds in code
3203 to set GCC=yes if CC=icc. This seems to end up at a point in the
3204 generated configure script that is early enough to affect the setting of
3205 compiler options, which is what is needed, but I have no means of testing
3206 whether it really works. (The user who reported this had patched the
3207 generated configure script, which of course I cannot do.)
3208
3209 LATER: After change 22 below (new libtool files), the configure script
3210 seems to know about icc (and also ecc). Therefore, I have commented out
3211 this hack in configure.in.
3212
3213 7. Added support for pkg-config (2 patches were sent in).
3214
3215 8. Negated POSIX character classes that used a combination of internal tables
3216 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3217 [[:^ascii]]. Typically, they would match almost any characters. The other
3218 POSIX classes were not broken in this way.
3219
3220 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3221 to find the match, as PCRE was deluded into thinking that the match had to
3222 start at the start point or following a newline. The same bug applied to
3223 patterns with negative forward assertions or any backward assertions
3224 preceding ".*" at the start, unless the pattern required a fixed first
3225 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3226
3227 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3228 starting at the last subject character, bytes beyond the end of the subject
3229 string were read.
3230
3231 11. Renamed the variable "class" as "classbits" to make life easier for C++
3232 users. (Previously there was a macro definition, but it apparently wasn't
3233 enough.)
3234
3235 12. Added the new field "tables" to the extra data so that tables can be passed
3236 in at exec time, or the internal tables can be re-selected. This allows
3237 a compiled regex to be saved and re-used at a later time by a different
3238 program that might have everything at different addresses.
3239
3240 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3241 -R library as well as a -L library.
3242
3243 14. The debugging options of pcretest (-d on the command line or D on a
3244 pattern) showed incorrect output for anything following an extended class
3245 that contained multibyte characters and which was followed by a quantifier.
3246
3247 15. Added optional support for general category Unicode character properties
3248 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3249 support. It adds about 90K to the size of the library. The meanings of the
3250 inbuilt class escapes such as \d and \s have NOT been changed.
3251
3252 16. Updated pcredemo.c to include calls to free() to release the memory for the
3253 compiled pattern.
3254
3255 17. The generated file chartables.c was being created in the source directory
3256 instead of in the building directory. This caused the build to fail if the
3257 source directory was different from the building directory, and was
3258 read-only.
3259
3260 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3261 file. No doubt somebody will tell me if they don't make sense... Also added
3262 Dan Mooney's comments about building on OpenVMS.
3263
3264 19. Added support for partial matching via the PCRE_PARTIAL option for
3265 pcre_exec() and the \P data escape in pcretest.
3266
3267 20. Extended pcretest with 3 new pattern features:
3268
3269 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3270 write the compiled pattern to the file whose name is "rest-of-line".
3271 This is a straight binary dump of the data, with the saved pointer to
3272 the character tables forced to be NULL. The study data, if any, is
3273 written too. After writing, pcretest reads a new pattern.
3274
3275 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3276 compiled pattern from the given file. There must not be any
3277 occurrences of "<" in the file name (pretty unlikely); if there are,
3278 pcretest will instead treat the initial "<" as a pattern delimiter.
3279 After reading in the pattern, pcretest goes on to read data lines as
3280 usual.
3281
3282 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3283 and 16-bit fields in a compiled pattern, to simulate a pattern that
3284 was compiled on a host of opposite endianness.
3285
3286 21. The pcre-exec() function can now cope with patterns that were compiled on
3287 hosts of opposite endianness, with this restriction:
3288
3289 As for any compiled expression that is saved and used later, the tables
3290 pointer field cannot be preserved; the extra_data field in the arguments
3291 to pcre_exec() should be used to pass in a tables address if a value
3292 other than the default internal tables were used at compile time.
3293
3294 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3295 now diagnosed as an error. Previously, most of the time, a negative number
3296 would have been treated as zero, but if in addition "ovector" was passed as
3297 NULL, a crash could occur.
3298
3299 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3300 new versions from the libtool 1.5 distribution (the last one is a copy of
3301 a file called libtool.m4). This seems to have fixed the need to patch
3302 "configure" to support Darwin 1.3 (which I used to do). However, I still
3303 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3304 workstation).
3305
3306 24. Changed the PCRE licence to be the more standard "BSD" licence.
3307
3308
3309 Version 4.5 01-Dec-03
3310 ---------------------
3311
3312 1. There has been some re-arrangement of the code for the match() function so
3313 that it can be compiled in a version that does not call itself recursively.
3314 Instead, it keeps those local variables that need separate instances for
3315 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3316 needs to "recurse". Keeping track of where control must go is done by means
3317 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3318 hide most of the details from the main code, and operates only if
3319 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3320 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3321 operating.
3322
3323 To make it easier for callers to provide specially tailored get/free
3324 functions for this usage, two new functions, pcre_stack_malloc, and
3325 pcre_stack_free, are used. They are always called in strict stacking order,
3326 and the size of block requested is always the same.
3327
3328 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3329 PCRE has been compiled to use the stack or the heap for recursion. The
3330 -C option of pcretest uses this to show which version is compiled.
3331
3332 A new data escape \S, is added to pcretest; it causes the amounts of store
3333 obtained and freed by both kinds of malloc/free at match time to be added
3334 to the output.
3335
3336 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3337 what's available on my current Linux desktop machine.
3338
3339 3. When matching a UTF-8 string, the test for a valid string at the start has
3340 been extended. If start_offset is not zero, PCRE now checks that it points
3341 to a byte that is the start of a UTF-8 character. If not, it returns
3342 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3343 this is necessary because there may be backward assertions in the pattern.
3344 When matching the same subject several times, it may save resources to use
3345 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3346
3347 4. The code for checking the validity of UTF-8 strings has been tightened so
3348 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3349 containing "overlong sequences".
3350
3351 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3352 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3353 should have been "&", but it just so happened that all the cases this let
3354 through by mistake were picked up later in the function.
3355
3356 6. I had used a variable called "isblank" - this is a C99 function, causing
3357 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3358
3359 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3360 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3361 so the version is shown; (c) stop "make test" echoing "./RunTest".
3362
3363 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3364
3365 9. The prototype for memmove() for systems that don't have it was using
3366 size_t, but the inclusion of the header that defines size_t was later. I've
3367 moved the #includes for the C headers earlier to avoid this.
3368
3369 10. Added some adjustments to the code to make it easier to compiler on certain
3370 special systems:
3371
3372 (a) Some "const" qualifiers were missing.
3373 (b) Added the macro EXPORT before all exported functions; by default this
3374 is defined to be empty.
3375 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3376 that it reads its output file name as an argument instead of writing
3377 to the standard output and assuming this can be redirected.
3378
3379 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3380 class containing characters with values greater than 255, PCRE compilation
3381 went into a loop.
3382
3383 12. A recursive reference to a subpattern that was within another subpattern
3384 that had a minimum quantifier of zero caused PCRE to crash. For example,
3385 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3386 recursion. If the recursively-called subpattern itself had a zero repeat,
3387 that was OK.
3388
3389 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3390 buffer into which it was copied (for escape processing) was still set at
3391 1024, so long lines caused crashes.
3392
3393 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3394 "internal error: code overflow...". This applied to any character class
3395 that was followed by a possessive quantifier.
3396
3397 15. Modified the Makefile to add libpcre.la as a prerequisite for
3398 libpcreposix.la because I was told this is needed for a parallel build to
3399 work.
3400
3401 16. If a pattern that contained .* following optional items at the start was
3402 studied, the wrong optimizing data was generated, leading to matching
3403 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3404 matching string must start with a or b or c. The correct conclusion for
3405 this pattern is that a match can start with any character.
3406
3407
3408 Version 4.4 13-Aug-03
3409 ---------------------
3410
3411 1. In UTF-8 mode, a character class containing characters with values between
3412 127 and 255 was not handled correctly if the compiled pattern was studied.
3413 In fixing this, I have also improved the studying algorithm for such
3414 classes (slightly).
3415
3416 2. Three internal functions had redundant arguments passed to them. Removal
3417 might give a very teeny performance improvement.
3418
3419 3. Documentation bug: the value of the capture_top field in a callout is *one
3420 more than* the number of the hightest numbered captured substring.
3421
3422 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3423 in incorrectly linking with a previously installed version. They now link
3424 explicitly with libpcre.la.
3425
3426 5. configure.in no longer needs to recognize Cygwin specially.
3427
3428 6. A problem in pcre.in for Windows platforms is fixed.
3429
3430 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3431 pcretest, it used to include the size of the study block as part of its
3432 output. Unfortunately, the structure contains a field that has a different
3433 size on different hardware architectures. This meant that the tests that
3434 showed this size failed. As the block is currently always of a fixed size,
3435 this information isn't actually particularly useful in pcretest output, so
3436 I have just removed it.
3437
3438 8. Three pre-processor statements accidentally did not start in column 1.
3439 Sadly, there are *still* compilers around that complain, even though
3440 standard C has not required this for well over a decade. Sigh.
3441
3442 9. In pcretest, the code for checking callouts passed small integers in the
3443 callout_data field, which is a void * field. However, some picky compilers
3444 complained about the casts involved for this on 64-bit systems. Now
3445 pcretest passes the address of the small integer instead, which should get
3446 rid of the warnings.
3447
3448 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3449 both compile and run time, and gives an error if an invalid UTF-8 sequence
3450 is found. There is a option for disabling this check in cases where the
3451 string is known to be correct and/or the maximum performance is wanted.
3452
3453 11. In response to a bug report, I changed one line in Makefile.in from
3454
3455 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3456 to
3457 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3458
3459 to look similar to other lines, but I have no way of telling whether this
3460 is the right thing to do, as I do not use Windows. No doubt I'll get told
3461 if it's wrong...
3462
3463
3464 Version 4.3 21-May-03
3465 ---------------------
3466
3467 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3468 Makefile.
3469
3470 2. Some refactoring to improve the quality of the code:
3471
3472 (i) The utf8_table... variables are now declared "const".
3473
3474 (ii) The code for \cx, which used the "case flipping" table to upper case
3475 lower case letters, now just substracts 32. This is ASCII-specific,
3476 but the whole concept of \cx is ASCII-specific, so it seems
3477 reasonable.
3478
3479 (iii) PCRE was using its character types table to recognize decimal and
3480 hexadecimal digits in the pattern. This is silly, because it handles
3481 only 0-9, a-f, and A-F, but the character types table is locale-
3482 specific, which means strange things might happen. A private
3483 table is now used for this - though it costs 256 bytes, a table is
3484 much faster than multiple explicit tests. Of course, the standard
3485 character types table is still used for matching digits in subject
3486 strings against \d.
3487
3488 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3489 ending in _t are). So I've renamed it as ESC_tee.
3490
3491 3. The first argument for regexec() in the POSIX wrapper should have been
3492 defined as "const".
3493
3494 4. Changed pcretest to use malloc() for its buffers so that they can be
3495 Electric Fenced for debugging.
3496
3497 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3498 to read one or more bytes before the start of the subject string. Often this
3499 had no effect on PCRE's behaviour, but in some circumstances it could
3500 provoke a segmentation fault.
3501
3502 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3503 to try to read one or more bytes before the start of the subject string.
3504
3505 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3506 UTF-8 support could misbehave in various ways if the subject string
3507 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3508 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3509 back over UTF-8 characters.)
3510
3511
3512 Version 4.2 14-Apr-03
3513 ---------------------
3514
3515 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3516
3517 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3518 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3519 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3520 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3521 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3522 and BUILD_EXEEXT
3523 Note: automatic setting of the BUILD variables is not yet working
3524 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3525 compile-time but not at link-time
3526 [LINK]: use for linking executables only
3527 make different versions for Windows and non-Windows
3528 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3529 libraries
3530 [LINK_FOR_BUILD]: new variable
3531 [OBJEXT]: use throughout
3532 [EXEEXT]: use throughout
3533 <winshared>: new target
3534 <wininstall>: new target
3535 <dftables.o>: use native compiler
3536 <dftables>: use native linker
3537 <install>: handle Windows platform correctly
3538 <clean>: ditto
3539 <check>: ditto
3540 copy DLL to top builddir before testing
3541
3542 As part of these changes, -no-undefined was removed again. This was reported
3543 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3544 in any case.
3545
3546 3. Some tidies to get rid of compiler warnings:
3547
3548 . In the match_data structure, match_limit was an unsigned long int, whereas
3549 match_call_count was an int. I've made them both unsigned long ints.
3550
3551 . In pcretest the fact that a const uschar * doesn't automatically cast to
3552 a void * provoked a warning.
3553
3554 . Turning on some more compiler warnings threw up some "shadow" variables
3555 and a few more missing casts.
3556
3557 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3558 option, a class that contained a single character with a value between 128
3559 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3560
3561 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3562 option, a class that contained several characters, but with at least one
3563 whose value was between 128 and 255 caused PCRE to crash.
3564
3565
3566 Version 4.1 12-Mar-03
3567 ---------------------
3568
3569 1. Compiling with gcc -pedantic found a couple of places where casts were
3570 needed, and a string in dftables.c that was longer than standard compilers are
3571 required to support.
3572
3573 2. Compiling with Sun's compiler found a few more places where the code could
3574 be tidied up in order to avoid warnings.
3575
3576 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3577 first of these names is deprecated in the latest Autoconf in favour of the name
3578 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3579 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3580 analogy I have changed it to CFLAGS_FOR_BUILD.
3581
3582 4. Added -no-undefined to the linking command in the Makefile, because this is
3583 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3584 linking step for the pcreposix library.
3585
3586 5. PCRE was failing to diagnose the case of two named groups with the same
3587 name.
3588
3589 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3590 literal character that is needed in the subject for a match, and scans along to
3591 ensure that it is present before embarking on the full matching process. This
3592 saves time in cases of nested unlimited repeats that are never going to match.
3593 Problem: the scan can take a lot of time if the subject is very long (e.g.
3594 megabytes), thus penalizing straightforward matches. It is now done only if the
3595 amount of subject to be scanned is less than 1000 bytes.
3596
3597 7. A lesser problem with the same optimization is that it was recording the
3598 first character of an anchored pattern as "needed", thus provoking a search
3599 right along the subject, even when the first match of the pattern was going to
3600 fail. The "needed" character is now not set for anchored patterns, unless it
3601 follows something in the pattern that is of non-fixed length. Thus, it still
3602 fulfils its original purpose of finding quick non-matches in cases of nested
3603 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3604
3605
3606 Version 4.0 17-Feb-03
3607 ---------------------
3608
3609 1. If a comment in an extended regex that started immediately after a meta-item
3610 extended to the end of string, PCRE compiled incorrect data. This could lead to
3611 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3612
3613 2. Moved to autoconf 2.53 and libtool 1.4.2.
3614
3615 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3616 the special perltest8 script is no longer needed - all the tests can be run
3617 from a single perltest script.
3618
3619 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3620 by \s. It has now been removed in PCRE. This means it isn't recognized as
3621 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3622 class [:space:] *does* include VT, thereby creating a mess.
3623
3624 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3625 space and tab.
3626
3627 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3628 its new features into the main test script, reducing the number of scripts.
3629
3630 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3631 were backward compatible, and made the (?i) apply to the whole pattern, as if
3632 /i were given. Now it behaves more logically, and applies the option setting
3633 only to what follows. PCRE has been changed to follow suit. However, if it
3634 finds options settings right at the start of the pattern, it extracts them into
3635 the global options, as before. Thus, they show up in the info data.
3636
3637 8. Added support for the \Q...\E escape sequence. Characters in between are
3638 treated as literals. This is slightly different from Perl in that $ and @ are
3639 also handled as literals inside the quotes. In Perl, they will cause variable
3640 interpolation. Note the following examples:
3641
3642 Pattern PCRE matches Perl matches
3643
3644 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3645 \Qabc\$xyz\E abc\$xyz abc\$xyz
3646 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3647
3648 For compatibility with Perl, \Q...\E sequences are recognized inside character
3649 classes as well as outside them.
3650
3651 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3652 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3653 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3654 signed/unsigned warnings.
3655
3656 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3657 option for pcretest, so I've replaced it by a simple function that does just
3658 that job.
3659
3660 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3661 "pcregrep -".
3662
3663 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3664 Java package. This provides some syntactic sugar for simple cases of what my
3665 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3666 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3667 item, you can use this simplified notation. Note that only makes sense with
3668 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3669 greediness, whatever the setting of the PCRE_UNGREEDY option.
3670
3671 13. A change of greediness default within a pattern was not taking effect at
3672 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3673 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3674 was abstracted outside.
3675
3676 14. PCRE now supports the \G assertion. It is true when the current matching
3677 position is at the start point of the match. This differs from \A when the
3678 starting offset is non-zero. Used with the /g option of pcretest (or similar
3679 code), it works in the same way as it does for Perl's /g option. If all
3680 alternatives of a regex begin with \G, the expression is anchored to the start
3681 match position, and the "anchored" flag is set in the compiled expression.
3682
3683 15. Some bugs concerning the handling of certain option changes within patterns
3684 have been fixed. These applied to options other than (?ims). For example,
3685 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3686 been the other way round. Some of this was related to change 7 above.
3687
3688 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3689 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3690 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3691 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3692
3693 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3694 mode. Unlike ".", it always matches newline, whatever the setting of
3695 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3696 assertions. Perl allows it, but it doesn't (in general) work because it can't
3697 calculate the length of the lookbehind. At least, that's the case for Perl
3698 5.8.0 - I've been told they are going to document that it doesn't work in
3699 future.
3700
3701 18. Added an error diagnosis for escapes that PCRE does not support: these are
3702 \L, \l, \N, \P, \p, \U, \u, and \X.
3703
3704 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3705 reading past the end of the pattern in cases such as /[abcd/.
3706
3707 20. PCRE was getting more memory than necessary for patterns with classes that
3708 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3709
3710 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3711 compiling PCRE for use with Virtual Pascal.
3712
3713 22. Small fix to the Makefile to make it work properly if the build is done
3714 outside the source tree.
3715
3716 23. Added a new extension: a condition to go with recursion. If a conditional
3717 subpattern starts with (?(R) the "true" branch is used if recursion has
3718 happened, whereas the "false" branch is used only at the top level.
3719
3720 24. When there was a very long string of literal characters (over 255 bytes
3721 without UTF support, over 250 bytes with UTF support), the computation of how
3722 much memory was required could be incorrect, leading to segfaults or other
3723 strange effects.
3724
3725 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3726 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3727 there was a subsequent back reference to those brackets. This meant that, for
3728 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3729 possible to check for precisely this case. All we can do is abandon the
3730 optimization if .* occurs inside capturing brackets when there are any back
3731 references whatsoever. (See below for a better fix that came later.)
3732
3733 26. The handling of the optimization for finding the first character of a
3734 non-anchored pattern, and for finding a character that is required later in the
3735 match were failing in some cases. This didn't break the matching; it just
3736 failed to optimize when it could. The way this is done has been re-implemented.
3737
3738 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3739
3740 28. Added a new feature that provides some of the functionality that Perl
3741 provides with (?{...}). The facility is termed a "callout". The way it is done
3742 in PCRE is for the caller to provide an optional function, by setting
3743 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3744 global variable. By default it is unset, which disables all calling out. To get
3745 the function called, the regex must include (?C) at appropriate points. This
3746 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3747 This provides a means of identifying different callout points. When PCRE
3748 reaches such a point in the regex, if pcre_callout has been set, the external
3749 function is called. It is provided with data in a structure called
3750 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3751 matching continues; if it returns a non-zero value, the match at the current
3752 point fails. However, backtracking will occur if possible. [This was changed
3753 later and other features added - see item 49 below.]
3754
3755 29. pcretest is upgraded to test the callout functionality. It provides a
3756 callout function that displays information. By default, it shows the start of
3757 the match and the current position in the text. There are some new data escapes
3758 to vary what happens:
3759
3760 \C+ in addition, show current contents of captured substrings
3761 \C- do not supply a callout function
3762 \C!n return 1 when callout number n is reached
3763 \C!n!m return 1 when callout number n is reached for the mth time
3764
3765 30. If pcregrep was called with the -l option and just a single file name, it
3766 output "<stdin>" if a match was found, instead of the file name.
3767
3768 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3769 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3770 pcre_exec(). This saves a malloc/free per call. The default value of
3771 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3772 when configuring.
3773
3774 32. The default maximum size of a compiled pattern is 64K. There have been a
3775 few cases of people hitting this limit. The code now uses macros to handle the
3776 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3777 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3778 configuring. Tests 2 and 5 work only with 2-byte links because they output
3779 debugging information about compiled patterns.
3780
3781 33. Internal code re-arrangements:
3782
3783 (a) Moved the debugging function for printing out a compiled regex into
3784 its own source file (printint.c) and used #include to pull it into
3785 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3786 separate copies.
3787
3788 (b) Defined the list of op-code names for debugging as a macro in
3789 internal.h so that it is next to the definition of the opcodes.
3790
3791 (c) Defined a table of op-code lengths for simpler skipping along compiled
3792 code. This is again a macro in internal.h so that it is next to the
3793 definition of the opcodes.
3794
3795 34. Added support for recursive calls to individual subpatterns, along the
3796 lines of Robin Houston's patch (but implemented somewhat differently).
3797
3798 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3799 allow it to read and process whole directories in Win32. This code was
3800 contributed by Lionel Fourquaux; it has not been tested by me.
3801
3802 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3803 used to name a group. Names consist of alphanumerics and underscores, and must
3804 be unique. Back references use the syntax (?P=name) and recursive calls use
3805 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3806 numbers. The function pcre_fullinfo() can be used after compilation to extract
3807 a name/number map. There are three relevant calls:
3808
3809 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3810 PCRE_INFO_NAMECOUNT yields the number of entries
3811 PCRE_INFO_NAMETABLE yields a pointer to the map.
3812
3813 The map is a vector of fixed-size entries. The size of each entry depends on
3814 the length of the longest name used. The first two bytes of each entry are the
3815 group number, most significant byte first. There follows the corresponding
3816 name, zero terminated. The names are in alphabetical order.
3817
3818 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3819 case instead of 255. Making it the same both with and without UTF-8 support
3820 means that the same test output works with both.
3821
3822 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3823 calling malloc() with a zero argument.
3824
3825 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3826 optimization. I've improved things by keeping a bitmap of backreferences with
3827 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3828 fact referenced, the optimization can be applied. It is unlikely that a
3829 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3830 the match to follow \n) will appear inside brackets with a number greater than
3831 31, but if it does, any back reference > 31 suppresses the optimization.
3832
3833 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3834 of disabling numbered capturing parentheses. Any opening parenthesis that is
3835 not followed by ? behaves as if it were followed by ?: but named parentheses
3836 can still be used for capturing (and they will acquire numbers in the usual
3837 way).
3838
3839 41. Redesigned the return codes from the match() function into yes/no/error so
3840 that errors can be passed back from deep inside the nested calls. A malloc
3841 failure while inside a recursive subpattern call now causes the
3842 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3843
3844 42. It is now possible to set a limit on the number of times the match()
3845 function is called in a call to pcre_exec(). This facility makes it possible to
3846 limit the amount of recursion and backtracking, though not in a directly
3847 obvious way, because the match() function is used in a number of different
3848 circumstances. The count starts from zero for each position in the subject
3849 string (for non-anchored patterns). The default limit is, for compatibility, a
3850 large number, namely 10 000 000. You can change this in two ways:
3851
3852 (a) When configuring PCRE before making, you can use --with-match-limit=n
3853 to set a default value for the compiled library.
3854
3855 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3856 a different value is set. See 45 below.
3857
3858 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3859
3860 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3861 of things that can be changed at compile time. The first argument specifies
3862 what is wanted and the second points to where the information is to be placed.
3863 The current list of available information is:
3864
3865 PCRE_CONFIG_UTF8
3866
3867 The output is an integer that is set to one if UTF-8 support is available;
3868 otherwise it is set to zero.
3869
3870 PCRE_CONFIG_NEWLINE
3871
3872 The output is an integer that it set to the value of the code that is used for
3873 newline. It is either LF (10) or CR (13).
3874
3875 PCRE_CONFIG_LINK_SIZE
3876
3877 The output is an integer that contains the number of bytes used for internal
3878 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3879
3880 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3881
3882 The output is an integer that contains the threshold above which the POSIX
3883 interface uses malloc() for output vectors. See item 31 above.
3884
3885 PCRE_CONFIG_MATCH_LIMIT
3886
3887 The output is an unsigned integer that contains the default limit of the number
3888 of match() calls in a pcre_exec() execution. See 42 above.
3889
3890 44. pcretest has been upgraded by the addition of the -C option. This causes it
3891 to extract all the available output from the new pcre_config() function, and to
3892 output it. The program then exits immediately.
3893
3894 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3895 order to support additional features. One way would have been to define
3896 pcre_exec2() (for example) with extra arguments, but this would not have been
3897 extensible, and would also have required all calls to the original function to
3898 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3899 is used for passing in "extra" data from pcre_study().
3900
3901 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3902 contains the following fields:
3903
3904 flags a bitmap indicating which of the following fields are set
3905 study_data opaque data from pcre_study()
3906 match_limit a way of specifying a limit on match() calls for a specific
3907 call to pcre_exec()
3908 callout_data data for callouts (see 49 below)
3909
3910 The flag bits are also defined in pcre.h, and are
3911
3912 PCRE_EXTRA_STUDY_DATA
3913 PCRE_EXTRA_MATCH_LIMIT
3914 PCRE_EXTRA_CALLOUT_DATA
3915
3916 The pcre_study() function now returns one of these new pcre_extra blocks, with
3917 the actual study data pointed to by the study_data field, and the
3918 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3919 before. That is, this change is entirely upwards-compatible and requires no
3920 change to existing code.
3921
3922 If you want to pass in additional data to pcre_exec(), you can either place it
3923 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3924 block.
3925
3926 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3927 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3928 times with different match limits, until it finds the minimum value needed for
3929 pcre_exec() to complete. The value is then output. This can be instructive; for
3930 most simple matches the number is quite small, but for pathological cases it
3931 gets very large very quickly.
3932
3933 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3934 returns the size of the data block pointed to by the study_data field in a
3935 pcre_extra block, that is, the value that was passed as the argument to
3936 pcre_malloc() when PCRE was getting memory in which to place the information
3937 created by pcre_study(). The fourth argument should point to a size_t variable.
3938 pcretest has been extended so that this information is shown after a successful
3939 pcre_study() call when information about the compiled regex is being displayed.
3940
3941 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3942 because what follows is always an absolute path. (Later: it turns out that this
3943 is more than cosmetic for MinGW, because it doesn't like empty path
3944 components.)
3945
3946 49. Some changes have been made to the callout feature (see 28 above):
3947
3948 (i) A callout function now has three choices for what it returns:
3949
3950 0 => success, carry on matching
3951 > 0 => failure at this point, but backtrack if possible
3952 < 0 => serious error, return this value from pcre_exec()
3953
3954 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3955 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3956 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3957 use by callout functions. It will never be used by PCRE itself.
3958
3959 (ii) The pcre_extra structure (see 45 above) has a void * field called
3960 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3961 pcre_callout_block structure has a field of the same name. The contents of
3962 the field passed in the pcre_extra structure are passed to the callout
3963 function in the corresponding field in the callout block. This makes it
3964 easier to use the same callout-containing regex from multiple threads. For
3965 testing, the pcretest program has a new data escape
3966
3967 \C*n pass the number n (may be negative) as callout_data
3968
3969 If the callout function in pcretest receives a non-zero value as
3970 callout_data, it returns that value.
3971
3972 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3973 there were some redundant $(CFLAGS) in commands that are now specified as
3974 $(LINK), which already includes $(CFLAGS).
3975
3976 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3977 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3978 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3979 one-byte characters throughout. Note that case-insensitive matching applies
3980 only to characters whose values are less than 256. PCRE doesn't support the
3981 notion of cases for higher-valued characters.
3982
3983 (i) A character class whose characters are all within 0-255 is handled as
3984 a bit map, and the map is inverted for negative classes. Previously, a
3985 character > 255 always failed to match such a class; however it should
3986 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3987
3988 (ii) A negated character class with a single character < 255 is coded as
3989 "not this character" (OP_NOT). This wasn't working properly when the test
3990 character was multibyte, either singly or repeated.
3991
3992 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3993 mode, for example: \x{100}{2,3}.
3994
3995 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3996 singly or repeated) now correctly test multibyte characters. However,
3997 PCRE doesn't recognize any characters with values greater than 255 as
3998 digits, spaces, or word characters. Such characters always match \D, \S,
3999 and \W, and never match \d, \s, or \w.
4000
4001 (v) Classes may now contain characters and character ranges with values
4002 greater than 255. For example: [ab\x{100}-\x{400}].
4003
4004 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
4005 PCRE in UTF-8 mode.
4006
4007 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
4008 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
4009 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
4010 value.)
4011
4012 53. The single man page has become too large. I have therefore split it up into
4013 a number of separate man pages. These also give rise to individual HTML pages;
4014 these are now put in a separate directory, and there is an index.html page that
4015 lists them all. Some hyperlinking between the pages has been installed.
4016
4017 54. Added convenience functions for handling named capturing parentheses.
4018
4019 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4020 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4021 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4022 are faulted.
4023
4024 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4025 calling configure. These values are used when compiling the dftables.c program
4026 which is run to generate the source of the default character tables. They
4027 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4028 you will need to set these values.
4029
4030 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4031
4032
4033 Version 3.9 02-Jan-02
4034 ---------------------
4035
4036 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4037
4038 2. If --disable-static was given, the building process failed when trying to
4039 build pcretest and pcregrep. (For some reason it was using libtool to compile
4040 them, which is not right, as they aren't part of the library.)
4041
4042
4043 Version 3.8 18-Dec-01
4044 ---------------------
4045
4046 1. The experimental UTF-8 code was completely screwed up. It was packing the
4047 bytes in the wrong order. How dumb can you get?
4048
4049
4050 Version 3.7 29-Oct-01
4051 ---------------------
4052
4053 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4054 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4055 this didn't happen under Solaris 8, where I normally test things.
4056
4057 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4058 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4059 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4060 this fix an hour or so after the initial 3.7 release.)
4061
4062
4063 Version 3.6 23-Oct-01
4064 ---------------------
4065
4066 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4067 offsets passed as NULL with zero offset count.
4068
4069 2. The config.guess and config.sub files had not been updated when I moved to
4070 the latest autoconf.
4071
4072
4073 Version 3.5 15-Aug-01
4074 ---------------------
4075
4076 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4077 had been forgotten.
4078
4079 2. By using declared but undefined structures, we can avoid using "void"
4080 definitions in pcre.h while keeping the internal definitions of the structures
4081 private.
4082
4083 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4084 user point of view, this means that both static and shared libraries are built
4085 by default, but this can be individually controlled. More of the work of
4086 handling this static/shared cases is now inside libtool instead of PCRE's make
4087 file.
4088
4089 4. The pcretest utility is now installed along with pcregrep because it is
4090 useful for users (to test regexs) and by doing this, it automatically gets
4091 relinked by libtool. The documentation has been turned into a man page, so
4092 there are now .1, .txt, and .html versions in /doc.
4093
4094 5. Upgrades to pcregrep:
4095 (i) Added long-form option names like gnu grep.
4096 (ii) Added --help to list all options with an explanatory phrase.
4097 (iii) Added -r, --recursive to recurse into sub-directories.
4098 (iv) Added -f, --file to read patterns from a file.
4099
4100 6. pcre_exec() was referring to its "code" argument before testing that
4101 argument for NULL (and giving an error if it was NULL).
4102
4103 7. Upgraded Makefile.in to allow for compiling in a different directory from
4104 the source directory.
4105
4106 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4107 options bits, the pointer it was passed was to an int instead of to an unsigned
4108 long int. This mattered only on 64-bit systems.
4109
4110 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4111 generated) instead of pcre.in, which it its source. Also made the same change
4112 in several of the .c files.
4113
4114 10. A new release of gcc defines printf() as a macro, which broke pcretest
4115 because it had an ifdef in the middle of a string argument for printf(). Fixed
4116 by using separate calls to printf().
4117
4118 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4119 script, to force use of CR or LF instead of \n in the source. On non-Unix
4120 systems, the value can be set in config.h.
4121
4122 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4123 absolute limit. Changed the text of the error message to make this clear, and
4124 likewise updated the man page.
4125
4126 13. The limit of 99 on the number of capturing subpatterns has been removed.
4127 The new limit is 65535, which I hope will not be a "real" limit.
4128
4129
4130 Version 3.4 22-Aug-00
4131 ---------------------
4132
4133 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4134
4135 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4136
4137
4138 Version 3.3 01-Aug-00
4139 ---------------------
4140
4141 1. If an octal character was given, but the value was greater than \377, it
4142 was not getting masked to the least significant bits, as documented. This could
4143 lead to crashes in some systems.
4144
4145 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4146 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4147
4148 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4149 These just pass their arguments on to (pcre_free)(), but they are provided
4150 because some uses of PCRE bind it to non-C systems that can call its functions,
4151 but cannot call free() or pcre_free() directly.
4152
4153 4. Add "make test" as a synonym for "make check". Corrected some comments in
4154 the Makefile.
4155
4156 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4157 Makefile.
4158
4159 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4160 command called pgrep for grepping around the active processes.
4161
4162 7. Added the beginnings of support for UTF-8 character strings.
4163
4164 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4165 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4166 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4167 out for the ar command.)
4168
4169
4170 Version 3.2 12-May-00
4171 ---------------------
4172
4173 This is purely a bug fixing release.
4174
4175 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4176 of ZA. This was just one example of several cases that could provoke this bug,
4177 which was introduced by change 9 of version 2.00. The code for breaking
4178 infinite loops after an iteration that matches an empty string was't working
4179 correctly.
4180
4181 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4182 when matched against abbab (for example). After matching an empty string, it
4183 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4184 caused it to match further down the string than it should.
4185
4186 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4187 was there because it doesn't seem to be needed, and it causes trouble on some
4188 systems, as it is not a Standard C header. It has been removed.
4189
4190 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4191 were reported on the Macintosh. The changes were from
4192
4193 while ((c = *(++ptr)) != 0 && c != '\n');
4194 to
4195 while ((c = *(++ptr)) != 0 && c != '\n') ;
4196
4197 Totally extraordinary, but if that's what it takes...
4198
4199 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4200 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4201 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4202 assumes the way PCRE uses memmove() (always moving upwards).
4203
4204 6. PCRE is being used in one environment where strchr() is not available. There
4205 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4206 faster code anyway.
4207
4208
4209 Version 3.1 09-Feb-00
4210 ---------------------
4211
4212 The only change in this release is the fixing of some bugs in Makefile.in for
4213 the "install" target:
4214
4215 (1) It was failing to install pcreposix.h.
4216
4217 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4218
4219
4220 Version 3.0 01-Feb-00
4221 ---------------------
4222
4223 1. Add support for the /+ modifier to perltest (to output $` like it does in
4224 pcretest).
4225
4226 2. Add support for the /g modifier to perltest.
4227
4228 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4229 matches null strings.
4230
4231 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4232 pattern. Perl treats empty patterns specially - it reuses the most recent
4233 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4234 effect.
4235
4236 5. The POSIX interface was broken in that it was just handing over the POSIX
4237 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4238 required a bigger vector, with some working space on the end. This means that
4239 the POSIX wrapper now has to get and free some memory, and copy the results.
4240
4241 6. Added some simple autoconf support, placing the test data and the
4242 documentation in separate directories, re-organizing some of the
4243 information files, and making it build pcre-config (a GNU standard). Also added
4244 libtool support for building PCRE as a shared library, which is now the
4245 default.
4246
4247 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4248 09 are not valid octal constants. Single digits will be used for minor values
4249 less than 10.
4250
4251 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4252 existing programs that set these in the POSIX interface can use PCRE without
4253 modification.
4254
4255 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4256 return all that pcre_info() returns, plus additional data. The pcre_info()
4257 function is retained for compatibility, but is considered to be obsolete.
4258
4259 10. Added experimental recursion feature (?R) to handle one common case that
4260 Perl 5.6 will be able to do with (?p{...}).
4261
4262 11. Added support for POSIX character classes like [:alpha:], which Perl is
4263 adopting.
4264
4265
4266 Version 2.08 31-Aug-99
4267 ----------------------
4268
4269 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4270 trying to match at the startoffset position, but instead was moving forward to
4271 the next newline as if a previous match had failed.
4272
4273 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4274 and could get into a loop if a null string was matched other than at the start
4275 of the subject.
4276
4277 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4278 be distinguished at compile time, and for completeness also added PCRE_DATE.
4279
4280 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4281 in GnuWin32 environments.
4282
4283
4284 Version 2.07 29-Jul-99
4285 ----------------------
4286
4287 1. The documentation is now supplied in plain text form and HTML as well as in
4288 the form of man page sources.
4289
4290 2. C++ compilers don't like assigning (void *) values to other pointer types.
4291 In particular this affects malloc(). Although there is no problem in Standard
4292 C, I've put in casts to keep C++ compilers happy.
4293
4294 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4295 should be (const char *).
4296
4297 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4298 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4299 However, I haven't made this a standard facility. The documentation doesn't
4300 mention it, and the Makefile doesn't support it.
4301
4302 5. The Makefile now contains an "install" target, with editable destinations at
4303 the top of the file. The pcretest program is not installed.
4304
4305 6. pgrep -V now gives the PCRE version number and date.
4306
4307 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4308 causing the entire string to be ignored, instead of just the last character.
4309
4310 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4311 non-matching string, it can take a very, very long time, even for strings of
4312 quite modest length, because of the nested recursion. PCRE now does better in
4313 some of these cases. It does this by remembering the last required literal
4314 character in the pattern, and pre-searching the subject to ensure it is present
4315 before running the real match. In other words, it applies a heuristic to detect
4316 some types of certain failure quickly, and in the above example, if presented
4317 with a string that has no trailing " it gives "no match" very quickly.
4318
4319 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4320 other alternatives are tried instead.
4321
4322
4323 Version 2.06 09-Jun-99
4324 ----------------------
4325
4326 1. Change pcretest's output for amount of store used to show just the code
4327 space, because the remainder (the data block) varies in size between 32-bit and
4328 64-bit systems.
4329
4330 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4331 start matching at. This allows lookbehinds to work when searching for multiple
4332 occurrences in a string.
4333
4334 3. Added additional options to pcretest for testing multiple occurrences:
4335
4336 /+ outputs the rest of the string that follows a match
4337 /g loops for multiple occurrences, using the new startoffset argument
4338 /G loops for multiple occurrences by passing an incremented pointer
4339
4340 4. PCRE wasn't doing the "first character" optimization for patterns starting
4341 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4342 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4343 the letter 'x'. On long subject strings, this gives a significant speed-up.
4344
4345
4346 Version 2.05 21-Apr-99
4347 ----------------------
4348
4349 1. Changed the type of magic_number from int to long int so that it works
4350 properly on 16-bit systems.
4351
4352 2. Fixed a bug which caused patterns starting with .* not to work correctly
4353 when the subject string contained newline characters. PCRE was assuming
4354 anchoring for such patterns in all cases, which is not correct because .* will
4355 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4356 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4357 must be retried after every newline in the subject.
4358
4359
4360 Version 2.04 18-Feb-99
4361 ----------------------
4362
4363 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4364 computation of the store needed to hold the pattern was incorrect (too large).
4365 If such patterns were nested a few deep, this could multiply and become a real
4366 problem.
4367
4368 2. Added /M option to pcretest to show the memory requirement of a specific
4369 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4370
4371 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4372 compiled in such a way that the backtracking after subsequent failure was
4373 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4374 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4375
4376
4377 Version 2.03 02-Feb-99
4378 ----------------------
4379
4380 1. Fixed typo and small mistake in man page.
4381
4382 2. Added 4th condition (GPL supersedes if conflict) and created separate
4383 LICENCE file containing the conditions.
4384
4385 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4386 Perl, that is the internal \ allows the delimiter to be included in the
4387 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4388 the final delimiter, add \ to the end of the pattern (to test the error).
4389
4390 4. Added the convenience functions for extracting substrings after a successful
4391 match. Updated pcretest to make it able to test these functions.
4392
4393
4394 Version 2.02 14-Jan-99
4395 ----------------------
4396
4397 1. Initialized the working variables associated with each extraction so that
4398 their saving and restoring doesn't refer to uninitialized store.
4399
4400 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4401 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4402 fix the problem.
4403
4404 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4405 calls, and wasn't printing the correct value for compiling calls. Increased the
4406 default value of LOOPREPEAT, and the number of significant figures in the
4407 times.
4408
4409 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4410
4411 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4412 a building problem on Windows NT with a FAT file system.
4413
4414
4415 Version 2.01 21-Oct-98
4416 ----------------------
4417
4418 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4419 to character tables built by pcre_maketables() in the current locale. If NULL
4420 is passed, the default tables are used.
4421
4422
4423 Version 2.00 24-Sep-98
4424 ----------------------
4425
4426 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4427 it any more.
4428
4429 2. Allow quantification of (?>) groups, and make it work correctly.
4430
4431 3. The first character computation wasn't working for (?>) groups.
4432
4433 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4434 end of the subject) and add 5.005's \z, which really does match only at the
4435 very end of the subject.
4436
4437 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4438
4439 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4440 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4441 localized options. All options to pcre_study() were also removed.
4442
4443 7. Add other new features from 5.005:
4444
4445 $(?<= positive lookbehind
4446 $(?<! negative lookbehind
4447 (?imsx-imsx) added the unsetting capability
4448 such a setting is global if at outer level; local otherwise
4449 (?imsx-imsx:) non-capturing groups with option setting
4450 (?(cond)re|re) conditional pattern matching
4451
4452 A backreference to itself in a repeated group matches the previous
4453 captured string.
4454
4455 8. General tidying up of studying (both automatic and via "study")
4456 consequential on the addition of new assertions.
4457
4458 9. As in 5.005, unlimited repeated groups that could match an empty substring
4459 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4460 runtime if any iteration does actually match an empty substring.
4461
4462 10. Include the RunTest script in the distribution.
4463
4464 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4465 discrepancies, some of which were old and were also with respect to 5.004. They
4466 have now been fixed.
4467
4468
4469 Version 1.09 28-Apr-98
4470 ----------------------
4471
4472 1. A negated single character class followed by a quantifier with a minimum
4473 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4474 program crashes, or just wrong answers. This did not apply to negated classes
4475 containing more than one character, or to minima other than one.
4476
4477
4478 Version 1.08 27-Mar-98
4479 ----------------------
4480
4481 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4482
4483 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4484 latter must appear before anything that relies on it in the pattern.
4485
4486
4487 Version 1.07 16-Feb-98
4488 ----------------------
4489
4490 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4491 repeat of a potentially empty string).
4492
4493
4494 Version 1.06 23-Jan-98
4495 ----------------------
4496
4497 1. Added Markus Oberhumer's little patches for C++.
4498
4499 2. Literal strings longer than 255 characters were broken.
4500
4501
4502 Version 1.05 23-Dec-97
4503 ----------------------
4504
4505 1. Negated character classes containing more than one character were failing if
4506 PCRE_CASELESS was set at run time.
4507
4508
4509 Version 1.04 19-Dec-97
4510 ----------------------
4511
4512 1. Corrected the man page, where some "const" qualifiers had been omitted.
4513
4514 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4515 input syntax.
4516
4517 3. Fixed memory leak which occurred when a regex with back references was
4518 matched with an offsets vector that wasn't big enough. The temporary memory
4519 that is used in this case wasn't being freed if the match failed.
4520
4521 4. Tidied pcretest to ensure it frees memory that it gets.
4522
4523 5. Temporary memory was being obtained in the case where the passed offsets
4524 vector was exactly big enough.
4525
4526 6. Corrected definition of offsetof() from change 5 below.
4527
4528 7. I had screwed up change 6 below and broken the rules for the use of
4529 setjmp(). Now fixed.
4530
4531
4532 Version 1.03 18-Dec-97
4533 ----------------------
4534
4535 1. A erroneous regex with a missing opening parenthesis was correctly
4536 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4537 on some systems.
4538
4539 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4540 it was reported that one broken compiler failed on the former because "code" is
4541 also an independent variable.
4542
4543 3. The erroneous regex a[]b caused an array overrun reference.
4544
4545 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4546 fail on data ending with that character. (It was going on too far, and checking
4547 the next character, typically a binary zero.) This was specific to the
4548 optimized code for single-character negative classes.
4549
4550 5. Added a contributed patch from the TIN world which does the following:
4551
4552 + Add an undef for memmove, in case the the system defines a macro for it.
4553
4554 + Add a definition of offsetof(), in case there isn't one. (I don't know
4555 the reason behind this - offsetof() is part of the ANSI standard - but
4556 it does no harm).
4557
4558 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4559 most of the places where whitespace preceded '#'. I have given up and
4560 allowed the remaining 2 cases to be at the margin.
4561
4562 + Rename some variables in pcre to eliminate shadowing. This seems very
4563 pedantic, but does no harm, of course.
4564
4565 6. Moved the call to setjmp() into its own function, to get rid of warnings
4566 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4567
4568 7. Constructs such as \d{8,} were compiling into the equivalent of
4569 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4570 outcome, but in this particular case used more store than had been allocated,
4571 which caused the bug to be discovered because it threw up an internal error.
4572
4573 8. The debugging code in both pcre and pcretest for outputting the compiled
4574 form of a regex was going wrong in the case of back references followed by
4575 curly-bracketed repeats.
4576
4577
4578 Version 1.02 12-Dec-97
4579 ----------------------
4580
4581 1. Typos in pcre.3 and comments in the source fixed.
4582
4583 2. Applied a contributed patch to get rid of places where it used to remove
4584 'const' from variables, and fixed some signed/unsigned and uninitialized
4585 variable warnings.
4586
4587 3. Added the "runtest" target to Makefile.
4588
4589 4. Set default compiler flag to -O2 rather than just -O.
4590
4591
4592 Version 1.01 19-Nov-97
4593 ----------------------
4594
4595 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4596 like /([ab]*)*/, that is, for classes with more than one character in them.
4597
4598 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4599 as /((?>a*))*/ (a PCRE_EXTRA facility).
4600
4601
4602 Version 1.00 18-Nov-97
4603 ----------------------
4604
4605 1. Added compile-time macros to support systems such as SunOS4 which don't have
4606 memmove() or strerror() but have other things that can be used instead.
4607
4608 2. Arranged that "make clean" removes the executables.
4609
4610
4611 Version 0.99 27-Oct-97
4612 ----------------------
4613
4614 1. Fixed bug in code for optimizing classes with only one character. It was
4615 initializing a 32-byte map regardless, which could cause it to run off the end
4616 of the memory it had got.
4617
4618 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4619
4620
4621 Version 0.98 22-Oct-97
4622 ----------------------
4623
4624 1. Fixed bug in code for handling temporary memory usage when there are more
4625 back references than supplied space in the ovector. This could cause segfaults.
4626
4627
4628 Version 0.97 21-Oct-97
4629 ----------------------
4630
4631 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4632
4633 2. Optimized negated single characters not to use a bit map.
4634
4635 3. Brought error texts together as macro definitions; clarified some of them;
4636 fixed one that was wrong - it said "range out of order" when it meant "invalid
4637 escape sequence".
4638
4639 4. Changed some char * arguments to const char *.
4640
4641 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4642
4643 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4644 pcretest.
4645
4646
4647 Version 0.96 16-Oct-97
4648 ----------------------
4649
4650 1. Added a simple "pgrep" utility to the distribution.
4651
4652 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4653 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4654 where "ddd" means "one or more decimal digits".
4655
4656 3. Fixed serious bug. If a pattern had a back reference, but the call to
4657 pcre_exec() didn't supply a large enough ovector to record the related
4658 identifying subpattern, the match always failed. PCRE now remembers the number
4659 of the largest back reference, and gets some temporary memory in which to save
4660 the offsets during matching if necessary, in order to ensure that
4661 backreferences always work.
4662
4663 4. Increased the compatibility with Perl in a number of ways:
4664
4665 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4666 to request this handling. The option can be set at compile or exec time.
4667
4668 (b) $ matches before a terminating newline by default; an option
4669 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4670 mode). The option can be set at compile or exec time.
4671
4672 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4673 the same as Perl's. If the decimal number it represents is less than 10
4674 or there aren't that many previous left capturing parentheses, an octal
4675 escape is read. Inside a character class, it's always an octal escape,
4676 even if it is a single digit.
4677
4678 (d) An escaped but undefined alphabetic character is taken as a literal,
4679 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4680 escapes.
4681
4682 (e) {0} is now permitted. (The previous item is removed from the compiled
4683 pattern).
4684
4685 5. Changed all the names of code files so that the basic parts are no longer
4686 than 10 characters, and abolished the teeny "globals.c" file.
4687
4688 6. Changed the handling of character classes; they are now done with a 32-byte
4689 bit map always.
4690
4691 7. Added the -d and /D options to pcretest to make it possible to look at the
4692 internals of compilation without having to recompile pcre.
4693
4694
4695 Version 0.95 23-Sep-97
4696 ----------------------
4697
4698 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4699 \x20 at the start of a run of normal characters. These were being treated as
4700 real characters, instead of the source characters being re-checked.
4701
4702
4703 Version 0.94 18-Sep-97
4704 ----------------------
4705
4706 1. The functions are now thread-safe, with the caveat that the global variables
4707 containing pointers to malloc() and free() or alternative functions are the
4708 same for all threads.
4709
4710 2. Get pcre_study() to generate a bitmap of initial characters for non-
4711 anchored patterns when this is possible, and use it if passed to pcre_exec().
4712
4713
4714 Version 0.93 15-Sep-97
4715 ----------------------
4716
4717 1. /(b)|(:+)/ was computing an incorrect first character.
4718
4719 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4720 but not actually doing anything yet.
4721
4722 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4723 as Perl does (e.g. [-az] or [az-]).
4724
4725 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4726 all possible positions.
4727
4728 5. Split up into different modules to avoid including unneeded functions in a
4729 compiled binary. However, compile and exec are still in one module. The "study"
4730 function is split off.
4731
4732 6. The character tables are now in a separate module whose source is generated
4733 by an auxiliary program - but can then be edited by hand if required. There are
4734 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4735 toupper() in the code.
4736
4737 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4738 make them global. Abolish the function for setting them, as the caller can now
4739 set them directly.
4740
4741
4742 Version 0.92 11-Sep-97
4743 ----------------------
4744
4745 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4746 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4747
4748 2. Caseless matching was not working in character classes if the characters in
4749 the pattern were in upper case.
4750
4751 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4752
4753 4. Make PCRE_ANCHORED public and accept as a compile option.
4754
4755 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4756 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4757 pass them.
4758
4759 6. Give an error if bad option bits passed at compile or run time.
4760
4761 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4762 pcretest to cause it to pass that flag.
4763
4764 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4765 options, and the first character, if set.
4766
4767 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4768
4769
4770 Version 0.91 10-Sep-97
4771 ----------------------
4772
4773 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4774 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4775
4776 2. PCRE was looping on encountering an indefinitely repeated back reference to
4777 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4778 Perl does - treats the match as successful.
4779
4780 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5