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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 969 - (show annotations)
Sun May 27 04:33:38 2012 UTC (7 years, 6 months ago) by zherczeg
File size: 230938 byte(s)
Put spaces around SLJIT_PRINT_D in the JIT compiler
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 32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
120
121
122 Version 8.30 04-February-2012
123 -----------------------------
124
125 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
126 name is defined in ctype.h.
127
128 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
129 only in quite long subpatterns.
130
131 3. Removed the function pcre_info(), which has been obsolete and deprecated
132 since it was replaced by pcre_fullinfo() in February 2000.
133
134 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
135 match a (*MARK), and the match failed at the start of the subject, a
136 reference to memory before the start of the subject could occur. This bug
137 was introduced by fix 17 of release 8.21.
138
139 5. A reference to an unset group with zero minimum repetition was giving
140 totally wrong answers (in non-JavaScript-compatibility mode). For example,
141 /(another)?(\1?)test/ matched against "hello world test". This bug was
142 introduced in release 8.13.
143
144 6. Add support for 16-bit character strings (a large amount of work involving
145 many changes and refactorings).
146
147 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
148 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
149 from a file.
150
151 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
152 rounding is not applied in this particular case).
153
154 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
155 if they appear, or are escaped, in patterns.
156
157 10. Get rid of a number of -Wunused-but-set-variable warnings.
158
159 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
160 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
161 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
162 also returns the mark "x". This bug applied to capturing parentheses,
163 non-capturing parentheses, and atomic parentheses. It also applied to some
164 assertions.
165
166 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
167 information out of configure.ac instead of relying on pcre.h.generic, which
168 is not stored in the repository.
169
170 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
171 -lreadline.
172
173 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
174
175 15. Applied Graycode's patch to put the top-level frame on the stack rather
176 than the heap when not using the stack for recursion. This gives a
177 performance improvement in many cases when recursion is not deep.
178
179 16. Experimental code added to "pcretest -C" to output the stack frame size.
180
181
182 Version 8.21 12-Dec-2011
183 ------------------------
184
185 1. Updating the JIT compiler.
186
187 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
188 are added as well.
189
190 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
191 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
192 calling _pcre_jit_exec. Some extra comments are added.
193
194 4. (*MARK) settings inside atomic groups that do not contain any capturing
195 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
196 was introduced by change 18 for 8.20.
197
198 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
199 ECMA-262 standard.
200
201 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
202 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
203 This bug was probably introduced by change 9 of 8.13.
204
205 7. While fixing 6 above, I noticed that a number of other items were being
206 incorrectly rejected as "not fixed length". This arose partly because newer
207 opcodes had not been added to the fixed-length checking code. I have (a)
208 corrected the bug and added tests for these items, and (b) arranged for an
209 error to occur if an unknown opcode is encountered while checking for fixed
210 length instead of just assuming "not fixed length". The items that were
211 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
212 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
213 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
214
215 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
216 being incorrectly compiled and would have given unpredicatble results.
217
218 9. A possessively repeated subpattern with minimum repeat count greater than
219 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
220 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
221 the first (A) could occur when it should not.
222
223 10. Add a cast and remove a redundant test from the code.
224
225 11. JIT should use pcre_malloc/pcre_free for allocation.
226
227 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
228 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
229 is anything other than /usr, -L is still shown).
230
231 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
232
233 14. Perl does not support \N without a following name in a [] class; PCRE now
234 also gives an error.
235
236 15. If a forward reference was repeated with an upper limit of around 2000,
237 it caused the error "internal error: overran compiling workspace". The
238 maximum number of forward references (including repeats) was limited by the
239 internal workspace, and dependent on the LINK_SIZE. The code has been
240 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
241 the default depends on LINK_SIZE. There is a new upper limit (for safety)
242 of around 200,000 forward references. While doing this, I also speeded up
243 the filling in of repeated forward references.
244
245 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
246 incorrectly expecting the subject to contain another "a" after the start.
247
248 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
249 in the match, the SKIP should be ignored. This was not happening; instead
250 the SKIP was being treated as NOMATCH. For patterns such as
251 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
252 tested.
253
254 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
255 now much more compatible with Perl, in particular in cases where the result
256 is a non-match for a non-anchored pattern. For example, if
257 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
258 "m", where previously it did not return a name. A side effect of this
259 change is that for partial matches, the last encountered mark name is
260 returned, as for non matches. A number of tests that were previously not
261 Perl-compatible have been moved into the Perl-compatible test files. The
262 refactoring has had the pleasing side effect of removing one argument from
263 the match() function, thus reducing its stack requirements.
264
265 19. If the /S+ option was used in pcretest to study a pattern using JIT,
266 subsequent uses of /S (without +) incorrectly behaved like /S+.
267
268 21. Retrieve executable code size support for the JIT compiler and fixing
269 some warnings.
270
271 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
272 not work when the shorter character appeared right at the end of the
273 subject string.
274
275 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
276 systems.
277
278 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
279 output it when the /M option is used in pcretest.
280
281 25. The CheckMan script was not being included in the distribution. Also, added
282 an explicit "perl" to run Perl scripts from the PrepareRelease script
283 because this is reportedly needed in Windows.
284
285 26. If study data was being save in a file and studying had not found a set of
286 "starts with" bytes for the pattern, the data written to the file (though
287 never used) was taken from uninitialized memory and so caused valgrind to
288 complain.
289
290 27. Updated RunTest.bat as provided by Sheri Pierce.
291
292 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
293
294 29. Computation of memory usage for the table of capturing group names was
295 giving an unnecessarily large value.
296
297
298 Version 8.20 21-Oct-2011
299 ------------------------
300
301 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
302 a POSIX class. After further experiments with Perl, which convinced me that
303 Perl has bugs and confusions, a closing square bracket is no longer allowed
304 in a POSIX name. This bug also affected patterns with classes that started
305 with full stops.
306
307 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
308 captured substring, but while checking the failing first alternative,
309 substring 1 is temporarily captured. If the output vector supplied to
310 pcre_exec() was not big enough for this capture, the yield of the function
311 was still zero ("insufficient space for captured substrings"). This cannot
312 be totally fixed without adding another stack variable, which seems a lot
313 of expense for a edge case. However, I have improved the situation in cases
314 such as /(a)(b)x|abc/ matched against "abc", where the return code
315 indicates that fewer than the maximum number of slots in the ovector have
316 been set.
317
318 3. Related to (2) above: when there are more back references in a pattern than
319 slots in the output vector, pcre_exec() uses temporary memory during
320 matching, and copies in the captures as far as possible afterwards. It was
321 using the entire output vector, but this conflicts with the specification
322 that only 2/3 is used for passing back captured substrings. Now it uses
323 only the first 2/3, for compatibility. This is, of course, another edge
324 case.
325
326 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
327 main code base, and can be used by building with --enable-jit. When this is
328 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
329 runtime --no-jit option is given.
330
331 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
332 ovector, the return from the function was zero, implying that there were
333 other matches that did not fit. The correct "exactly full" value is now
334 returned.
335
336 6. If a subpattern that was called recursively or as a subroutine contained
337 (*PRUNE) or any other control that caused it to give a non-standard return,
338 invalid errors such as "Error -26 (nested recursion at the same subject
339 position)" or even infinite loops could occur.
340
341 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
342 computing the minimum length on reaching *ACCEPT, and so ended up with the
343 wrong value of 1 rather than 0. Further investigation indicates that
344 computing a minimum subject length in the presence of *ACCEPT is difficult
345 (think back references, subroutine calls), and so I have changed the code
346 so that no minimum is registered for a pattern that contains *ACCEPT.
347
348 8. If (*THEN) was present in the first (true) branch of a conditional group,
349 it was not handled as intended. [But see 16 below.]
350
351 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
352 Sheri Pierce.
353
354 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
355 the first byte in a match must be "a".
356
357 11. Change 17 for 8.13 increased the recursion depth for patterns like
358 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
359 pattern contains any instances of (*THEN). If it does not, the old
360 optimizations are restored. It would be nice to do this on a per-group
361 basis, but at the moment that is not feasible.
362
363 12. In some environments, the output of pcretest -C is CRLF terminated. This
364 broke RunTest's code that checks for the link size. A single white space
365 character after the value is now allowed for.
366
367 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
368 For "fr", it uses the Windows-specific input and output files.
369
370 14. If (*THEN) appeared in a group that was called recursively or as a
371 subroutine, it did not work as intended. [But see next item.]
372
373 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
374 pattern fragments (but not containing any | characters). If A and B are
375 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
376 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
377 D. In other words, Perl considers parentheses that do not contain any |
378 characters to be part of a surrounding alternative, whereas PCRE was
379 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
380 differently. PCRE now behaves in the same way as Perl, except in the case
381 of subroutine/recursion calls such as (?1) which have in any case always
382 been different (but PCRE had them first :-).
383
384 16. Related to 15 above: Perl does not treat the | in a conditional group as
385 creating alternatives. Such a group is treated in the same way as an
386 ordinary group without any | characters when processing (*THEN). PCRE has
387 been changed to match Perl's behaviour.
388
389 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
390 RunGrepTest script failed.
391
392 18. Change 22 for version 13 caused atomic groups to use more stack. This is
393 inevitable for groups that contain captures, but it can lead to a lot of
394 stack use in large patterns. The old behaviour has been restored for atomic
395 groups that do not contain any capturing parentheses.
396
397 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
398 suppress the check for a minimum subject length at run time. (If it was
399 given to pcre_exec() or pcre_dfa_exec() it did work.)
400
401 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
402 fail to work when decoding hex characters in data strings in EBCDIC
403 environments.
404
405 21. It appears that in at least one Mac OS environment, the isxdigit() function
406 is implemented as a macro that evaluates to its argument more than once,
407 contravening the C 90 Standard (I haven't checked a later standard). There
408 was an instance in pcretest which caused it to go wrong when processing
409 \x{...} escapes in subject strings. The has been rewritten to avoid using
410 things like p++ in the argument of isxdigit().
411
412
413 Version 8.13 16-Aug-2011
414 ------------------------
415
416 1. The Unicode data tables have been updated to Unicode 6.0.0.
417
418 2. Two minor typos in pcre_internal.h have been fixed.
419
420 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
421 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
422 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
423
424 4. There were a number of related bugs in the code for matching backrefences
425 caselessly in UTF-8 mode when codes for the characters concerned were
426 different numbers of bytes. For example, U+023A and U+2C65 are an upper
427 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
428 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
429 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
430 2-byte code at the end of the subject (it thought there wasn't enough data
431 left).
432
433 5. Comprehensive information about what went wrong is now returned by
434 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
435 as the output vector has at least 2 elements. The offset of the start of
436 the failing character and a reason code are placed in the vector.
437
438 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
439 now returned is for the first byte of the failing character, instead of the
440 last byte inspected. This is an incompatible change, but I hope it is small
441 enough not to be a problem. It makes the returned offset consistent with
442 pcre_exec() and pcre_dfa_exec().
443
444 7. pcretest now gives a text phrase as well as the error number when
445 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
446 failure, the offset and reason code are output.
447
448 8. When \R was used with a maximizing quantifier it failed to skip backwards
449 over a \r\n pair if the subsequent match failed. Instead, it just skipped
450 back over a single character (\n). This seems wrong (because it treated the
451 two characters as a single entity when going forwards), conflicts with the
452 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
453 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
454 has been changed.
455
456 9. Some internal refactoring has changed the processing so that the handling
457 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
458 time (the PCRE_DOTALL option was changed this way some time ago: version
459 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
460 which was always a bit of a fudge. It also means that there is one less
461 argument for the match() function, which reduces its stack requirements
462 slightly. This change also fixes an incompatibility with Perl: the pattern
463 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
464
465 10. More internal refactoring has drastically reduced the number of recursive
466 calls to match() for possessively repeated groups such as (abc)++ when
467 using pcre_exec().
468
469 11. While implementing 10, a number of bugs in the handling of groups were
470 discovered and fixed:
471
472 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
473 (a|)*(?1) gave a compile-time internal error.
474 ((a|)+)+ did not notice that the outer group could match an empty string.
475 (^a|^)+ was not marked as anchored.
476 (.*a|.*)+ was not marked as matching at start or after a newline.
477
478 12. Yet more internal refactoring has removed another argument from the match()
479 function. Special calls to this function are now indicated by setting a
480 value in a variable in the "match data" data block.
481
482 13. Be more explicit in pcre_study() instead of relying on "default" for
483 opcodes that mean there is no starting character; this means that when new
484 ones are added and accidentally left out of pcre_study(), testing should
485 pick them up.
486
487 14. The -s option of pcretest has been documented for ages as being an old
488 synonym of -m (show memory usage). I have changed it to mean "force study
489 for every regex", that is, assume /S for every regex. This is similar to -i
490 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
491 using it. It makes it easier to run collections of tests with and without
492 study enabled, and thereby test pcre_study() more easily. All the standard
493 tests are now run with and without -s (but some patterns can be marked as
494 "never study" - see 20 below).
495
496 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
497 restoration of the capturing data to the outer values was not happening
498 correctly.
499
500 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
501 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
502 pattern had matched an empty string, and so incorrectly returned a no
503 match.
504
505 17. There was optimizing code for the last branch of non-capturing parentheses,
506 and also for the obeyed branch of a conditional subexpression, which used
507 tail recursion to cut down on stack usage. Unfortunately, now that there is
508 the possibility of (*THEN) occurring in these branches, tail recursion is
509 no longer possible because the return has to be checked for (*THEN). These
510 two optimizations have therefore been removed. [But see 8.20/11 above.]
511
512 18. If a pattern containing \R was studied, it was assumed that \R always
513 matched two bytes, thus causing the minimum subject length to be
514 incorrectly computed because \R can also match just one byte.
515
516 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
517 was incorrectly computed.
518
519 20. If /S is present twice on a test pattern in pcretest input, it now
520 *disables* studying, thereby overriding the use of -s on the command line
521 (see 14 above). This is necessary for one or two tests to keep the output
522 identical in both cases.
523
524 21. When (*ACCEPT) was used in an assertion that matched an empty string and
525 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
526
527 22. When an atomic group that contained a capturing parenthesis was
528 successfully matched, but the branch in which it appeared failed, the
529 capturing was not being forgotten if a higher numbered group was later
530 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
531 group 1 to "a", when in fact it should be unset. This applied to multi-
532 branched capturing and non-capturing groups, repeated or not, and also to
533 positive assertions (capturing in negative assertions does not happen
534 in PCRE) and also to nested atomic groups.
535
536 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
537 subject after a captured substring, to make it easier to tell which of a
538 number of identical substrings has been captured.
539
540 24. The way atomic groups are processed by pcre_exec() has been changed so that
541 if they are repeated, backtracking one repetition now resets captured
542 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
543 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
544 "aaa". Previously, it would have been "a". As part of this code
545 refactoring, the way recursive calls are handled has also been changed.
546
547 25. If an assertion condition captured any substrings, they were not passed
548 back unless some other capturing happened later. For example, if
549 (?(?=(a))a) was matched against "a", no capturing was returned.
550
551 26. When studying a pattern that contained subroutine calls or assertions,
552 the code for finding the minimum length of a possible match was handling
553 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
554 group 1 called group 2 while simultaneously a separate group 2 called group
555 1). A stack overflow occurred in this case. I have fixed this by limiting
556 the recursion depth to 10.
557
558 27. Updated RunTest.bat in the distribution to the version supplied by Tom
559 Fortmann. This supports explicit test numbers on the command line, and has
560 argument validation and error reporting.
561
562 28. An instance of \X with an unlimited repeat could fail if at any point the
563 first character it looked at was a mark character.
564
565 29. Some minor code refactoring concerning Unicode properties and scripts
566 should reduce the stack requirement of match() slightly.
567
568 30. Added the '=' option to pcretest to check the setting of unused capturing
569 slots at the end of the pattern, which are documented as being -1, but are
570 not included in the return count.
571
572 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
573 compiled something random. Now it gives a compile-time error (as does
574 Perl).
575
576 32. A *MARK encountered during the processing of a positive assertion is now
577 recorded and passed back (compatible with Perl).
578
579 33. If --only-matching or --colour was set on a pcregrep call whose pattern
580 had alternative anchored branches, the search for a second match in a line
581 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
582 matched the line "0102" twice. The same bug affected patterns that started
583 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
584 twice.
585
586 34. Previously, PCRE did not allow quantification of assertions. However, Perl
587 does, and because of capturing effects, quantifying parenthesized
588 assertions may at times be useful. Quantifiers are now allowed for
589 parenthesized assertions.
590
591 35. A minor code tidy in pcre_compile() when checking options for \R usage.
592
593 36. \g was being checked for fancy things in a character class, when it should
594 just be a literal "g".
595
596 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
597 appearance of a nested POSIX class supersedes an apparent external class.
598 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
599 unescaped square brackets may also appear as part of class names. For
600 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
601 more like Perl. (But see 8.20/1 above.)
602
603 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
604 was because it thought it was \N{name}, which is not supported).
605
606 39. Add minix to OS list not supporting the -S option in pcretest.
607
608 40. PCRE tries to detect cases of infinite recursion at compile time, but it
609 cannot analyze patterns in sufficient detail to catch mutual recursions
610 such as ((?1))((?2)). There is now a runtime test that gives an error if a
611 subgroup is called recursively as a subpattern for a second time at the
612 same position in the subject string. In previous releases this might have
613 been caught by the recursion limit, or it might have run out of stack.
614
615 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
616 happen only once. PCRE was, however incorrectly giving a compile time error
617 "recursive call could loop indefinitely" because it cannot analyze the
618 pattern in sufficient detail. The compile time test no longer happens when
619 PCRE is compiling a conditional subpattern, but actual runaway loops are
620 now caught at runtime (see 40 above).
621
622 42. It seems that Perl allows any characters other than a closing parenthesis
623 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
624 has been changed to be the same.
625
626 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
627 as not to get warnings when autogen.sh is called. Also changed
628 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
629
630 44. To help people who use pcregrep to scan files containing exceedingly long
631 lines, the following changes have been made:
632
633 (a) The default value of the buffer size parameter has been increased from
634 8K to 20K. (The actual buffer used is three times this size.)
635
636 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
637 PCRE is built.
638
639 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
640 to be set at run time.
641
642 (d) Numerical values in pcregrep options can be followed by K or M, for
643 example --buffer-size=50K.
644
645 (e) If a line being scanned overflows pcregrep's buffer, an error is now
646 given and the return code is set to 2.
647
648 45. Add a pointer to the latest mark to the callout data block.
649
650 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
651 partial match of an empty string instead of no match. This was specific to
652 the use of ".".
653
654 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
655 complete match instead of a partial match. This bug was dependent on both
656 the PCRE_UTF8 and PCRE_DOTALL options being set.
657
658 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
659 starting byte set, because \b was not being ignored.
660
661
662 Version 8.12 15-Jan-2011
663 ------------------------
664
665 1. Fixed some typos in the markup of the man pages, and wrote a script that
666 checks for such things as part of the documentation building process.
667
668 2. On a big-endian 64-bit system, pcregrep did not correctly process the
669 --match-limit and --recursion-limit options (added for 8.11). In
670 particular, this made one of the standard tests fail. (The integer value
671 went into the wrong half of a long int.)
672
673 3. If the --colour option was given to pcregrep with -v (invert match), it
674 did strange things, either producing crazy output, or crashing. It should,
675 of course, ignore a request for colour when reporting lines that do not
676 match.
677
678 4. Another pcregrep bug caused similar problems if --colour was specified with
679 -M (multiline) and the pattern match finished with a line ending.
680
681 5. In pcregrep, when a pattern that ended with a literal newline sequence was
682 matched in multiline mode, the following line was shown as part of the
683 match. This seems wrong, so I have changed it.
684
685 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
686 the check for further matches in the same line (so they could be coloured)
687 to overrun the end of the current line. If another match was found, it was
688 incorrectly shown (and then shown again when found in the next line).
689
690 7. If pcregrep was compiled under Windows, there was a reference to the
691 function pcregrep_exit() before it was defined. I am assuming this was
692 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
693 reported by a user. I've moved the definition above the reference.
694
695
696 Version 8.11 10-Dec-2010
697 ------------------------
698
699 1. (*THEN) was not working properly if there were untried alternatives prior
700 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
701 backtracked to try for "b" instead of moving to the next alternative branch
702 at the same level (in this case, to look for "c"). The Perl documentation
703 is clear that when (*THEN) is backtracked onto, it goes to the "next
704 alternative in the innermost enclosing group".
705
706 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
707 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
708 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
709 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
710 (*THEN).
711
712 3. If \s appeared in a character class, it removed the VT character from
713 the class, even if it had been included by some previous item, for example
714 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
715 of \s, but is part of the POSIX "space" class.)
716
717 4. A partial match never returns an empty string (because you can always
718 match an empty string at the end of the subject); however the checking for
719 an empty string was starting at the "start of match" point. This has been
720 changed to the "earliest inspected character" point, because the returned
721 data for a partial match starts at this character. This means that, for
722 example, /(?<=abc)def/ gives a partial match for the subject "abc"
723 (previously it gave "no match").
724
725 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
726 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
727 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
728 has an implication that the given string is incomplete (because a partial
729 match is preferred over a full match). For this reason, these items now
730 give a partial match in this situation. [Aside: previously, the one case
731 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
732 match rather than a full match, which was wrong by the old rules, but is
733 now correct.]
734
735 6. There was a bug in the handling of #-introduced comments, recognized when
736 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
737 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
738 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
739 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
740 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
741 places in pcre_compile().
742
743 7. Related to (6) above, when pcre_compile() was skipping #-introduced
744 comments when looking ahead for named forward references to subpatterns,
745 the only newline sequence it recognized was NL. It now handles newlines
746 according to the set newline convention.
747
748 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
749 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
750 cater for a lack of strerror(). These oversights have been fixed.
751
752 9. Added --match-limit and --recursion-limit to pcregrep.
753
754 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
755
756 11. When the -o option was used, pcregrep was setting a return code of 1, even
757 when matches were found, and --line-buffered was not being honoured.
758
759 12. Added an optional parentheses number to the -o and --only-matching options
760 of pcregrep.
761
762 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
763 can match an empty string. The code to do it in pcretest and pcredemo
764 needed fixing:
765
766 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
767 only one byte after an empty string match just before CRLF (this case
768 just got forgotten; "any" and "anycrlf" were OK).
769
770 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
771 mode when an empty string match preceded an ASCII character followed by
772 a non-ASCII character. (The code for advancing by one character rather
773 than one byte was nonsense.)
774
775 (c) The pcredemo.c sample program did not have any code at all to handle
776 the cases when CRLF is a valid newline sequence.
777
778 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
779 as a starting offset was within the subject string. There is now a new
780 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
781 negative or greater than the length of the string. In order to test this,
782 pcretest is extended to allow the setting of negative starting offsets.
783
784 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
785 starting offset points to the beginning of a UTF-8 character was
786 unnecessarily clumsy. I tidied it up.
787
788 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
789 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
790
791 17. Nobody had reported that the --include_dir option, which was added in
792 release 7.7 should have been called --include-dir (hyphen, not underscore)
793 for compatibility with GNU grep. I have changed it to --include-dir, but
794 left --include_dir as an undocumented synonym, and the same for
795 --exclude-dir, though that is not available in GNU grep, at least as of
796 release 2.5.4.
797
798 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
799 characters from a string of bytes) have been redefined so as not to use
800 loops, in order to improve performance in some environments. At the same
801 time, I abstracted some of the common code into auxiliary macros to save
802 repetition (this should not affect the compiled code).
803
804 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
805 compile-time error is now given if \c is not followed by an ASCII
806 character, that is, a byte less than 128. (In EBCDIC mode, the code is
807 different, and any byte value is allowed.)
808
809 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
810 START_OPTIMIZE option, which is now allowed at compile time - but just
811 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
812 to pcregrep and other applications that have no direct access to PCRE
813 options. The new /Y option in pcretest sets this option when calling
814 pcre_compile().
815
816 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
817 back references. Groups containing recursive back references were forced to
818 be atomic by that change, but in the case of named groups, the amount of
819 memory required was incorrectly computed, leading to "Failed: internal
820 error: code overflow". This has been fixed.
821
822 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
823 pcretest.c, to avoid build problems in some Borland environments.
824
825
826 Version 8.10 25-Jun-2010
827 ------------------------
828
829 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
830 THEN.
831
832 2. (*ACCEPT) was not working when inside an atomic group.
833
834 3. Inside a character class, \B is treated as a literal by default, but
835 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
836 causes the error). The code is unchanged, but I tidied the documentation.
837
838 4. Inside a character class, PCRE always treated \R and \X as literals,
839 whereas Perl faults them if its -w option is set. I have changed PCRE so
840 that it faults them when PCRE_EXTRA is set.
841
842 5. Added support for \N, which always matches any character other than
843 newline. (It is the same as "." when PCRE_DOTALL is not set.)
844
845 6. When compiling pcregrep with newer versions of gcc which may have
846 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
847 declared with attribute warn_unused_result" were given. Just casting the
848 result to (void) does not stop the warnings; a more elaborate fudge is
849 needed. I've used a macro to implement this.
850
851 7. Minor change to pcretest.c to avoid a compiler warning.
852
853 8. Added four artifical Unicode properties to help with an option to make
854 \s etc use properties (see next item). The new properties are: Xan
855 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
856
857 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
858 use Unicode properties. (*UCP) at the start of a pattern can be used to set
859 this option. Modified pcretest to add /W to test this facility. Added
860 REG_UCP to make it available via the POSIX interface.
861
862 10. Added --line-buffered to pcregrep.
863
864 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
865 studied, and the match started with a letter with a code point greater than
866 127 whose first byte was different to the first byte of the other case of
867 the letter, the other case of this starting letter was not recognized
868 (#976).
869
870 12. If a pattern that was studied started with a repeated Unicode property
871 test, for example, \p{Nd}+, there was the theoretical possibility of
872 setting up an incorrect bitmap of starting bytes, but fortunately it could
873 not have actually happened in practice until change 8 above was made (it
874 added property types that matched character-matching opcodes).
875
876 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
877 possible starting bytes for non-anchored patterns.
878
879 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
880 \R, and also a number of cases that involve Unicode properties, both
881 explicit and implicit when PCRE_UCP is set.
882
883 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
884 input, it could crash or give wrong results if characters with values
885 greater than 0xc0 were present in the subject string. (Detail: it assumed
886 UTF-8 input when processing these items.)
887
888 16. Added a lot of (int) casts to avoid compiler warnings in systems where
889 size_t is 64-bit (#991).
890
891 17. Added a check for running out of memory when PCRE is compiled with
892 --disable-stack-for-recursion (#990).
893
894 18. If the last data line in a file for pcretest does not have a newline on
895 the end, a newline was missing in the output.
896
897 19. The default pcre_chartables.c file recognizes only ASCII characters (values
898 less than 128) in its various bitmaps. However, there is a facility for
899 generating tables according to the current locale when PCRE is compiled. It
900 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
901 characters, are recognized by isspace() and therefore were getting set in
902 these tables, and indeed these tables seem to approximate to ISO 8859. This
903 caused a problem in UTF-8 mode when pcre_study() was used to create a list
904 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
905 which of course cannot start UTF-8 characters. I have changed the code so
906 that only real ASCII characters (less than 128) and the correct starting
907 bytes for UTF-8 encodings are set for characters greater than 127 when in
908 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
909 altogether.)
910
911 20. Added the /T option to pcretest so as to be able to run tests with non-
912 standard character tables, thus making it possible to include the tests
913 used for 19 above in the standard set of tests.
914
915 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
916 reference to a subpattern the other side of a comment that contains an
917 opening parenthesis caused either an internal compiling error, or a
918 reference to the wrong subpattern.
919
920
921 Version 8.02 19-Mar-2010
922 ------------------------
923
924 1. The Unicode data tables have been updated to Unicode 5.2.0.
925
926 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
927 configured.
928
929 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
930 original author of that file, following a query about its status.
931
932 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
933 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
934
935 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
936 quantifier applied to a forward-referencing subroutine call, could compile
937 incorrect code or give the error "internal error: previously-checked
938 referenced subpattern not found".
939
940 6. Both MS Visual Studio and Symbian OS have problems with initializing
941 variables to point to external functions. For these systems, therefore,
942 pcre_malloc etc. are now initialized to local functions that call the
943 relevant global functions.
944
945 7. There were two entries missing in the vectors called coptable and poptable
946 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
947 I've fixed the data, and added a kludgy way of testing at compile time that
948 the lengths are correct (equal to the number of opcodes).
949
950 8. Following on from 7, I added a similar kludge to check the length of the
951 eint vector in pcreposix.c.
952
953 9. Error texts for pcre_compile() are held as one long string to avoid too
954 much relocation at load time. To find a text, the string is searched,
955 counting zeros. There was no check for running off the end of the string,
956 which could happen if a new error number was added without updating the
957 string.
958
959 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
960
961 11. \K was not working if it appeared in an atomic group or in a group that
962 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
963 \K is "not well defined" if used in an assertion. PCRE now accepts it if
964 the assertion is positive, but not if it is negative.
965
966 12. Change 11 fortuitously reduced the size of the stack frame used in the
967 "match()" function of pcre_exec.c by one pointer. Forthcoming
968 implementation of support for (*MARK) will need an extra pointer on the
969 stack; I have reserved it now, so that the stack frame size does not
970 decrease.
971
972 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
973 item in branch that calls a recursion is a subroutine call - as in the
974 second branch in the above example - was incorrectly given the compile-
975 time error "recursive call could loop indefinitely" because pcre_compile()
976 was not correctly checking the subroutine for matching a non-empty string.
977
978 14. The checks for overrunning compiling workspace could trigger after an
979 overrun had occurred. This is a "should never occur" error, but it can be
980 triggered by pathological patterns such as hundreds of nested parentheses.
981 The checks now trigger 100 bytes before the end of the workspace.
982
983 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
984
985
986 Version 8.01 19-Jan-2010
987 ------------------------
988
989 1. If a pattern contained a conditional subpattern with only one branch (in
990 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
991 computed the wrong minimum data length (which is of course zero for such
992 subpatterns). This could cause incorrect "no match" results.
993
994 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
995 the pattern is reset in the first branch, pcre_compile() failed with
996 "internal error: code overflow at offset...". This happened only when
997 the reset was to the original external option setting. (An optimization
998 abstracts leading options settings into an external setting, which was the
999 cause of this.)
1000
1001 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1002 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1003 assertion pattern did not match (meaning that the assertion was true), it
1004 was incorrectly treated as false if the SKIP had been reached during the
1005 matching. This also applied to assertions used as conditions.
1006
1007 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
1008 assertion subpattern, including such a pattern used as a condition,
1009 unpredictable results occurred, instead of the error return
1010 PCRE_ERROR_DFA_UITEM.
1011
1012 5. The C++ GlobalReplace function was not working like Perl for the special
1013 situation when an empty string is matched. It now does the fancy magic
1014 stuff that is necessary.
1015
1016 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1017 removed. (These were left over from very, very early versions of PCRE.)
1018
1019 7. Some cosmetic changes to the code to make life easier when compiling it
1020 as part of something else:
1021
1022 (a) Change DEBUG to PCRE_DEBUG.
1023
1024 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1025 called "current" as "current_branch", to prevent a collision with the
1026 Linux macro when compiled as a kernel module.
1027
1028 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1029 prevent a collision with the Linux macro when compiled as a kernel
1030 module.
1031
1032 8. In pcre_compile() there are some checks for integer overflows that used to
1033 cast potentially large values to (double). This has been changed to that
1034 when building, a check for int64_t is made, and if it is found, it is used
1035 instead, thus avoiding the use of floating point arithmetic. (There is no
1036 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1037 double.
1038
1039 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1040 2005 (difference between two addresses compared to an unsigned value).
1041
1042 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1043 custom one, because of the following reported problem in Windows:
1044
1045 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1046 under Win32.
1047 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1048 therefore missing the function definition.
1049 - The compiler thus generates a "C" signature for the test function.
1050 - The linker fails to find the "C" function.
1051 - PCRE fails to configure if asked to do so against libbz2.
1052
1053 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1054 messages were output:
1055
1056 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1057 rerunning libtoolize, to keep the correct libtool macros in-tree.
1058 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1059
1060 I have done both of these things.
1061
1062 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1063 most of the time, it *can* run out if it is given a pattern that contains a
1064 runaway infinite recursion. I updated the discussion in the pcrestack man
1065 page.
1066
1067 13. Now that we have gone to the x.xx style of version numbers, the minor
1068 version may start with zero. Using 08 or 09 is a bad idea because users
1069 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1070 interpreted as invalid octal numbers. I've updated the previous comment in
1071 configure.ac, and also added a check that gives an error if 08 or 09 are
1072 used.
1073
1074 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1075 causing partial matching to fail when the end of the subject matched \W
1076 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1077
1078 15. There were some discrepancies between the declarations in pcre_internal.h
1079 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1080 their definitions. The declarations used "const uschar *" and the
1081 definitions used USPTR. Even though USPTR is normally defined as "const
1082 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1083 reported that: "This difference in casting confuses some C++ compilers, for
1084 example, SunCC recognizes above declarations as different functions and
1085 generates broken code for hbpcre." I have changed the declarations to use
1086 USPTR.
1087
1088 16. GNU libtool is named differently on some systems. The autogen.sh script now
1089 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1090 (FreeBSD).
1091
1092 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1093 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1094 comment: "Figure out how to create a longlong from a string: strtoll and
1095 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1096 instance, but it only takes 2 args instead of 3!"
1097
1098 18. A subtle bug concerned with back references has been fixed by a change of
1099 specification, with a corresponding code fix. A pattern such as
1100 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1101 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1102 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1103 same bug. Such groups have to be quantified to be useful, or contained
1104 inside another quantified group. (If there's no repetition, the reference
1105 can never match.) The problem arises because, having left the group and
1106 moved on to the rest of the pattern, a later failure that backtracks into
1107 the group uses the captured value from the final iteration of the group
1108 rather than the correct earlier one. I have fixed this in PCRE by forcing
1109 any group that contains a reference to itself to be an atomic group; that
1110 is, there cannot be any backtracking into it once it has completed. This is
1111 similar to recursive and subroutine calls.
1112
1113
1114 Version 8.00 19-Oct-09
1115 ----------------------
1116
1117 1. The table for translating pcre_compile() error codes into POSIX error codes
1118 was out-of-date, and there was no check on the pcre_compile() error code
1119 being within the table. This could lead to an OK return being given in
1120 error.
1121
1122 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1123 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1124 in a Windows environment.
1125
1126 3. The pcregrep --count option prints the count for each file even when it is
1127 zero, as does GNU grep. However, pcregrep was also printing all files when
1128 --files-with-matches was added. Now, when both options are given, it prints
1129 counts only for those files that have at least one match. (GNU grep just
1130 prints the file name in this circumstance, but including the count seems
1131 more useful - otherwise, why use --count?) Also ensured that the
1132 combination -clh just lists non-zero counts, with no names.
1133
1134 4. The long form of the pcregrep -F option was incorrectly implemented as
1135 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1136 but it seems right to fix it, and I didn't think it was worth preserving
1137 the old behaviour.
1138
1139 5. The command line items --regex=pattern and --regexp=pattern were not
1140 recognized by pcregrep, which required --regex pattern or --regexp pattern
1141 (with a space rather than an '='). The man page documented the '=' forms,
1142 which are compatible with GNU grep; these now work.
1143
1144 6. No libpcreposix.pc file was created for pkg-config; there was just
1145 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1146
1147 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1148 when UCP support is not needed, by modifying the Python script that
1149 generates it from Unicode data files. This should not matter if the module
1150 is correctly used as a library, but I received one complaint about 50K of
1151 unwanted data. My guess is that the person linked everything into his
1152 program rather than using a library. Anyway, it does no harm.
1153
1154 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1155 was a minimum greater than 1 for a wide character in a possessive
1156 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1157 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1158 character. Chaos in the form of incorrect output or a compiling loop could
1159 result.
1160
1161 9. The restrictions on what a pattern can contain when partial matching is
1162 requested for pcre_exec() have been removed. All patterns can now be
1163 partially matched by this function. In addition, if there are at least two
1164 slots in the offset vector, the offset of the earliest inspected character
1165 for the match and the offset of the end of the subject are set in them when
1166 PCRE_ERROR_PARTIAL is returned.
1167
1168 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1169 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1170 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1171 and may be more useful for multi-segment matching.
1172
1173 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1174 used to be given if ever the end of the subject was reached; now it is
1175 given only if matching could not proceed because another character was
1176 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1177 string "Z", which now yields "no match" instead of "partial match". In the
1178 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1179 final character ended with (*FAIL).
1180
1181 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1182 if the pattern had a "must contain" character that was already found in the
1183 earlier partial match, unless partial matching was again requested. For
1184 example, with the pattern /dog.(body)?/, the "must contain" character is
1185 "g". If the first part-match was for the string "dog", restarting with
1186 "sbody" failed. This bug has been fixed.
1187
1188 13. The string returned by pcre_dfa_exec() after a partial match has been
1189 changed so that it starts at the first inspected character rather than the
1190 first character of the match. This makes a difference only if the pattern
1191 starts with a lookbehind assertion or \b or \B (\K is not supported by
1192 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1193 matching functions compatible, and I think it's the right thing to do.
1194
1195 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1196 so that the demonstration program is easily available in environments where
1197 PCRE has not been installed from source.
1198
1199 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1200 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1201 library.
1202
1203 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1204 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1205 is not the first non-POSIX option to be added. Clearly some people find
1206 these options useful.
1207
1208 17. If a caller to the POSIX matching function regexec() passes a non-zero
1209 value for nmatch with a NULL value for pmatch, the value of
1210 nmatch is forced to zero.
1211
1212 18. RunGrepTest did not have a test for the availability of the -u option of
1213 the diff command, as RunTest does. It now checks in the same way as
1214 RunTest, and also checks for the -b option.
1215
1216 19. If an odd number of negated classes containing just a single character
1217 interposed, within parentheses, between a forward reference to a named
1218 subpattern and the definition of the subpattern, compilation crashed with
1219 an internal error, complaining that it could not find the referenced
1220 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1221 [The bug was that it was starting one character too far in when skipping
1222 over the character class, thus treating the ] as data rather than
1223 terminating the class. This meant it could skip too much.]
1224
1225 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1226 /g option in pcretest when the pattern contains \K, which makes it possible
1227 to have an empty string match not at the start, even when the pattern is
1228 anchored. Updated pcretest and pcredemo to use this option.
1229
1230 21. If the maximum number of capturing subpatterns in a recursion was greater
1231 than the maximum at the outer level, the higher number was returned, but
1232 with unset values at the outer level. The correct (outer level) value is
1233 now given.
1234
1235 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1236 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1237 make it do so. The string so far is captured, making this feature
1238 compatible with Perl.
1239
1240 23. The tests have been re-organized, adding tests 11 and 12, to make it
1241 possible to check the Perl 5.10 features against Perl 5.10.
1242
1243 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1244 pattern matches a fixed length string. PCRE did not allow this; now it
1245 does. Neither allows recursion.
1246
1247 25. I finally figured out how to implement a request to provide the minimum
1248 length of subject string that was needed in order to match a given pattern.
1249 (It was back references and recursion that I had previously got hung up
1250 on.) This code has now been added to pcre_study(); it finds a lower bound
1251 to the length of subject needed. It is not necessarily the greatest lower
1252 bound, but using it to avoid searching strings that are too short does give
1253 some useful speed-ups. The value is available to calling programs via
1254 pcre_fullinfo().
1255
1256 26. While implementing 25, I discovered to my embarrassment that pcretest had
1257 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1258 study optimizations had never been tested with that matching function.
1259 Oops. What is worse, even when it was passed study data, there was a bug in
1260 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1261 were also very few tests of studied patterns with pcre_dfa_exec().
1262
1263 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1264 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1265 on the other side of the coin, they are no longer allowed to have different
1266 names, because these cannot be distinguished in PCRE, and this has caused
1267 confusion. (This is a difference from Perl.)
1268
1269 28. When duplicate subpattern names are present (necessarily with different
1270 numbers, as required by 27 above), and a test is made by name in a
1271 conditional pattern, either for a subpattern having been matched, or for
1272 recursion in such a pattern, all the associated numbered subpatterns are
1273 tested, and the overall condition is true if the condition is true for any
1274 one of them. This is the way Perl works, and is also more like the way
1275 testing by number works.
1276
1277
1278 Version 7.9 11-Apr-09
1279 ---------------------
1280
1281 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1282 (pcretest), all targets were linked against these libraries. This included
1283 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1284 libraries. This caused unwanted dependencies to be created. This problem
1285 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1286 pcretest is linked with readline.
1287
1288 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1289 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1290 moved outside it again, because FALSE and TRUE are already defined in AIX,
1291 but BOOL is not.
1292
1293 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1294 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1295
1296 4. The pcregrep documentation said spaces were inserted as well as colons (or
1297 hyphens) following file names and line numbers when outputting matching
1298 lines. This is not true; no spaces are inserted. I have also clarified the
1299 wording for the --colour (or --color) option.
1300
1301 5. In pcregrep, when --colour was used with -o, the list of matching strings
1302 was not coloured; this is different to GNU grep, so I have changed it to be
1303 the same.
1304
1305 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1306 each matching line was coloured. Now it goes on to look for further matches
1307 of any of the test patterns, which is the same behaviour as GNU grep.
1308
1309 7. A pattern that could match an empty string could cause pcregrep to loop; it
1310 doesn't make sense to accept an empty string match in pcregrep, so I have
1311 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1312 seems to be how GNU grep behaves.
1313
1314 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1315 start or after a newline", because the conditional assertion was not being
1316 correctly handled. The rule now is that both the assertion and what follows
1317 in the first alternative must satisfy the test.
1318
1319 9. If auto-callout was enabled in a pattern with a conditional group whose
1320 condition was an assertion, PCRE could crash during matching, both with
1321 pcre_exec() and pcre_dfa_exec().
1322
1323 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1324 used for matching.
1325
1326 11. Unicode property support in character classes was not working for
1327 characters (bytes) greater than 127 when not in UTF-8 mode.
1328
1329 12. Added the -M command line option to pcretest.
1330
1331 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1332
1333 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1334
1335 16. Added comments and documentation about mis-use of no_arg in the C++
1336 wrapper.
1337
1338 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1339 from Martin Jerabek that uses macro names for all relevant character and
1340 string constants.
1341
1342 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1343 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1344 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1345 these, but not everybody uses configure.
1346
1347 19. A conditional group that had only one branch was not being correctly
1348 recognized as an item that could match an empty string. This meant that an
1349 enclosing group might also not be so recognized, causing infinite looping
1350 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1351 with the subject "ab", where knowledge that the repeated group can match
1352 nothing is needed in order to break the loop.
1353
1354 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1355 exec(), but without supplying a callout function, matching went wrong.
1356
1357 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1358 leak if the size of the offset vector was greater than 30. When the vector
1359 is smaller, the saved offsets during recursion go onto a local stack
1360 vector, but for larger vectors malloc() is used. It was failing to free
1361 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1362 error, in fact).
1363
1364 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1365 heapframe that is used only when UTF-8 support is enabled. This caused no
1366 problem, but was untidy.
1367
1368 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1369 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1370 included within another project.
1371
1372 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1373 slightly modified by me:
1374
1375 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1376 not building pcregrep.
1377
1378 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1379 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1380
1381 25. Forward references, both numeric and by name, in patterns that made use of
1382 duplicate group numbers, could behave incorrectly or give incorrect errors,
1383 because when scanning forward to find the reference group, PCRE was not
1384 taking into account the duplicate group numbers. A pattern such as
1385 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1386
1387 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1388 the feature of a custom pointer more persuasive (as requested by a user).
1389
1390 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1391 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1392 pre-defined.
1393
1394 28. Added support for (*UTF8) at the start of a pattern.
1395
1396 29. Arrange for flags added by the "release type" setting in CMake to be shown
1397 in the configuration summary.
1398
1399
1400 Version 7.8 05-Sep-08
1401 ---------------------
1402
1403 1. Replaced UCP searching code with optimized version as implemented for Ad
1404 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1405 stage table and inline lookup instead of a function, giving speed ups of 2
1406 to 5 times on some simple patterns that I tested. Permission was given to
1407 distribute the MultiStage2.py script that generates the tables (it's not in
1408 the tarball, but is in the Subversion repository).
1409
1410 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1411 scripts.
1412
1413 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1414 a group with a zero qualifier. The result of the study could be incorrect,
1415 or the function might crash, depending on the pattern.
1416
1417 4. Caseless matching was not working for non-ASCII characters in back
1418 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1419 It now works when Unicode Property Support is available.
1420
1421 5. In pcretest, an escape such as \x{de} in the data was always generating
1422 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1423 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1424 truncation.
1425
1426 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1427
1428 7. Added two (int) casts to pcregrep when printing the difference of two
1429 pointers, in case they are 64-bit values.
1430
1431 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1432 test 2 if it fails.
1433
1434 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1435 and a #define of that name to empty if it is not externally set. This is to
1436 allow users of MSVC to set it if necessary.
1437
1438 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1439 the convenience functions in the pcre_get.c source file.
1440
1441 11. An option change at the start of a pattern that had top-level alternatives
1442 could cause overwriting and/or a crash. This command provoked a crash in
1443 some environments:
1444
1445 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1446
1447 This potential security problem was recorded as CVE-2008-2371.
1448
1449 12. For a pattern where the match had to start at the beginning or immediately
1450 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1451 pcre_dfa_exec() could read past the end of the passed subject if there was
1452 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1453 pcretest so that it places the subject at the end of its malloc-ed buffer.
1454
1455 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1456 exec() might read past the end of the data buffer in UTF-8 mode.
1457
1458 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1459 the data contained the byte 0x85 as part of a UTF-8 character within its
1460 first line. This applied both to normal and DFA matching.
1461
1462 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1463 /^[^d]*?$/8 failed to match "abc".
1464
1465 16. Added a missing copyright notice to pcrecpp_internal.h.
1466
1467 17. Make it more clear in the documentation that values returned from
1468 pcre_exec() in ovector are byte offsets, not character counts.
1469
1470 18. Tidied a few places to stop certain compilers from issuing warnings.
1471
1472 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1473 supplied by Stefan Weber. I made a further small update for 7.8 because
1474 there is a change of source arrangements: the pcre_searchfuncs.c module is
1475 replaced by pcre_ucd.c.
1476
1477
1478 Version 7.7 07-May-08
1479 ---------------------
1480
1481 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1482 a string to a long long, pretend we don't even have a long long." This is
1483 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1484
1485 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1486 pre-7.6 versions, which defined a global no_arg variable instead of putting
1487 it in the RE class. (See also #8 below.)
1488
1489 3. Remove a line of dead code, identified by coverity and reported by Nuno
1490 Lopes.
1491
1492 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1493
1494 (1) The include/exclude patterns were being applied to the whole pathnames
1495 of files, instead of just to the final components.
1496
1497 (2) If there was more than one level of directory, the subdirectories were
1498 skipped unless they satisfied the include/exclude conditions. This is
1499 inconsistent with GNU grep (and could even be seen as contrary to the
1500 pcregrep specification - which I improved to make it absolutely clear).
1501 The action now is always to scan all levels of directory, and just
1502 apply the include/exclude patterns to regular files.
1503
1504 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1505 --exclude_dir in the tests to avoid scanning .svn directories.
1506
1507 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1508 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1509 doesn't support NULs in patterns.
1510
1511 7. Added some missing "const"s to declarations of static tables in
1512 pcre_compile.c and pcre_dfa_exec.c.
1513
1514 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1515 caused by fix #2 above. (Subsequently also a second patch to fix the
1516 first patch. And a third patch - this was a messy problem.)
1517
1518 9. Applied Craig's patch to remove the use of push_back().
1519
1520 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1521 matching function regexec().
1522
1523 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1524 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1525 references. PCRE supports relative numbers with this syntax (I don't think
1526 Oniguruma does).
1527
1528 12. Previously, a group with a zero repeat such as (...){0} was completely
1529 omitted from the compiled regex. However, this means that if the group
1530 was called as a subroutine from elsewhere in the pattern, things went wrong
1531 (an internal error was given). Such groups are now left in the compiled
1532 pattern, with a new opcode that causes them to be skipped at execution
1533 time.
1534
1535 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1536 to the way PCRE behaves:
1537
1538 (a) A lone ] character is dis-allowed (Perl treats it as data).
1539
1540 (b) A back reference to an unmatched subpattern matches an empty string
1541 (Perl fails the current match path).
1542
1543 (c) A data ] in a character class must be notated as \] because if the
1544 first data character in a class is ], it defines an empty class. (In
1545 Perl it is not possible to have an empty class.) The empty class []
1546 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1547 The negative empty class [^] matches any one character, independently
1548 of the DOTALL setting.
1549
1550 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1551 non-existent subpattern following a character class starting with ']' and
1552 containing () gave an internal compiling error instead of "reference to
1553 non-existent subpattern". Fortunately, when the pattern did exist, the
1554 compiled code was correct. (When scanning forwards to check for the
1555 existencd of the subpattern, it was treating the data ']' as terminating
1556 the class, so got the count wrong. When actually compiling, the reference
1557 was subsequently set up correctly.)
1558
1559 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1560 it was being rejected as not supported by pcre_dfa_exec(), even though
1561 other assertions are supported. I have made pcre_dfa_exec() support
1562 (*FAIL).
1563
1564 16. The implementation of 13c above involved the invention of a new opcode,
1565 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1566 cannot be changed at match time, I realized I could make a small
1567 improvement to matching performance by compiling OP_ALLANY instead of
1568 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1569 on the OP_ANY path.
1570
1571 17. Compiling pcretest on Windows with readline support failed without the
1572 following two fixes: (1) Make the unistd.h include conditional on
1573 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1574
1575 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1576 ncurses library to be included for pcretest when ReadLine support is
1577 requested, but also to allow for it to be overridden. This patch came from
1578 Daniel Bergström.
1579
1580 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1581 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1582 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1583 spotting this.
1584
1585
1586 Version 7.6 28-Jan-08
1587 ---------------------
1588
1589 1. A character class containing a very large number of characters with
1590 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1591 overflow.
1592
1593 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1594 HAVE_LONG_LONG is not defined.
1595
1596 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1597 bring them up to date and include new features. This patch includes:
1598
1599 - Fixed PH's badly added libz and libbz2 support.
1600 - Fixed a problem with static linking.
1601 - Added pcredemo. [But later removed - see 7 below.]
1602 - Fixed dftables problem and added an option.
1603 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1604 HAVE_LONG_LONG.
1605 - Added readline support for pcretest.
1606 - Added an listing of the option settings after cmake has run.
1607
1608 4. A user submitted a patch to Makefile that makes it easy to create
1609 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1610 Makefile.am that cause it to include this special target, without
1611 affecting anything else. Note that the same mingw target plus all
1612 the other distribution libraries and programs are now supported
1613 when configuring with CMake (see 6 below) instead of with
1614 Configure/Make.
1615
1616 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1617 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1618 exported in the Windows port". It has not yet been confirmed that the patch
1619 solves the problem, but it does no harm.
1620
1621 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1622 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1623 with CMake, and also correct the comment about stack recursion.
1624
1625 7. Remove the automatic building of pcredemo from the ./configure system and
1626 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1627 of a program that users should build themselves after PCRE is installed, so
1628 building it automatically is not really right. What is more, it gave
1629 trouble in some build environments.
1630
1631 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1632
1633
1634 Version 7.5 10-Jan-08
1635 ---------------------
1636
1637 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1638 values in parens when parsing an RE using the C++ wrapper."
1639
1640 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1641 Characters greater than 255 were excluded from the class instead of being
1642 included.
1643
1644 3. The same bug as (2) above applied to negated POSIX classes such as
1645 [:^space:].
1646
1647 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1648 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1649 I have changed it.
1650
1651 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1652 first named subpattern) and a construct such as (?&a) would reference the
1653 first named subpattern whose name started with "a" (in other words, the
1654 length check was missing). Both these problems are fixed. "Subpattern name
1655 expected" is now given for (?&) (a zero-length name), and this patch also
1656 makes it give the same error for \k'' (previously it complained that that
1657 was a reference to a non-existent subpattern).
1658
1659 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1660 this is right because (?- can be followed by option settings as well as by
1661 digits. I have, however, made the messages clearer.
1662
1663 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1664 than the number used in the conditional) now cause a compile-time error.
1665 This is actually not compatible with Perl, which accepts such patterns, but
1666 treats the conditional as always being FALSE (as PCRE used to), but it
1667 seems to me that giving a diagnostic is better.
1668
1669 8. Change "alphameric" to the more common word "alphanumeric" in comments
1670 and messages.
1671
1672 9. Fix two occurrences of "backslash" in comments that should have been
1673 "backspace".
1674
1675 10. Remove two redundant lines of code that can never be obeyed (their function
1676 was moved elsewhere).
1677
1678 11. The program that makes PCRE's Unicode character property table had a bug
1679 which caused it to generate incorrect table entries for sequences of
1680 characters that have the same character type, but are in different scripts.
1681 It amalgamated them into a single range, with the script of the first of
1682 them. In other words, some characters were in the wrong script. There were
1683 thirteen such cases, affecting characters in the following ranges:
1684
1685 U+002b0 - U+002c1
1686 U+0060c - U+0060d
1687 U+0061e - U+00612
1688 U+0064b - U+0065e
1689 U+0074d - U+0076d
1690 U+01800 - U+01805
1691 U+01d00 - U+01d77
1692 U+01d9b - U+01dbf
1693 U+0200b - U+0200f
1694 U+030fc - U+030fe
1695 U+03260 - U+0327f
1696 U+0fb46 - U+0fbb1
1697 U+10450 - U+1049d
1698
1699 12. The -o option (show only the matching part of a line) for pcregrep was not
1700 compatible with GNU grep in that, if there was more than one match in a
1701 line, it showed only the first of them. It now behaves in the same way as
1702 GNU grep.
1703
1704 13. If the -o and -v options were combined for pcregrep, it printed a blank
1705 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1706 does the same. The return code can be used to tell if there were any
1707 non-matching lines.
1708
1709 14. Added --file-offsets and --line-offsets to pcregrep.
1710
1711 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1712 infinitely looping recursion. The bug was that positive lookaheads were not
1713 being skipped when checking for a possible empty match (negative lookaheads
1714 and both kinds of lookbehind were skipped).
1715
1716 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1717 inclusion of <windows.h> to before rather than after the definition of
1718 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1719
1720 17. Specifying a possessive quantifier with a specific limit for a Unicode
1721 character property caused pcre_compile() to compile bad code, which led at
1722 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1723 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1724 caused the error; without that there was no problem.
1725
1726 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1727
1728 19. Added --enable-pcretest-libreadline.
1729
1730 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1731 RE::GlobalReplace(). As a result, the number of replacements returned was
1732 double what it should be. I removed one of the increments, but Craig sent a
1733 later patch that removed the other one (the right fix) and added unit tests
1734 that check the return values (which was not done before).
1735
1736 21. Several CMake things:
1737
1738 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1739 the names libpcre and libpcreposix, not just pcre and pcreposix.
1740
1741 (2) The above change means that pcretest and pcregrep are now correctly
1742 linked with the newly-built libraries, not previously installed ones.
1743
1744 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1745
1746 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1747 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1748 UTF-8 newline character). The key issue is that the pattern starts .*;
1749 this means that the match must be either at the beginning, or after a
1750 newline. The bug was in the code for advancing after a failed match and
1751 checking that the new position followed a newline. It was not taking
1752 account of UTF-8 characters correctly.
1753
1754 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1755 character classes. PCRE was not treating the sequence [:...:] as a
1756 character class unless the ... were all letters. Perl, however, seems to
1757 allow any characters between [: and :], though of course it rejects as
1758 unknown any "names" that contain non-letters, because all the known class
1759 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1760 for example, whereas PCRE did not - it did not recognize a POSIX character
1761 class. This seemed a bit dangerous, so the code has been changed to be
1762 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1763 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1764 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1765 Perl does, and where it didn't before.
1766
1767 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1768 Windows environments %n is disabled by default.
1769
1770
1771 Version 7.4 21-Sep-07
1772 ---------------------
1773
1774 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1775 means that a class such as [\s] counted as "explicit reference to CR or
1776 LF". That isn't really right - the whole point of the change was to try to
1777 help when there was an actual mention of one of the two characters. So now
1778 the change happens only if \r or \n (or a literal CR or LF) character is
1779 encountered.
1780
1781 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1782 of both had grown to the point where there were only 3 bits left.
1783 Fortunately, there was spare space in the data structure, and so I have
1784 moved the internal flags into a new 16-bit field to free up more option
1785 bits.
1786
1787 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1788 but did not set the internal JCHANGED flag - either of these is enough to
1789 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1790 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1791 start sets both bits.
1792
1793 4. Added options (at build time, compile time, exec time) to change \R from
1794 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1795
1796 5. doc/pcresyntax.html was missing from the distribution.
1797
1798 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1799 compatibility, even though it is no longer used.
1800
1801 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1802 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1803 windows.h file is present (where different names are used). [This was
1804 reversed later after testing - see 16 below.]
1805
1806 8. Changed all #include <config.h> to #include "config.h". There were also
1807 some further <pcre.h> cases that I changed to "pcre.h".
1808
1809 9. When pcregrep was used with the --colour option, it missed the line ending
1810 sequence off the lines that it output.
1811
1812 10. It was pointed out to me that arrays of string pointers cause lots of
1813 relocations when a shared library is dynamically loaded. A technique of
1814 using a single long string with a table of offsets can drastically reduce
1815 these. I have refactored PCRE in four places to do this. The result is
1816 dramatic:
1817
1818 Originally: 290
1819 After changing UCP table: 187
1820 After changing error message table: 43
1821 After changing table of "verbs" 36
1822 After changing table of Posix names 22
1823
1824 Thanks to the folks working on Gregex for glib for this insight.
1825
1826 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1827 unicode-properties was also set.
1828
1829 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1830
1831 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1832 checked only for CRLF.
1833
1834 14. Added casts to pcretest.c to avoid compiler warnings.
1835
1836 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1837
1838 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1839 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1840 entirely. This removes changes made in 7 above.
1841
1842 17. The CMake files have been updated, and there is now more information about
1843 building with CMake in the NON-UNIX-USE document.
1844
1845
1846 Version 7.3 28-Aug-07
1847 ---------------------
1848
1849 1. In the rejigging of the build system that eventually resulted in 7.1, the
1850 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1851 brackets there is not right, since it causes compilers to look for an
1852 installed pcre.h, not the version that is in the source that is being
1853 compiled (which of course may be different). I have changed it back to:
1854
1855 #include "pcre.h"
1856
1857 I have a vague recollection that the change was concerned with compiling in
1858 different directories, but in the new build system, that is taken care of
1859 by the VPATH setting the Makefile.
1860
1861 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1862 when the subject happened to end in the byte 0x85 (e.g. if the last
1863 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1864 characters but of course it shouldn't be taken as a newline when it is part
1865 of another character. The bug was that, for an unlimited repeat of . in
1866 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1867 characters when looking for a newline.
1868
1869 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1870
1871 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1872 in debug output.
1873
1874 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1875 long printing in the pcrecpp unittest when running under MinGW.
1876
1877 6. ESC_K was left out of the EBCDIC table.
1878
1879 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1880 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1881 limit also applies to "virtual nesting" when a pattern is recursive, and in
1882 this case 1000 isn't so big. I have been able to remove this limit at the
1883 expense of backing off one optimization in certain circumstances. Normally,
1884 when pcre_exec() would call its internal match() function recursively and
1885 immediately return the result unconditionally, it uses a "tail recursion"
1886 feature to save stack. However, when a subpattern that can match an empty
1887 string has an unlimited repetition quantifier, it no longer makes this
1888 optimization. That gives it a stack frame in which to save the data for
1889 checking that an empty string has been matched. Previously this was taken
1890 from the 1000-entry workspace that had been reserved. So now there is no
1891 explicit limit, but more stack is used.
1892
1893 8. Applied Daniel's patches to solve problems with the import/export magic
1894 syntax that is required for Windows, and which was going wrong for the
1895 pcreposix and pcrecpp parts of the library. These were overlooked when this
1896 problem was solved for the main library.
1897
1898 9. There were some crude static tests to avoid integer overflow when computing
1899 the size of patterns that contain repeated groups with explicit upper
1900 limits. As the maximum quantifier is 65535, the maximum group length was
1901 set at 30,000 so that the product of these two numbers did not overflow a
1902 32-bit integer. However, it turns out that people want to use groups that
1903 are longer than 30,000 bytes (though not repeat them that many times).
1904 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1905 made it possible to implement the integer overflow checks in a much more
1906 dynamic way, which I have now done. The artificial limitation on group
1907 length has been removed - we now have only the limit on the total length of
1908 the compiled pattern, which depends on the LINK_SIZE setting.
1909
1910 10. Fixed a bug in the documentation for get/copy named substring when
1911 duplicate names are permitted. If none of the named substrings are set, the
1912 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1913 empty string.
1914
1915 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1916 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1917 because the ] is interpreted as the first data character and the
1918 terminating ] is not found. PCRE has been made compatible with Perl in this
1919 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1920 cause memory overwriting.
1921
1922 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1923 string has been matched (to stop an infinite loop). It was not recognizing
1924 a conditional subpattern that could match an empty string if that
1925 subpattern was within another subpattern. For example, it looped when
1926 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1927 condition was not nested. This bug has been fixed.
1928
1929 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1930 past the start of the subject in the presence of bytes with the top bit
1931 set, for example "\x8aBCD".
1932
1933 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1934 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1935
1936 14. Optimized (?!) to (*FAIL).
1937
1938 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1939 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1940 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1941 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1942 does: it's just the validity check that is more restrictive.
1943
1944 16. Inserted checks for integer overflows during escape sequence (backslash)
1945 processing, and also fixed erroneous offset values for syntax errors during
1946 backslash processing.
1947
1948 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1949 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1950
1951 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1952 caused an overrun.
1953
1954 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1955 something other than just ASCII characters) inside a group that had an
1956 unlimited repeat caused a loop at compile time (while checking to see
1957 whether the group could match an empty string).
1958
1959 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1960 [\P{Any}] did so. (Error in the code for printing property names.)
1961
1962 21. An orphan \E inside a character class could cause a crash.
1963
1964 22. A repeated capturing bracket such as (A)? could cause a wild memory
1965 reference during compilation.
1966
1967 23. There are several functions in pcre_compile() that scan along a compiled
1968 expression for various reasons (e.g. to see if it's fixed length for look
1969 behind). There were bugs in these functions when a repeated \p or \P was
1970 present in the pattern. These operators have additional parameters compared
1971 with \d, etc, and these were not being taken into account when moving along
1972 the compiled data. Specifically:
1973
1974 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1975 length.
1976
1977 (b) An item such as \pL+ within a repeated group could cause crashes or
1978 loops.
1979
1980 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1981 "reference to non-existent subpattern" error.
1982
1983 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1984
1985 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1986 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1987
1988 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1989
1990 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1991 character were causing crashes (broken optimization).
1992
1993 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1994 \p or \P) caused a compile-time loop.
1995
1996 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1997 break. For example, the unstudied pattern [\r\n]A does not match the string
1998 "\r\nA" because change 7.0/46 below moves the current point on by two
1999 characters after failing to match at the start. However, the pattern \nA
2000 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2001 the same is true. There doesn't seem any very clean way out of this, but
2002 what I have chosen to do makes the common cases work: PCRE now takes note
2003 of whether there can be an explicit match for \r or \n anywhere in the
2004 pattern, and if so, 7.0/46 no longer applies. As part of this change,
2005 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2006 pattern has explicit CR or LF references.
2007
2008 29. Added (*CR) etc for changing newline setting at start of pattern.
2009
2010
2011 Version 7.2 19-Jun-07
2012 ---------------------
2013
2014 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2015 which is apparently normally available under Windows.
2016
2017 2. Re-jig the pcregrep tests with different newline settings in an attempt
2018 to make them independent of the local environment's newline setting.
2019
2020 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2021
2022 4. Some of the "internals" tests were previously cut out when the link size
2023 was not 2, because the output contained actual offsets. The recent new
2024 "Z" feature of pcretest means that these can be cut out, making the tests
2025 usable with all link sizes.
2026
2027 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2028 stack recursion. This gives a massive performance boost under BSD, but just
2029 a small improvement under Linux. However, it saves one field in the frame
2030 in all cases.
2031
2032 6. Added more features from the forthcoming Perl 5.10:
2033
2034 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2035 recursion call. It refers to the nth most recently opened parentheses.
2036
2037 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2038 to be opened parentheses.
2039
2040 (c) Conditions that refer to capturing parentheses can be specified
2041 relatively, for example, (?(-2)... or (?(+3)...
2042
2043 (d) \K resets the start of the current match so that everything before
2044 is not part of it.
2045
2046 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2047
2048 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2049 reference syntax.
2050
2051 (g) (?| introduces a group in which the numbering of parentheses in each
2052 alternative starts with the same number.
2053
2054 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2055
2056 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2057 PCRE_INFO_JCHANGED.
2058
2059 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2060 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2061 for detecting groups that can match an empty string.
2062
2063 9. A pattern with a very large number of alternatives (more than several
2064 hundred) was running out of internal workspace during the pre-compile
2065 phase, where pcre_compile() figures out how much memory will be needed. A
2066 bit of new cunning has reduced the workspace needed for groups with
2067 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2068 workspace instead of running out of the 4096 that are available.
2069
2070 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2071
2072 11. Applied patch from Google to remove an optimization that didn't quite work.
2073 The report of the bug said:
2074
2075 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2076 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2077 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2078
2079 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2080 it matched the wrong number of bytes.
2081
2082
2083 Version 7.1 24-Apr-07
2084 ---------------------
2085
2086 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2087 that is more "standard", making use of automake and other Autotools. There
2088 is some re-arrangement of the files and adjustment of comments consequent
2089 on this.
2090
2091 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2092 for recursive directory scanning broke on some systems because the files
2093 are not scanned in any specific order and on different systems the order
2094 was different. A call to "sort" has been inserted into RunGrepTest for the
2095 approprate test as a short-term fix. In the longer term there may be an
2096 alternative.
2097
2098 3. I had an email from Eric Raymond about problems translating some of PCRE's
2099 man pages to HTML (despite the fact that I distribute HTML pages, some
2100 people do their own conversions for various reasons). The problems
2101 concerned the use of low-level troff macros .br and .in. I have therefore
2102 removed all such uses from the man pages (some were redundant, some could
2103 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2104 HTML has been updated to handle .nf/.fi and to complain if it encounters
2105 .br or .in.
2106
2107 4. Updated comments in configure.ac that get placed in config.h.in and also
2108 arranged for config.h to be included in the distribution, with the name
2109 config.h.generic, for the benefit of those who have to compile without
2110 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2111
2112 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2113 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2114 makevp.bat for the current PCRE, using the additional files
2115 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2116
2117 6. A Windows user reported a minor discrepancy with test 2, which turned out
2118 to be caused by a trailing space on an input line that had got lost in his
2119 copy. The trailing space was an accident, so I've just removed it.
2120
2121 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2122 that is needed.
2123
2124 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2125 as "const" (a) because they are and (b) because it helps the PHP
2126 maintainers who have recently made a script to detect big data structures
2127 in the php code that should be moved to the .rodata section. I remembered
2128 to update Builducptable as well, so it won't revert if ucptable.h is ever
2129 re-created.
2130
2131 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2132 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2133 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2134 support is not required. This saves 1.5-2K of code, which is important in
2135 some applications.
2136
2137 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2138 so as not to refer to the tables, even though these functions will never be
2139 called when UTF-8 support is disabled. Otherwise there are problems with a
2140 shared library.
2141
2142 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2143
2144 (a) It was defining its arguments as char * instead of void *.
2145
2146 (b) It was assuming that all moves were upwards in memory; this was true
2147 a long time ago when I wrote it, but is no longer the case.
2148
2149 The emulated memove() is provided for those environments that have neither
2150 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2151 is clearly not the case, as these two bugs were recently reported.
2152
2153 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2154 and Detrail to create the HTML documentation, the .txt form of the man
2155 pages, and it removes trailing spaces from listed files. It also creates
2156 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2157 case, it wraps all the #defines with #ifndefs. This script should be run
2158 before "make dist".
2159
2160 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2161 with Unicode property support.
2162
2163 (a) For a maximizing quantifier, if the two different cases of the
2164 character were of different lengths in their UTF-8 codings (there are
2165 some cases like this - I found 11), and the matching function had to
2166 back up over a mixture of the two cases, it incorrectly assumed they
2167 were both the same length.
2168
2169 (b) When PCRE was configured to use the heap rather than the stack for
2170 recursion during matching, it was not correctly preserving the data for
2171 the other case of a UTF-8 character when checking ahead for a match
2172 while processing a minimizing repeat. If the check also involved
2173 matching a wide character, but failed, corruption could cause an
2174 erroneous result when trying to check for a repeat of the original
2175 character.
2176
2177 13. Some tidying changes to the testing mechanism:
2178
2179 (a) The RunTest script now detects the internal link size and whether there
2180 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2181 values substituted by "configure". (The RunGrepTest script already did
2182 this for UTF-8.) The configure.ac script no longer substitutes the
2183 relevant variables.
2184
2185 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2186 with length and offset values. This means that the output is different
2187 for different internal link sizes. Test 2 is skipped for link sizes
2188 other than 2 because of this, bypassing the problem. Unfortunately,
2189 there was also a test in test 3 (the locale tests) that used /B and
2190 failed for link sizes other than 2. Rather than cut the whole test out,
2191 I have added a new /Z option to pcretest that replaces the length and
2192 offset values with spaces. This is now used to make test 3 independent
2193 of link size. (Test 2 will be tidied up later.)
2194
2195 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2196 segmentation fault instead of returning the appropriate error message.
2197
2198 15. In multiline mode when the newline sequence was set to "any", the pattern
2199 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2200 This doesn't seem right; it now treats the CRLF combination as the line
2201 ending, and so does not match in that case. It's only a pattern such as ^$
2202 that would hit this one: something like ^ABC$ would have failed after \r
2203 and then tried again after \r\n.
2204
2205 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2206 in an attempt to make files that differ only in their line terminators
2207 compare equal. This works on Linux.
2208
2209 17. Under certain error circumstances pcregrep might try to free random memory
2210 as it exited. This is now fixed, thanks to valgrind.
2211
2212 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2213 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2214 was because its rules for how to advance for /g after matching an empty
2215 string at the end of a line did not allow for this case. They now check for
2216 it specially.
2217
2218 20. pcretest is supposed to handle patterns and data of any length, by
2219 extending its buffers when necessary. It was getting this wrong when the
2220 buffer for a data line had to be extended.
2221
2222 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2223 CRLF as a newline sequence.
2224
2225 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2226 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2227 I have nevertheless tidied it up.
2228
2229 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2230
2231 24. Added a man page for pcre-config.
2232
2233
2234 Version 7.0 19-Dec-06
2235 ---------------------
2236
2237 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2238 moving to gcc 4.1.1.
2239
2240 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2241 sys/time.h, which is documented as needed for this function. It doesn't
2242 seem to matter on Linux, but it showed up on some releases of OS X.
2243
2244 3. It seems that there are systems where bytes whose values are greater than
2245 127 match isprint() in the "C" locale. The "C" locale should be the
2246 default when a C program starts up. In most systems, only ASCII printing
2247 characters match isprint(). This difference caused the output from pcretest
2248 to vary, making some of the tests fail. I have changed pcretest so that:
2249
2250 (a) When it is outputting text in the compiled version of a pattern, bytes
2251 other than 32-126 are always shown as hex escapes.
2252
2253 (b) When it is outputting text that is a matched part of a subject string,
2254 it does the same, unless a different locale has been set for the match
2255 (using the /L modifier). In this case, it uses isprint() to decide.
2256
2257 4. Fixed a major bug that caused incorrect computation of the amount of memory
2258 required for a compiled pattern when options that changed within the
2259 pattern affected the logic of the preliminary scan that determines the
2260 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2261 that the computed length was too small. The symptoms of this bug were
2262 either the PCRE error "internal error: code overflow" from pcre_compile(),
2263 or a glibc crash with a message such as "pcretest: free(): invalid next
2264 size (fast)". Examples of patterns that provoked this bug (shown in
2265 pcretest format) are:
2266
2267 /(?-x: )/x
2268 /(?x)(?-x: \s*#\s*)/
2269 /((?i)[\x{c0}])/8
2270 /(?i:[\x{c0}])/8
2271
2272 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2273 is now done differently.
2274
2275 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2276 wrapper classes; (b) implement a new function in the C++ scanner that is
2277 more efficient than the old way of doing things because it avoids levels of
2278 recursion in the regex matching; (c) add a paragraph to the documentation
2279 for the FullMatch() function.
2280
2281 6. The escape sequence \n was being treated as whatever was defined as
2282 "newline". Not only was this contrary to the documentation, which states
2283 that \n is character 10 (hex 0A), but it also went horribly wrong when
2284 "newline" was defined as CRLF. This has been fixed.
2285
2286 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2287 was being set to -1 for the "end of line" case (supposedly a value that no
2288 character can have). Though this value is never used (the check for end of
2289 line is "zero bytes in current character"), it caused compiler complaints.
2290 I've changed it to 0xffffffff.
2291
2292 8. In pcre_version.c, the version string was being built by a sequence of
2293 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2294 string (as it is for production releases) called a macro with an empty
2295 argument. The C standard says the result of this is undefined. The gcc
2296 compiler treats it as an empty string (which was what was wanted) but it is
2297 reported that Visual C gives an error. The source has been hacked around to
2298 avoid this problem.
2299
2300 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2301 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2302 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2303 of them did).
2304
2305 10. Originally, pcretest opened its input and output without "b"; then I was
2306 told that "b" was needed in some environments, so it was added for release
2307 5.0 to both the input and output. (It makes no difference on Unix-like
2308 systems.) Later I was told that it is wrong for the input on Windows. I've
2309 now abstracted the modes into two macros, to make it easier to fiddle with
2310 them, and removed "b" from the input mode under Windows.
2311
2312 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2313
2314 12. Added -help and --help to pcretest as an official way of being reminded
2315 of the options.
2316
2317 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2318 and pcrecpp.cc because they annoy compilers at high warning levels.
2319
2320 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2321
2322 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2323 scripts are not C programs :-) and which was not noticed because it works
2324 on Linux.
2325
2326 16. pcretest is supposed to handle any length of pattern and data line (as one
2327 line or as a continued sequence of lines) by extending its input buffer if
2328 necessary. This feature was broken for very long pattern lines, leading to
2329 a string of junk being passed to pcre_compile() if the pattern was longer
2330 than about 50K.
2331
2332 17. I have done a major re-factoring of the way pcre_compile() computes the
2333 amount of memory needed for a compiled pattern. Previously, there was code
2334 that made a preliminary scan of the pattern in order to do this. That was
2335 OK when PCRE was new, but as the facilities have expanded, it has become
2336 harder and harder to keep it in step with the real compile phase, and there
2337 have been a number of bugs (see for example, 4 above). I have now found a
2338 cunning way of running the real compile function in a "fake" mode that
2339 enables it to compute how much memory it would need, while actually only
2340 ever using a few hundred bytes of working memory and without too many
2341 tests of the mode. This should make future maintenance and development
2342 easier. A side effect of this work is that the limit of 200 on the nesting
2343 depth of parentheses has been removed (though this was never a serious
2344 limitation, I suspect). However, there is a downside: pcre_compile() now
2345 runs more slowly than before (30% or more, depending on the pattern). I
2346 hope this isn't a big issue. There is no effect on runtime performance.
2347
2348 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2349 newline (only possible for the last line of a file) and it was a
2350 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2351
2352 19. Added additional timing features to pcretest. (1) The -tm option now times
2353 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2354 separate command line item, by a number that specifies the number of
2355 repeats to use when timing. The default is 50000; this gives better
2356 precision, but takes uncomfortably long for very large patterns.
2357
2358 20. Extended pcre_study() to be more clever in cases where a branch of a
2359 subpattern has no definite first character. For example, (a*|b*)[cd] would
2360 previously give no result from pcre_study(). Now it recognizes that the
2361 first character must be a, b, c, or d.
2362
2363 21. There was an incorrect error "recursive call could loop indefinitely" if
2364 a subpattern (or the entire pattern) that was being tested for matching an
2365 empty string contained only one non-empty item after a nested subpattern.
2366 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2367 incorrectly, because the \d was being skipped in the check.
2368
2369 22. The pcretest program now has a new pattern option /B and a command line
2370 option -b, which is equivalent to adding /B to every pattern. This causes
2371 it to show the compiled bytecode, without the additional information that
2372 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2373 is the same as /B/I).
2374
2375 23. A new optimization is now able automatically to treat some sequences such
2376 as a*b as a*+b. More specifically, if something simple (such as a character
2377 or a simple class like \d) has an unlimited quantifier, and is followed by
2378 something that cannot possibly match the quantified thing, the quantifier
2379 is automatically "possessified".
2380
2381 24. A recursive reference to a subpattern whose number was greater than 39
2382 went wrong under certain circumstances in UTF-8 mode. This bug could also
2383 have affected the operation of pcre_study().
2384
2385 25. Realized that a little bit of performance could be had by replacing
2386 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2387
2388 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2389
2390 27. Possessive quantifiers such as a++ were previously implemented by turning
2391 them into atomic groups such as ($>a+). Now they have their own opcodes,
2392 which improves performance. This includes the automatically created ones
2393 from 23 above.
2394
2395 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2396 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2397 the first matched character to be a colon. This applied both to named and
2398 numbered groups.
2399
2400 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2401
2402 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2403 building PCRE on Windows easier, so I have included it in the distribution.
2404
2405 31. There is now a check in pcretest against a ridiculously large number being
2406 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2407 loop, the loop is abandoned.
2408
2409 32. Forward references to subpatterns in conditions such as (?(2)...) where
2410 subpattern 2 is defined later cause pcre_compile() to search forwards in
2411 the pattern for the relevant set of parentheses. This search went wrong
2412 when there were unescaped parentheses in a character class, parentheses
2413 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2414
2415 33. "Subroutine" calls and backreferences were previously restricted to
2416 referencing subpatterns earlier in the regex. This restriction has now
2417 been removed.
2418
2419 34. Added a number of extra features that are going to be in Perl 5.10. On the
2420 whole, these are just syntactic alternatives for features that PCRE had
2421 previously implemented using the Python syntax or my own invention. The
2422 other formats are all retained for compatibility.
2423
2424 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2425 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2426 also .NET compatible.
2427
2428 (b) A recursion or subroutine call to a named group can now be defined as
2429 (?&name) as well as (?P>name).
2430
2431 (c) A backreference to a named group can now be defined as \k<name> or
2432 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2433 5.10, are also .NET compatible.
2434
2435 (d) A conditional reference to a named group can now use the syntax
2436 (?(<name>) or (?('name') as well as (?(name).
2437
2438 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2439 groups (named and numbered) that are never evaluated inline, but can be
2440 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2441 is always false. There may be only one alternative in such a group.
2442
2443 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2444 as the simple (?(R). The condition is true only if the most recent
2445 recursion is that of the given number or name. It does not search out
2446 through the entire recursion stack.
2447
2448 (g) The escape \gN or \g{N} has been added, where N is a positive or
2449 negative number, specifying an absolute or relative reference.
2450
2451 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2452 some "unreachable code" warnings.
2453
2454 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2455 things, this adds five new scripts.
2456
2457 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2458 There were also incompatibilities regarding the handling of \Q..\E inside
2459 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2460 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2461
2462 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2463 matches an empty string, and forcibly breaks the loop. There were bugs in
2464 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2465 against aaaa the result was just "a" rather than "aaaa", for example. Two
2466 separate and independent bugs (that affected different cases) have been
2467 fixed.
2468
2469 39. Refactored the code to abolish the use of different opcodes for small
2470 capturing bracket numbers. This is a tidy that I avoided doing when I
2471 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2472 The new approach is not only tidier, it makes it possible to reduce the
2473 memory needed to fix the previous bug (38).
2474
2475 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2476 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2477 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2478 mode.
2479
2480 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2481 report.
2482
2483 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2484 copy construction and assignment in the C++ wrapper.
2485
2486 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2487 couple of bugs that could have given wrong results in the "--newline=crlf"
2488 case.
2489
2490 44. Added a number of casts and did some reorganization of signed/unsigned int
2491 variables following suggestions from Dair Grant. Also renamed the variable
2492 "this" as "item" because it is a C++ keyword.
2493
2494 45. Arranged for dftables to add
2495
2496 #include "pcre_internal.h"
2497
2498 to pcre_chartables.c because without it, gcc 4.x may remove the array
2499 definition from the final binary if PCRE is built into a static library and
2500 dead code stripping is activated.
2501
2502 46. For an unanchored pattern, if a match attempt fails at the start of a
2503 newline sequence, and the newline setting is CRLF or ANY, and the next two
2504 characters are CRLF, advance by two characters instead of one.
2505
2506
2507 Version 6.7 04-Jul-06
2508 ---------------------
2509
2510 1. In order to handle tests when input lines are enormously long, pcretest has
2511 been re-factored so that it automatically extends its buffers when
2512 necessary. The code is crude, but this _is_ just a test program. The
2513 default size has been increased from 32K to 50K.
2514
2515 2. The code in pcre_study() was using the value of the re argument before
2516 testing it for NULL. (Of course, in any sensible call of the function, it
2517 won't be NULL.)
2518
2519 3. The memmove() emulation function in pcre_internal.h, which is used on
2520 systems that lack both memmove() and bcopy() - that is, hardly ever -
2521 was missing a "static" storage class specifier.
2522
2523 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2524 containing an extended class (one that cannot be represented by a bitmap
2525 because it contains high-valued characters or Unicode property items, e.g.
2526 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2527 pattern, but PCRE should not loop if you do not (it no longer does).
2528 [Detail: two cases were found: (a) a repeated subpattern containing an
2529 extended class; (b) a recursive reference to a subpattern that followed a
2530 previous extended class. It wasn't skipping over the extended class
2531 correctly when UTF-8 mode was not set.]
2532
2533 5. A negated single-character class was not being recognized as fixed-length
2534 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2535 compile error "lookbehind assertion is not fixed length".
2536
2537 6. The RunPerlTest auxiliary script was showing an unexpected difference
2538 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2539 write a Perl script that can interpret lines of an input file either as
2540 byte characters or as UTF-8, which is what "perltest" was being required to
2541 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2542 can't do is switch easily at run time between having the "use utf8;" pragma
2543 or not. In the end, I fudged it by using the RunPerlTest script to insert
2544 "use utf8;" explicitly for the UTF-8 tests.
2545
2546 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2547 the end of the subject string, contrary to the documentation and to what
2548 Perl does. This was true of both matching functions. Now it matches only at
2549 the start of the subject and immediately after *internal* newlines.
2550
2551 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2552 a pointer to an int instead of a pointer to an unsigned long int. This
2553 caused problems on 64-bit systems.
2554
2555 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2556 instance of the 'standard' template library not being so standard".
2557
2558 10. There was no check on the number of named subpatterns nor the maximum
2559 length of a subpattern name. The product of these values is used to compute
2560 the size of the memory block for a compiled pattern. By supplying a very
2561 long subpattern name and a large number of named subpatterns, the size
2562 computation could be caused to overflow. This is now prevented by limiting
2563 the length of names to 32 characters, and the number of named subpatterns
2564 to 10,000.
2565
2566 11. Subpatterns that are repeated with specific counts have to be replicated in
2567 the compiled pattern. The size of memory for this was computed from the
2568 length of the subpattern and the repeat count. The latter is limited to
2569 65535, but there was no limit on the former, meaning that integer overflow
2570 could in principle occur. The compiled length of a repeated subpattern is
2571 now limited to 30,000 bytes in order to prevent this.
2572
2573 12. Added the optional facility to have named substrings with the same name.
2574
2575 13. Added the ability to use a named substring as a condition, using the
2576 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2577 are numbers (not recommended). Forward references are permitted.
2578
2579 14. Added forward references in named backreferences (if you see what I mean).
2580
2581 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2582 pattern could run off the end of the subject. For example, the pattern
2583 "(?s)(.{1,5})"8 did this with the subject "ab".
2584
2585 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2586 PCRE_CASELESS was set when matching characters that were quantified with ?
2587 or *.
2588
2589 17. A character class other than a single negated character that had a minimum
2590 but no maximum quantifier - for example [ab]{6,} - was not handled
2591 correctly by pce_dfa_exec(). It would match only one character.
2592
2593 18. A valid (though odd) pattern that looked like a POSIX character
2594 class but used an invalid character after [ (for example [[,abc,]]) caused
2595 pcre_compile() to give the error "Failed: internal error: code overflow" or
2596 in some cases to crash with a glibc free() error. This could even happen if
2597 the pattern terminated after [[ but there just happened to be a sequence of
2598 letters, a binary zero, and a closing ] in the memory that followed.
2599
2600 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2601 over the years. Originally (before any Unicode support), just the bottom 8
2602 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2603 output from "man perlunicode" includes this:
2604
2605 The regular expression compiler produces polymorphic opcodes. That
2606 is, the pattern adapts to the data and automatically switches to
2607 the Unicode character scheme when presented with Unicode data--or
2608 instead uses a traditional byte scheme when presented with byte
2609 data.
2610
2611 Sadly, a wide octal escape does not cause a switch, and in a string with
2612 no other multibyte characters, these octal escapes are treated as before.
2613 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2614 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2615 Unicode string.
2616
2617 I have not perpetrated such confusion in PCRE. Up till now, it took just
2618 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2619 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2620 translate to the appropriate multibyte character.
2621
2622 29. Applied some refactoring to reduce the number of warnings from Microsoft
2623 and Borland compilers. This has included removing the fudge introduced
2624 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2625 a warning about an unused variable.
2626
2627 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2628 characters since release 4.0, because Perl (from release 5.004) does not.
2629 [Or at least, is documented not to: some releases seem to be in conflict
2630 with the documentation.] However, when a pattern was studied with
2631 pcre_study() and all its branches started with \s, PCRE still included VT
2632 as a possible starting character. Of course, this did no harm; it just
2633 caused an unnecessary match attempt.
2634
2635 22. Removed a now-redundant internal flag bit that recorded the fact that case
2636 dependency changed within the pattern. This was once needed for "required
2637 byte" processing, but is no longer used. This recovers a now-scarce options
2638 bit. Also moved the least significant internal flag bit to the most-
2639 significant bit of the word, which was not previously used (hangover from
2640 the days when it was an int rather than a uint) to free up another bit for
2641 the future.
2642
2643 23. Added support for CRLF line endings as well as CR and LF. As well as the
2644 default being selectable at build time, it can now be changed at runtime
2645 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2646 specify that it is scanning data with non-default line endings.
2647
2648 24. Changed the definition of CXXLINK to make it agree with the definition of
2649 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2650
2651 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2652 recursions. This makes a big different to stack usage for some patterns.
2653
2654 26. If a subpattern containing a named recursion or subroutine reference such
2655 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2656 the space required for the compiled pattern went wrong and gave too small a
2657 value. Depending on the environment, this could lead to "Failed: internal
2658 error: code overflow at offset 49" or "glibc detected double free or
2659 corruption" errors.
2660
2661 27. Applied patches from Google (a) to support the new newline modes and (b) to
2662 advance over multibyte UTF-8 characters in GlobalReplace.
2663
2664 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2665 difference for some implementation of PCRE in some Windows version.
2666
2667 29. Added some extra testing facilities to pcretest:
2668
2669 \q<number> in a data line sets the "match limit" value
2670 \Q<number> in a data line sets the "match recursion limt" value
2671 -S <number> sets the stack size, where <number> is in megabytes
2672
2673 The -S option isn't available for Windows.
2674
2675
2676 Version 6.6 06-Feb-06
2677 ---------------------
2678
2679 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2680 in pcreposix.h. I have copied the definition from pcre.h.
2681
2682 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2683 because pcre.h is no longer a built file.
2684
2685 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2686 not normally included in the compiled code.
2687
2688
2689 Version 6.5 01-Feb-06
2690 ---------------------
2691
2692 1. When using the partial match feature with pcre_dfa_exec(), it was not
2693 anchoring the second and subsequent partial matches at the new starting
2694 point. This could lead to incorrect results. For example, with the pattern
2695 /1234/, partially matching against "123" and then "a4" gave a match.
2696
2697 2. Changes to pcregrep:
2698
2699 (a) All non-match returns from pcre_exec() were being treated as failures
2700 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2701 error message is output. Some extra information is given for the
2702 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2703 probably the only errors that are likely to be caused by users (by
2704 specifying a regex that has nested indefinite repeats, for instance).
2705 If there are more than 20 of these errors, pcregrep is abandoned.
2706
2707 (b) A binary zero was treated as data while matching, but terminated the
2708 output line if it was written out. This has been fixed: binary zeroes
2709 are now no different to any other data bytes.
2710
2711 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2712 used to set a locale for matching. The --locale=xxxx long option has
2713 been added (no short equivalent) to specify a locale explicitly on the
2714 pcregrep command, overriding the environment variables.
2715
2716 (d) When -B was used with -n, some line numbers in the output were one less
2717 than they should have been.
2718
2719 (e) Added the -o (--only-matching) option.
2720
2721 (f) If -A or -C was used with -c (count only), some lines of context were
2722 accidentally printed for the final match.
2723
2724 (g) Added the -H (--with-filename) option.
2725
2726 (h) The combination of options -rh failed to suppress file names for files
2727 that were found from directory arguments.
2728
2729 (i) Added the -D (--devices) and -d (--directories) options.
2730
2731 (j) Added the -F (--fixed-strings) option.
2732
2733 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2734
2735 (l) Added the --colo(u)r option.
2736
2737 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2738 is not present by default.
2739
2740 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2741 items such as (?R) or (?1), when the recursion could match a number of
2742 alternatives. If it matched one of the alternatives, but subsequently,
2743 outside the recursion, there was a failure, the code tried to back up into
2744 the recursion. However, because of the way PCRE is implemented, this is not
2745 possible, and the result was an incorrect result from the match.
2746
2747 In order to prevent this happening, the specification of recursion has
2748 been changed so that all such subpatterns are automatically treated as
2749 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2750
2751 4. I had overlooked the fact that, in some locales, there are characters for
2752 which isalpha() is true but neither isupper() nor islower() are true. In
2753 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2754 and ordfeminine) are like this. This affected the treatment of \w and \W
2755 when they appeared in character classes, but not when they appeared outside
2756 a character class. The bit map for "word" characters is now created
2757 separately from the results of isalnum() instead of just taking it from the
2758 upper, lower, and digit maps. (Plus the underscore character, of course.)
2759
2760 5. The above bug also affected the handling of POSIX character classes such as
2761 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2762 permanent tables. Instead, the bit maps for such a class were previously
2763 created as the appropriate unions of the upper, lower, and digit bitmaps.
2764 Now they are created by subtraction from the [[:word:]] class, which has
2765 its own bitmap.
2766
2767 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2768 It is created by subtracting the vertical space characters (\x09, \x0a,
2769 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2770 subtraction was done in the overall bitmap for a character class, meaning
2771 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2772 be recognized. This bug has been fixed.
2773
2774 7. Patches from the folks at Google:
2775
2776 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2777 real life, but is still worth protecting against".
2778
2779 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2780 regular expressions".
2781
2782 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2783 have it.
2784
2785 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2786 "configure" and the latter not, in order to fix a problem somebody had
2787 with compiling the Arg class on HP-UX.
2788
2789 (e) Improve the error-handling of the C++ wrapper a little bit.
2790
2791 (f) New tests for checking recursion limiting.
2792
2793 8. The pcre_memmove() function, which is used only if the environment does not
2794 have a standard memmove() function (and is therefore rarely compiled),
2795 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2796 returning a result (though PCRE never actually uses the result).
2797
2798 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2799 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2800 returned instead of calling malloc() with an overflowing number that would
2801 most likely cause subsequent chaos.
2802
2803 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2804
2805 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2806 with this option is matched, the nmatch and pmatch options of regexec() are
2807 ignored.
2808
2809 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2810 provided in case anyone wants to the the POSIX interface with UTF-8
2811 strings.
2812
2813 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2814 C++ linking (needed for some HP-UX environments).
2815
2816 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2817 (unused parameter) and in the pcre_printint() function (omitted "default"
2818 switch label when the default is to do nothing).
2819
2820 15. Added some code to make it possible, when PCRE is compiled as a C++
2821 library, to replace subject pointers for pcre_exec() with a smart pointer
2822 class, thus making it possible to process discontinuous strings.
2823
2824 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2825 much the same function. They were added by different people who were trying
2826 to make PCRE easy to compile on non-Unix systems. It has been suggested
2827 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2828 for compiling on Windows systems. I have therefore replaced it with
2829 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2830 defaults to "extern" for C or "extern C" for C++, which works fine on
2831 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2832 SCOPE with something explicit in config.h. In addition:
2833
2834 (a) pcreposix.h still had just "extern" instead of either of these macros;
2835 I have replaced it with PCRE_DATA_SCOPE.
2836
2837 (b) Functions such as _pcre_xclass(), which are internal to the library,
2838 but external in the C sense, all had PCRE_EXPORT in their definitions.
2839 This is apparently wrong for the Windows case, so I have removed it.
2840 (It makes no difference on Unix-like systems.)
2841
2842 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2843 of recursive calls to match(). This is different to MATCH_LIMIT because
2844 that limits the total number of calls to match(), not all of which increase
2845 the depth of recursion. Limiting the recursion depth limits the amount of
2846 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2847 when PCRE is compiled, and changed at run time. A patch from Google adds
2848 this functionality to the C++ interface.
2849
2850 18. Changes to the handling of Unicode character properties:
2851
2852 (a) Updated the table to Unicode 4.1.0.
2853
2854 (b) Recognize characters that are not in the table as "Cn" (undefined).
2855
2856 (c) I revised the way the table is implemented to a much improved format
2857 which includes recognition of ranges. It now supports the ranges that
2858 are defined in UnicodeData.txt, and it also amalgamates other
2859 characters into ranges. This has reduced the number of entries in the
2860 table from around 16,000 to around 3,000, thus reducing its size
2861 considerably. I realized I did not need to use a tree structure after
2862 all - a binary chop search is just as efficient. Having reduced the
2863 number of entries, I extended their size from 6 bytes to 8 bytes to
2864 allow for more data.
2865
2866 (d) Added support for Unicode script names via properties such as \p{Han}.
2867
2868 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2869 matching that character.
2870
2871 20. When matching a repeated Unicode property with a minimum greater than zero,
2872 (for example \pL{2,}), PCRE could look past the end of the subject if it
2873 reached it while seeking the minimum number of characters. This could
2874 happen only if some of the characters were more than one byte long, because
2875 there is a check for at least the minimum number of bytes.
2876
2877 21. Refactored the implementation of \p and \P so as to be more general, to
2878 allow for more different types of property in future. This has changed the
2879 compiled form incompatibly. Anybody with saved compiled patterns that use
2880 \p or \P will have to recompile them.
2881
2882 22. Added "Any" and "L&" to the supported property types.
2883
2884 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2885 but give a compile time error if the value is greater than 0xff.
2886
2887 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2888 accidentally not being installed or uninstalled.
2889
2890 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2891 made were to insert the current release number. This seemed silly, because
2892 it made things harder for people building PCRE on systems that don't run
2893 "configure". I have turned pcre.h into a distributed file, no longer built
2894 by "configure", with the version identification directly included. There is
2895 no longer a pcre.h.in file.
2896
2897 However, this change necessitated a change to the pcre-config script as
2898 well. It is built from pcre-config.in, and one of the substitutions was the
2899 release number. I have updated configure.ac so that ./configure now finds
2900 the release number by grepping pcre.h.
2901
2902 26. Added the ability to run the tests under valgrind.
2903
2904
2905 Version 6.4 05-Sep-05
2906 ---------------------
2907
2908 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2909 "--" to be printed when multiple files were scanned, even when none of the
2910 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2911 consider it to be a bug, and have restored the previous behaviour.
2912
2913 2. A couple of code tidies to get rid of compiler warnings.
2914
2915 3. The pcretest program used to cheat by referring to symbols in the library
2916 whose names begin with _pcre_. These are internal symbols that are not
2917 really supposed to be visible externally, and in some environments it is
2918 possible to suppress them. The cheating is now confined to including
2919 certain files from the library's source, which is a bit cleaner.
2920
2921 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2922 file's purpose clearer.
2923
2924 5. Reorganized pcre_ucp_findchar().
2925
2926
2927 Version 6.3 15-Aug-05
2928 ---------------------
2929
2930 1. The file libpcre.pc.in did not have general read permission in the tarball.
2931
2932 2. There were some problems when building without C++ support:
2933
2934 (a) If C++ support was not built, "make install" and "make test" still
2935 tried to test it.
2936
2937 (b) There were problems when the value of CXX was explicitly set. Some
2938 changes have been made to try to fix these, and ...
2939
2940 (c) --disable-cpp can now be used to explicitly disable C++ support.
2941
2942 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2943 backslash in a target when C++ was disabled. This confuses some
2944 versions of "make", apparently. Using an intermediate variable solves
2945 this. (Same for CPP_LOBJ.)
2946
2947 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2948 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2949 necessary on certain architectures.
2950
2951 4. Added a setting of -export-symbols-regex to the link command to remove
2952 those symbols that are exported in the C sense, but actually are local
2953 within the library, and not documented. Their names all begin with
2954 "_pcre_". This is not a perfect job, because (a) we have to except some
2955 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2956 available (and never for static libraries). I have made a note to try to
2957 find a way round (a) in the future.
2958
2959
2960 Version 6.2 01-Aug-05
2961 ---------------------
2962
2963 1. There was no test for integer overflow of quantifier values. A construction
2964 such as {1111111111111111} would give undefined results. What is worse, if
2965 a minimum quantifier for a parenthesized subpattern overflowed and became
2966 negative, the calculation of the memory size went wrong. This could have
2967 led to memory overwriting.
2968
2969 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2970
2971 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2972 operating environments where this matters.
2973
2974 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2975 PCRE options from within the C++ wrapper.
2976
2977 5. Named capturing subpatterns were not being correctly counted when a pattern
2978 was compiled. This caused two problems: (a) If there were more than 100
2979 such subpatterns, the calculation of the memory needed for the whole
2980 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2981 back references of the form \12, where the number was greater than 9, were
2982 not recognized as back references, even though there were sufficient
2983 previous subpatterns.
2984
2985 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2986 versions of gcc, e.g. 2.95.4.
2987
2988
2989 Version 6.1 21-Jun-05
2990 ---------------------
2991
2992 1. There was one reference to the variable "posix" in pcretest.c that was not
2993 surrounded by "#if !defined NOPOSIX".
2994
2995 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2996 the cross-check on the old pcre_info() function, for the benefit of the
2997 cut-down version of PCRE that is currently imported into Exim.
2998
2999 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3000 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3001 patterns that start (?i) but I don't think that matters. The use of (?i) is
3002 just an example; this all applies to the other options as well.
3003
3004 4. Since libtool seems to echo the compile commands it is issuing, the output
3005 from "make" can be reduced a bit by putting "@" in front of each libtool
3006 compile command.
3007
3008 5. Patch from the folks at Google for configure.in to be a bit more thorough
3009 in checking for a suitable C++ installation before trying to compile the
3010 C++ stuff. This should fix a reported problem when a compiler was present,
3011 but no suitable headers.
3012
3013 6. The man pages all had just "PCRE" as their title. I have changed them to
3014 be the relevant file name. I have also arranged that these names are
3015 retained in the file doc/pcre.txt, which is a concatenation in text format
3016 of all the man pages except the little individual ones for each function.
3017
3018 7. The NON-UNIX-USE file had not been updated for the different set of source
3019 files that come with release 6. I also added a few comments about the C++
3020 wrapper.
3021
3022
3023 Version 6.0 07-Jun-05
3024 ---------------------
3025
3026 1. Some minor internal re-organization to help with my DFA experiments.
3027
3028 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3029 didn't matter for the library itself when fully configured, but did matter
3030 when compiling without UCP support, or within Exim, where the ucp files are
3031 not imported.
3032
3033 3. Refactoring of the library code to split up the various functions into
3034 different source modules. The addition of the new DFA matching code (see
3035 below) to a single monolithic source would have made it really too
3036 unwieldy, quite apart from causing all the code to be include in a
3037 statically linked application, when only some functions are used. This is
3038 relevant even without the DFA addition now that patterns can be compiled in
3039 one application and matched in another.
3040
3041 The downside of splitting up is that there have to be some external
3042 functions and data tables that are used internally in different modules of
3043 the library but which are not part of the API. These have all had their
3044 names changed to start with "_pcre_" so that they are unlikely to clash
3045 with other external names.
3046
3047 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3048 a different (DFA) algorithm. Although it is slower than the original
3049 function, it does have some advantages for certain types of matching
3050 problem.
3051
3052 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3053 including restarting after a partial match.
3054
3055 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3056 defined when compiling for Windows was sent to me. I have put it into the
3057 code, though I have no means of testing or verifying it.
3058
3059 7. Added the pcre_refcount() auxiliary function.
3060
3061 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3062 match before or at the first newline in the subject string. In pcretest,
3063 the /f option on a pattern can be used to set this.
3064
3065 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3066 would behave wrongly. This has been present in PCRE since release 4.0.
3067
3068 10. A number of changes to the pcregrep command:
3069
3070 (a) Refactored how -x works; insert ^(...)$ instead of setting
3071 PCRE_ANCHORED and checking the length, in preparation for adding
3072 something similar for -w.
3073
3074 (b) Added the -w (match as a word) option.
3075
3076 (c) Refactored the way lines are read and buffered so as to have more
3077 than one at a time available.
3078
3079 (d) Implemented a pcregrep test script.
3080
3081 (e) Added the -M (multiline match) option. This allows patterns to match
3082 over several lines of the subject. The buffering ensures that at least
3083 8K, or the rest of the document (whichever is the shorter) is available
3084 for matching (and similarly the previous 8K for lookbehind assertions).
3085
3086 (f) Changed the --help output so that it now says
3087
3088 -w, --word-regex(p)
3089
3090 instead of two lines, one with "regex" and the other with "regexp"
3091 because that confused at least one person since the short forms are the
3092 same. (This required a bit of code, as the output is generated
3093 automatically from a table. It wasn't just a text change.)
3094
3095 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3096 option but starts with a hyphen. Could be a pattern or a path name
3097 starting with a hyphen, for instance.
3098
3099 (h) "-" can be given as a file name to represent stdin.
3100
3101 (i) When file names are being printed, "(standard input)" is used for
3102 the standard input, for compatibility with GNU grep. Previously
3103 "<stdin>" was used.
3104
3105 (j) The option --label=xxx can be used to supply a name to be used for
3106 stdin when file names are being printed. There is no short form.
3107
3108 (k) Re-factored the options decoding logic because we are going to add
3109 two more options that take data. Such options can now be given in four
3110 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3111
3112 (l) Added the -A, -B, and -C options for requesting that lines of context
3113 around matches be printed.
3114
3115 (m) Added the -L option to print the names of files that do not contain
3116 any matching lines, that is, the complement of -l.
3117
3118 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3119 continue to scan other files.
3120
3121 (o) The -s option was incorrectly implemented. For compatibility with other
3122 greps, it now suppresses the error message for a non-existent or non-
3123 accessible file (but not the return code). There is a new option called
3124 -q that suppresses the output of matching lines, which was what -s was
3125 previously doing.
3126
3127 (p) Added --include and --exclude options to specify files for inclusion
3128 and exclusion when recursing.
3129
3130 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3131 Hopefully, it now does.
3132
3133 12. Missing cast in pcre_study().
3134
3135 13. Added an "uninstall" target to the makefile.
3136
3137 14. Replaced "extern" in the function prototypes in Makefile.in with
3138 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3139 world, but is set differently for Windows.
3140
3141 15. Added a second compiling function called pcre_compile2(). The only
3142 difference is that it has an extra argument, which is a pointer to an
3143 integer error code. When there is a compile-time failure, this is set
3144 non-zero, in addition to the error test pointer being set to point to an
3145 error message. The new argument may be NULL if no error number is required
3146 (but then you may as well call pcre_compile(), which is now just a
3147 wrapper). This facility is provided because some applications need a
3148 numeric error indication, but it has also enabled me to tidy up the way
3149 compile-time errors are handled in the POSIX wrapper.
3150
3151 16. Added VPATH=.libs to the makefile; this should help when building with one
3152 prefix path and installing with another. (Or so I'm told by someone who
3153 knows more about this stuff than I do.)
3154
3155 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3156 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3157 match everything, including newlines. This is not POSIX-compatible, but
3158 somebody wanted the feature. From pcretest it can be activated by using
3159 both the P and the s flags.
3160
3161 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3162
3163 19. libpcre.pc was being incorrectly installed as executable.
3164
3165 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3166 it now also looks for '\r' so that it will work unmodified on Windows.
3167
3168 21. Added Google's contributed C++ wrapper to the distribution.
3169
3170 22. Added some untidy missing memory free() calls in pcretest, to keep
3171 Electric Fence happy when testing.
3172
3173
3174
3175 Version 5.0 13-Sep-04
3176 ---------------------
3177
3178 1. Internal change: literal characters are no longer packed up into items
3179 containing multiple characters in a single byte-string. Each character
3180 is now matched using a separate opcode. However, there may be more than one
3181 byte in the character in UTF-8 mode.
3182
3183 2. The pcre_callout_block structure has two new fields: pattern_position and
3184 next_item_length. These contain the offset in the pattern to the next match
3185 item, and its length, respectively.
3186
3187 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3188 insertion of callouts before each pattern item. Added the /C option to
3189 pcretest to make use of this.
3190
3191 4. On the advice of a Windows user, the lines
3192
3193 #if defined(_WIN32) || defined(WIN32)
3194 _setmode( _fileno( stdout ), 0x8000 );
3195 #endif /* defined(_WIN32) || defined(WIN32) */
3196
3197 have been added to the source of pcretest. This apparently does useful
3198 magic in relation to line terminators.
3199
3200 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3201 for the benefit of those environments where the "b" makes a difference.
3202
3203 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3204 to know about it. I have put a hack into configure.in that adds in code
3205 to set GCC=yes if CC=icc. This seems to end up at a point in the
3206 generated configure script that is early enough to affect the setting of
3207 compiler options, which is what is needed, but I have no means of testing
3208 whether it really works. (The user who reported this had patched the
3209 generated configure script, which of course I cannot do.)
3210
3211 LATER: After change 22 below (new libtool files), the configure script
3212 seems to know about icc (and also ecc). Therefore, I have commented out
3213 this hack in configure.in.
3214
3215 7. Added support for pkg-config (2 patches were sent in).
3216
3217 8. Negated POSIX character classes that used a combination of internal tables
3218 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3219 [[:^ascii]]. Typically, they would match almost any characters. The other
3220 POSIX classes were not broken in this way.
3221
3222 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3223 to find the match, as PCRE was deluded into thinking that the match had to
3224 start at the start point or following a newline. The same bug applied to
3225 patterns with negative forward assertions or any backward assertions
3226 preceding ".*" at the start, unless the pattern required a fixed first
3227 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3228
3229 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3230 starting at the last subject character, bytes beyond the end of the subject
3231 string were read.
3232
3233 11. Renamed the variable "class" as "classbits" to make life easier for C++
3234 users. (Previously there was a macro definition, but it apparently wasn't
3235 enough.)
3236
3237 12. Added the new field "tables" to the extra data so that tables can be passed
3238 in at exec time, or the internal tables can be re-selected. This allows
3239 a compiled regex to be saved and re-used at a later time by a different
3240 program that might have everything at different addresses.
3241
3242 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3243 -R library as well as a -L library.
3244
3245 14. The debugging options of pcretest (-d on the command line or D on a
3246 pattern) showed incorrect output for anything following an extended class
3247 that contained multibyte characters and which was followed by a quantifier.
3248
3249 15. Added optional support for general category Unicode character properties
3250 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3251 support. It adds about 90K to the size of the library. The meanings of the
3252 inbuilt class escapes such as \d and \s have NOT been changed.
3253
3254 16. Updated pcredemo.c to include calls to free() to release the memory for the
3255 compiled pattern.
3256
3257 17. The generated file chartables.c was being created in the source directory
3258 instead of in the building directory. This caused the build to fail if the
3259 source directory was different from the building directory, and was
3260 read-only.
3261
3262 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3263 file. No doubt somebody will tell me if they don't make sense... Also added
3264 Dan Mooney's comments about building on OpenVMS.
3265
3266 19. Added support for partial matching via the PCRE_PARTIAL option for
3267 pcre_exec() and the \P data escape in pcretest.
3268
3269 20. Extended pcretest with 3 new pattern features:
3270
3271 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3272 write the compiled pattern to the file whose name is "rest-of-line".
3273 This is a straight binary dump of the data, with the saved pointer to
3274 the character tables forced to be NULL. The study data, if any, is
3275 written too. After writing, pcretest reads a new pattern.
3276
3277 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3278 compiled pattern from the given file. There must not be any
3279 occurrences of "<" in the file name (pretty unlikely); if there are,
3280 pcretest will instead treat the initial "<" as a pattern delimiter.
3281 After reading in the pattern, pcretest goes on to read data lines as
3282 usual.
3283
3284 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3285 and 16-bit fields in a compiled pattern, to simulate a pattern that
3286 was compiled on a host of opposite endianness.
3287
3288 21. The pcre-exec() function can now cope with patterns that were compiled on
3289 hosts of opposite endianness, with this restriction:
3290
3291 As for any compiled expression that is saved and used later, the tables
3292 pointer field cannot be preserved; the extra_data field in the arguments
3293 to pcre_exec() should be used to pass in a tables address if a value
3294 other than the default internal tables were used at compile time.
3295
3296 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3297 now diagnosed as an error. Previously, most of the time, a negative number
3298 would have been treated as zero, but if in addition "ovector" was passed as
3299 NULL, a crash could occur.
3300
3301 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3302 new versions from the libtool 1.5 distribution (the last one is a copy of
3303 a file called libtool.m4). This seems to have fixed the need to patch
3304 "configure" to support Darwin 1.3 (which I used to do). However, I still
3305 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3306 workstation).
3307
3308 24. Changed the PCRE licence to be the more standard "BSD" licence.
3309
3310
3311 Version 4.5 01-Dec-03
3312 ---------------------
3313
3314 1. There has been some re-arrangement of the code for the match() function so
3315 that it can be compiled in a version that does not call itself recursively.
3316 Instead, it keeps those local variables that need separate instances for
3317 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3318 needs to "recurse". Keeping track of where control must go is done by means
3319 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3320 hide most of the details from the main code, and operates only if
3321 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3322 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3323 operating.
3324
3325 To make it easier for callers to provide specially tailored get/free
3326 functions for this usage, two new functions, pcre_stack_malloc, and
3327 pcre_stack_free, are used. They are always called in strict stacking order,
3328 and the size of block requested is always the same.
3329
3330 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3331 PCRE has been compiled to use the stack or the heap for recursion. The
3332 -C option of pcretest uses this to show which version is compiled.
3333
3334 A new data escape \S, is added to pcretest; it causes the amounts of store
3335 obtained and freed by both kinds of malloc/free at match time to be added
3336 to the output.
3337
3338 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3339 what's available on my current Linux desktop machine.
3340
3341 3. When matching a UTF-8 string, the test for a valid string at the start has
3342 been extended. If start_offset is not zero, PCRE now checks that it points
3343 to a byte that is the start of a UTF-8 character. If not, it returns
3344 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3345 this is necessary because there may be backward assertions in the pattern.
3346 When matching the same subject several times, it may save resources to use
3347 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3348
3349 4. The code for checking the validity of UTF-8 strings has been tightened so
3350 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3351 containing "overlong sequences".
3352
3353 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3354 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3355 should have been "&", but it just so happened that all the cases this let
3356 through by mistake were picked up later in the function.
3357
3358 6. I had used a variable called "isblank" - this is a C99 function, causing
3359 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3360
3361 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3362 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3363 so the version is shown; (c) stop "make test" echoing "./RunTest".
3364
3365 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3366
3367 9. The prototype for memmove() for systems that don't have it was using
3368 size_t, but the inclusion of the header that defines size_t was later. I've
3369 moved the #includes for the C headers earlier to avoid this.
3370
3371 10. Added some adjustments to the code to make it easier to compiler on certain
3372 special systems:
3373
3374 (a) Some "const" qualifiers were missing.
3375 (b) Added the macro EXPORT before all exported functions; by default this
3376 is defined to be empty.
3377 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3378 that it reads its output file name as an argument instead of writing
3379 to the standard output and assuming this can be redirected.
3380
3381 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3382 class containing characters with values greater than 255, PCRE compilation
3383 went into a loop.
3384
3385 12. A recursive reference to a subpattern that was within another subpattern
3386 that had a minimum quantifier of zero caused PCRE to crash. For example,
3387 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3388 recursion. If the recursively-called subpattern itself had a zero repeat,
3389 that was OK.
3390
3391 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3392 buffer into which it was copied (for escape processing) was still set at
3393 1024, so long lines caused crashes.
3394
3395 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3396 "internal error: code overflow...". This applied to any character class
3397 that was followed by a possessive quantifier.
3398
3399 15. Modified the Makefile to add libpcre.la as a prerequisite for
3400 libpcreposix.la because I was told this is needed for a parallel build to
3401 work.
3402
3403 16. If a pattern that contained .* following optional items at the start was
3404 studied, the wrong optimizing data was generated, leading to matching
3405 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3406 matching string must start with a or b or c. The correct conclusion for
3407 this pattern is that a match can start with any character.
3408
3409
3410 Version 4.4 13-Aug-03
3411 ---------------------
3412
3413 1. In UTF-8 mode, a character class containing characters with values between
3414 127 and 255 was not handled correctly if the compiled pattern was studied.
3415 In fixing this, I have also improved the studying algorithm for such
3416 classes (slightly).
3417
3418 2. Three internal functions had redundant arguments passed to them. Removal
3419 might give a very teeny performance improvement.
3420
3421 3. Documentation bug: the value of the capture_top field in a callout is *one
3422 more than* the number of the hightest numbered captured substring.
3423
3424 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3425 in incorrectly linking with a previously installed version. They now link
3426 explicitly with libpcre.la.
3427
3428 5. configure.in no longer needs to recognize Cygwin specially.
3429
3430 6. A problem in pcre.in for Windows platforms is fixed.
3431
3432 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3433 pcretest, it used to include the size of the study block as part of its
3434 output. Unfortunately, the structure contains a field that has a different
3435 size on different hardware architectures. This meant that the tests that
3436 showed this size failed. As the block is currently always of a fixed size,
3437 this information isn't actually particularly useful in pcretest output, so
3438 I have just removed it.
3439
3440 8. Three pre-processor statements accidentally did not start in column 1.
3441 Sadly, there are *still* compilers around that complain, even though
3442 standard C has not required this for well over a decade. Sigh.
3443
3444 9. In pcretest, the code for checking callouts passed small integers in the
3445 callout_data field, which is a void * field. However, some picky compilers
3446 complained about the casts involved for this on 64-bit systems. Now
3447 pcretest passes the address of the small integer instead, which should get
3448 rid of the warnings.
3449
3450 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3451 both compile and run time, and gives an error if an invalid UTF-8 sequence
3452 is found. There is a option for disabling this check in cases where the
3453 string is known to be correct and/or the maximum performance is wanted.
3454
3455 11. In response to a bug report, I changed one line in Makefile.in from
3456
3457 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3458 to
3459 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3460
3461 to look similar to other lines, but I have no way of telling whether this
3462 is the right thing to do, as I do not use Windows. No doubt I'll get told
3463 if it's wrong...
3464
3465
3466 Version 4.3 21-May-03
3467 ---------------------
3468
3469 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3470 Makefile.
3471
3472 2. Some refactoring to improve the quality of the code:
3473
3474 (i) The utf8_table... variables are now declared "const".
3475
3476 (ii) The code for \cx, which used the "case flipping" table to upper case
3477 lower case letters, now just substracts 32. This is ASCII-specific,
3478 but the whole concept of \cx is ASCII-specific, so it seems
3479 reasonable.
3480
3481 (iii) PCRE was using its character types table to recognize decimal and
3482 hexadecimal digits in the pattern. This is silly, because it handles
3483 only 0-9, a-f, and A-F, but the character types table is locale-
3484 specific, which means strange things might happen. A private
3485 table is now used for this - though it costs 256 bytes, a table is
3486 much faster than multiple explicit tests. Of course, the standard
3487 character types table is still used for matching digits in subject
3488 strings against \d.
3489
3490 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3491 ending in _t are). So I've renamed it as ESC_tee.
3492
3493 3. The first argument for regexec() in the POSIX wrapper should have been
3494 defined as "const".
3495
3496 4. Changed pcretest to use malloc() for its buffers so that they can be
3497 Electric Fenced for debugging.
3498
3499 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3500 to read one or more bytes before the start of the subject string. Often this
3501 had no effect on PCRE's behaviour, but in some circumstances it could
3502 provoke a segmentation fault.
3503
3504 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3505 to try to read one or more bytes before the start of the subject string.
3506
3507 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3508 UTF-8 support could misbehave in various ways if the subject string
3509 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3510 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3511 back over UTF-8 characters.)
3512
3513
3514 Version 4.2 14-Apr-03
3515 ---------------------
3516
3517 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3518
3519 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3520 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3521 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3522 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3523 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3524 and BUILD_EXEEXT
3525 Note: automatic setting of the BUILD variables is not yet working
3526 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3527 compile-time but not at link-time
3528 [LINK]: use for linking executables only
3529 make different versions for Windows and non-Windows
3530 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3531 libraries
3532 [LINK_FOR_BUILD]: new variable
3533 [OBJEXT]: use throughout
3534 [EXEEXT]: use throughout
3535 <winshared>: new target
3536 <wininstall>: new target
3537 <dftables.o>: use native compiler
3538 <dftables>: use native linker
3539 <install>: handle Windows platform correctly
3540 <clean>: ditto
3541 <check>: ditto
3542 copy DLL to top builddir before testing
3543
3544 As part of these changes, -no-undefined was removed again. This was reported
3545 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3546 in any case.
3547
3548 3. Some tidies to get rid of compiler warnings:
3549
3550 . In the match_data structure, match_limit was an unsigned long int, whereas
3551 match_call_count was an int. I've made them both unsigned long ints.
3552
3553 . In pcretest the fact that a const uschar * doesn't automatically cast to
3554 a void * provoked a warning.
3555
3556 . Turning on some more compiler warnings threw up some "shadow" variables
3557 and a few more missing casts.
3558
3559 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3560 option, a class that contained a single character with a value between 128
3561 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3562
3563 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3564 option, a class that contained several characters, but with at least one
3565 whose value was between 128 and 255 caused PCRE to crash.
3566
3567
3568 Version 4.1 12-Mar-03
3569 ---------------------
3570
3571 1. Compiling with gcc -pedantic found a couple of places where casts were
3572 needed, and a string in dftables.c that was longer than standard compilers are
3573 required to support.
3574
3575 2. Compiling with Sun's compiler found a few more places where the code could
3576 be tidied up in order to avoid warnings.
3577
3578 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3579 first of these names is deprecated in the latest Autoconf in favour of the name
3580 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3581 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3582 analogy I have changed it to CFLAGS_FOR_BUILD.
3583
3584 4. Added -no-undefined to the linking command in the Makefile, because this is
3585 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3586 linking step for the pcreposix library.
3587
3588 5. PCRE was failing to diagnose the case of two named groups with the same
3589 name.
3590
3591 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3592 literal character that is needed in the subject for a match, and scans along to
3593 ensure that it is present before embarking on the full matching process. This
3594 saves time in cases of nested unlimited repeats that are never going to match.
3595 Problem: the scan can take a lot of time if the subject is very long (e.g.
3596 megabytes), thus penalizing straightforward matches. It is now done only if the
3597 amount of subject to be scanned is less than 1000 bytes.
3598
3599 7. A lesser problem with the same optimization is that it was recording the
3600 first character of an anchored pattern as "needed", thus provoking a search
3601 right along the subject, even when the first match of the pattern was going to
3602 fail. The "needed" character is now not set for anchored patterns, unless it
3603 follows something in the pattern that is of non-fixed length. Thus, it still
3604 fulfils its original purpose of finding quick non-matches in cases of nested
3605 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3606
3607
3608 Version 4.0 17-Feb-03
3609 ---------------------
3610
3611 1. If a comment in an extended regex that started immediately after a meta-item
3612 extended to the end of string, PCRE compiled incorrect data. This could lead to
3613 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3614
3615 2. Moved to autoconf 2.53 and libtool 1.4.2.
3616
3617 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3618 the special perltest8 script is no longer needed - all the tests can be run
3619 from a single perltest script.
3620
3621 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3622 by \s. It has now been removed in PCRE. This means it isn't recognized as
3623 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3624 class [:space:] *does* include VT, thereby creating a mess.
3625
3626 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3627 space and tab.
3628
3629 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3630 its new features into the main test script, reducing the number of scripts.
3631
3632 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3633 were backward compatible, and made the (?i) apply to the whole pattern, as if
3634 /i were given. Now it behaves more logically, and applies the option setting
3635 only to what follows. PCRE has been changed to follow suit. However, if it
3636 finds options settings right at the start of the pattern, it extracts them into
3637 the global options, as before. Thus, they show up in the info data.
3638
3639 8. Added support for the \Q...\E escape sequence. Characters in between are
3640 treated as literals. This is slightly different from Perl in that $ and @ are
3641 also handled as literals inside the quotes. In Perl, they will cause variable
3642 interpolation. Note the following examples:
3643
3644 Pattern PCRE matches Perl matches
3645
3646 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3647 \Qabc\$xyz\E abc\$xyz abc\$xyz
3648 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3649
3650 For compatibility with Perl, \Q...\E sequences are recognized inside character
3651 classes as well as outside them.
3652
3653 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3654 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3655 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3656 signed/unsigned warnings.
3657
3658 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3659 option for pcretest, so I've replaced it by a simple function that does just
3660 that job.
3661
3662 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3663 "pcregrep -".
3664
3665 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3666 Java package. This provides some syntactic sugar for simple cases of what my
3667 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3668 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3669 item, you can use this simplified notation. Note that only makes sense with
3670 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3671 greediness, whatever the setting of the PCRE_UNGREEDY option.
3672
3673 13. A change of greediness default within a pattern was not taking effect at
3674 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3675 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3676 was abstracted outside.
3677
3678 14. PCRE now supports the \G assertion. It is true when the current matching
3679 position is at the start point of the match. This differs from \A when the
3680 starting offset is non-zero. Used with the /g option of pcretest (or similar
3681 code), it works in the same way as it does for Perl's /g option. If all
3682 alternatives of a regex begin with \G, the expression is anchored to the start
3683 match position, and the "anchored" flag is set in the compiled expression.
3684
3685 15. Some bugs concerning the handling of certain option changes within patterns
3686 have been fixed. These applied to options other than (?ims). For example,
3687 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3688 been the other way round. Some of this was related to change 7 above.
3689
3690 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3691 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3692 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3693 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3694
3695 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3696 mode. Unlike ".", it always matches newline, whatever the setting of
3697 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3698 assertions. Perl allows it, but it doesn't (in general) work because it can't
3699 calculate the length of the lookbehind. At least, that's the case for Perl
3700 5.8.0 - I've been told they are going to document that it doesn't work in
3701 future.
3702
3703 18. Added an error diagnosis for escapes that PCRE does not support: these are
3704 \L, \l, \N, \P, \p, \U, \u, and \X.
3705
3706 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3707 reading past the end of the pattern in cases such as /[abcd/.
3708
3709 20. PCRE was getting more memory than necessary for patterns with classes that
3710 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3711
3712 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3713 compiling PCRE for use with Virtual Pascal.
3714
3715 22. Small fix to the Makefile to make it work properly if the build is done
3716 outside the source tree.
3717
3718 23. Added a new extension: a condition to go with recursion. If a conditional
3719 subpattern starts with (?(R) the "true" branch is used if recursion has
3720 happened, whereas the "false" branch is used only at the top level.
3721
3722 24. When there was a very long string of literal characters (over 255 bytes
3723 without UTF support, over 250 bytes with UTF support), the computation of how
3724 much memory was required could be incorrect, leading to segfaults or other
3725 strange effects.
3726
3727 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3728 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3729 there was a subsequent back reference to those brackets. This meant that, for
3730 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3731 possible to check for precisely this case. All we can do is abandon the
3732 optimization if .* occurs inside capturing brackets when there are any back
3733 references whatsoever. (See below for a better fix that came later.)
3734
3735 26. The handling of the optimization for finding the first character of a
3736 non-anchored pattern, and for finding a character that is required later in the
3737 match were failing in some cases. This didn't break the matching; it just
3738 failed to optimize when it could. The way this is done has been re-implemented.
3739
3740 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3741
3742 28. Added a new feature that provides some of the functionality that Perl
3743 provides with (?{...}). The facility is termed a "callout". The way it is done
3744 in PCRE is for the caller to provide an optional function, by setting
3745 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3746 global variable. By default it is unset, which disables all calling out. To get
3747 the function called, the regex must include (?C) at appropriate points. This
3748 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3749 This provides a means of identifying different callout points. When PCRE
3750 reaches such a point in the regex, if pcre_callout has been set, the external
3751 function is called. It is provided with data in a structure called
3752 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3753 matching continues; if it returns a non-zero value, the match at the current
3754 point fails. However, backtracking will occur if possible. [This was changed
3755 later and other features added - see item 49 below.]
3756
3757 29. pcretest is upgraded to test the callout functionality. It provides a
3758 callout function that displays information. By default, it shows the start of
3759 the match and the current position in the text. There are some new data escapes
3760 to vary what happens:
3761
3762 \C+ in addition, show current contents of captured substrings
3763 \C- do not supply a callout function
3764 \C!n return 1 when callout number n is reached
3765 \C!n!m return 1 when callout number n is reached for the mth time
3766
3767 30. If pcregrep was called with the -l option and just a single file name, it
3768 output "<stdin>" if a match was found, instead of the file name.
3769
3770 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3771 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3772 pcre_exec(). This saves a malloc/free per call. The default value of
3773 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3774 when configuring.
3775
3776 32. The default maximum size of a compiled pattern is 64K. There have been a
3777 few cases of people hitting this limit. The code now uses macros to handle the
3778 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3779 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3780 configuring. Tests 2 and 5 work only with 2-byte links because they output
3781 debugging information about compiled patterns.
3782
3783 33. Internal code re-arrangements:
3784
3785 (a) Moved the debugging function for printing out a compiled regex into
3786 its own source file (printint.c) and used #include to pull it into
3787 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3788 separate copies.
3789
3790 (b) Defined the list of op-code names for debugging as a macro in
3791 internal.h so that it is next to the definition of the opcodes.
3792
3793 (c) Defined a table of op-code lengths for simpler skipping along compiled
3794 code. This is again a macro in internal.h so that it is next to the
3795 definition of the opcodes.
3796
3797 34. Added support for recursive calls to individual subpatterns, along the
3798 lines of Robin Houston's patch (but implemented somewhat differently).
3799
3800 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3801 allow it to read and process whole directories in Win32. This code was
3802 contributed by Lionel Fourquaux; it has not been tested by me.
3803
3804 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3805 used to name a group. Names consist of alphanumerics and underscores, and must
3806 be unique. Back references use the syntax (?P=name) and recursive calls use
3807 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3808 numbers. The function pcre_fullinfo() can be used after compilation to extract
3809 a name/number map. There are three relevant calls:
3810
3811 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3812 PCRE_INFO_NAMECOUNT yields the number of entries
3813 PCRE_INFO_NAMETABLE yields a pointer to the map.
3814
3815 The map is a vector of fixed-size entries. The size of each entry depends on
3816 the length of the longest name used. The first two bytes of each entry are the
3817 group number, most significant byte first. There follows the corresponding
3818 name, zero terminated. The names are in alphabetical order.
3819
3820 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3821 case instead of 255. Making it the same both with and without UTF-8 support
3822 means that the same test output works with both.
3823
3824 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3825 calling malloc() with a zero argument.
3826
3827 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3828 optimization. I've improved things by keeping a bitmap of backreferences with
3829 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3830 fact referenced, the optimization can be applied. It is unlikely that a
3831 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3832 the match to follow \n) will appear inside brackets with a number greater than
3833 31, but if it does, any back reference > 31 suppresses the optimization.
3834
3835 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3836 of disabling numbered capturing parentheses. Any opening parenthesis that is
3837 not followed by ? behaves as if it were followed by ?: but named parentheses
3838 can still be used for capturing (and they will acquire numbers in the usual
3839 way).
3840
3841 41. Redesigned the return codes from the match() function into yes/no/error so
3842 that errors can be passed back from deep inside the nested calls. A malloc
3843 failure while inside a recursive subpattern call now causes the
3844 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3845
3846 42. It is now possible to set a limit on the number of times the match()
3847 function is called in a call to pcre_exec(). This facility makes it possible to
3848 limit the amount of recursion and backtracking, though not in a directly
3849 obvious way, because the match() function is used in a number of different
3850 circumstances. The count starts from zero for each position in the subject
3851 string (for non-anchored patterns). The default limit is, for compatibility, a
3852 large number, namely 10 000 000. You can change this in two ways:
3853
3854 (a) When configuring PCRE before making, you can use --with-match-limit=n
3855 to set a default value for the compiled library.
3856
3857 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3858 a different value is set. See 45 below.
3859
3860 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3861
3862 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3863 of things that can be changed at compile time. The first argument specifies
3864 what is wanted and the second points to where the information is to be placed.
3865 The current list of available information is:
3866
3867 PCRE_CONFIG_UTF8
3868
3869 The output is an integer that is set to one if UTF-8 support is available;
3870 otherwise it is set to zero.
3871
3872 PCRE_CONFIG_NEWLINE
3873
3874 The output is an integer that it set to the value of the code that is used for
3875 newline. It is either LF (10) or CR (13).
3876
3877 PCRE_CONFIG_LINK_SIZE
3878
3879 The output is an integer that contains the number of bytes used for internal
3880 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3881
3882 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3883
3884 The output is an integer that contains the threshold above which the POSIX
3885 interface uses malloc() for output vectors. See item 31 above.
3886
3887 PCRE_CONFIG_MATCH_LIMIT
3888
3889 The output is an unsigned integer that contains the default limit of the number
3890 of match() calls in a pcre_exec() execution. See 42 above.
3891
3892 44. pcretest has been upgraded by the addition of the -C option. This causes it
3893 to extract all the available output from the new pcre_config() function, and to
3894 output it. The program then exits immediately.
3895
3896 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3897 order to support additional features. One way would have been to define
3898 pcre_exec2() (for example) with extra arguments, but this would not have been
3899 extensible, and would also have required all calls to the original function to
3900 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3901 is used for passing in "extra" data from pcre_study().
3902
3903 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3904 contains the following fields:
3905
3906 flags a bitmap indicating which of the following fields are set
3907 study_data opaque data from pcre_study()
3908 match_limit a way of specifying a limit on match() calls for a specific
3909 call to pcre_exec()
3910 callout_data data for callouts (see 49 below)
3911
3912 The flag bits are also defined in pcre.h, and are
3913
3914 PCRE_EXTRA_STUDY_DATA
3915 PCRE_EXTRA_MATCH_LIMIT
3916 PCRE_EXTRA_CALLOUT_DATA
3917
3918 The pcre_study() function now returns one of these new pcre_extra blocks, with
3919 the actual study data pointed to by the study_data field, and the
3920 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3921 before. That is, this change is entirely upwards-compatible and requires no
3922 change to existing code.
3923
3924 If you want to pass in additional data to pcre_exec(), you can either place it
3925 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3926 block.
3927
3928 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3929 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3930 times with different match limits, until it finds the minimum value needed for
3931 pcre_exec() to complete. The value is then output. This can be instructive; for
3932 most simple matches the number is quite small, but for pathological cases it
3933 gets very large very quickly.
3934
3935 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3936 returns the size of the data block pointed to by the study_data field in a
3937 pcre_extra block, that is, the value that was passed as the argument to
3938 pcre_malloc() when PCRE was getting memory in which to place the information
3939 created by pcre_study(). The fourth argument should point to a size_t variable.
3940 pcretest has been extended so that this information is shown after a successful
3941 pcre_study() call when information about the compiled regex is being displayed.
3942
3943 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3944 because what follows is always an absolute path. (Later: it turns out that this
3945 is more than cosmetic for MinGW, because it doesn't like empty path
3946 components.)
3947
3948 49. Some changes have been made to the callout feature (see 28 above):
3949
3950 (i) A callout function now has three choices for what it returns:
3951
3952 0 => success, carry on matching
3953 > 0 => failure at this point, but backtrack if possible
3954 < 0 => serious error, return this value from pcre_exec()
3955
3956 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3957 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3958 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3959 use by callout functions. It will never be used by PCRE itself.
3960
3961 (ii) The pcre_extra structure (see 45 above) has a void * field called
3962 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3963 pcre_callout_block structure has a field of the same name. The contents of
3964 the field passed in the pcre_extra structure are passed to the callout
3965 function in the corresponding field in the callout block. This makes it
3966 easier to use the same callout-containing regex from multiple threads. For
3967 testing, the pcretest program has a new data escape
3968
3969 \C*n pass the number n (may be negative) as callout_data
3970
3971 If the callout function in pcretest receives a non-zero value as
3972 callout_data, it returns that value.
3973
3974 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3975 there were some redundant $(CFLAGS) in commands that are now specified as
3976 $(LINK), which already includes $(CFLAGS).
3977
3978 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3979 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3980 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3981 one-byte characters throughout. Note that case-insensitive matching applies
3982 only to characters whose values are less than 256. PCRE doesn't support the
3983 notion of cases for higher-valued characters.
3984
3985 (i) A character class whose characters are all within 0-255 is handled as
3986 a bit map, and the map is inverted for negative classes. Previously, a
3987 character > 255 always failed to match such a class; however it should
3988 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3989
3990 (ii) A negated character class with a single character < 255 is coded as
3991 "not this character" (OP_NOT). This wasn't working properly when the test
3992 character was multibyte, either singly or repeated.
3993
3994 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3995 mode, for example: \x{100}{2,3}.
3996
3997 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3998 singly or repeated) now correctly test multibyte characters. However,
3999 PCRE doesn't recognize any characters with values greater than 255 as
4000 digits, spaces, or word characters. Such characters always match \D, \S,
4001 and \W, and never match \d, \s, or \w.
4002
4003 (v) Classes may now contain characters and character ranges with values
4004 greater than 255. For example: [ab\x{100}-\x{400}].
4005
4006 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
4007 PCRE in UTF-8 mode.
4008
4009 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
4010 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
4011 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
4012 value.)
4013
4014 53. The single man page has become too large. I have therefore split it up into
4015 a number of separate man pages. These also give rise to individual HTML pages;
4016 these are now put in a separate directory, and there is an index.html page that
4017 lists them all. Some hyperlinking between the pages has been installed.
4018
4019 54. Added convenience functions for handling named capturing parentheses.
4020
4021 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4022 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4023 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4024 are faulted.
4025
4026 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4027 calling configure. These values are used when compiling the dftables.c program
4028 which is run to generate the source of the default character tables. They
4029 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4030 you will need to set these values.
4031
4032 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4033
4034
4035 Version 3.9 02-Jan-02
4036 ---------------------
4037
4038 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4039
4040 2. If --disable-static was given, the building process failed when trying to
4041 build pcretest and pcregrep. (For some reason it was using libtool to compile
4042 them, which is not right, as they aren't part of the library.)
4043
4044
4045 Version 3.8 18-Dec-01
4046 ---------------------
4047
4048 1. The experimental UTF-8 code was completely screwed up. It was packing the
4049 bytes in the wrong order. How dumb can you get?
4050
4051
4052 Version 3.7 29-Oct-01
4053 ---------------------
4054
4055 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4056 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4057 this didn't happen under Solaris 8, where I normally test things.
4058
4059 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4060 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4061 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4062 this fix an hour or so after the initial 3.7 release.)
4063
4064
4065 Version 3.6 23-Oct-01
4066 ---------------------
4067
4068 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4069 offsets passed as NULL with zero offset count.
4070
4071 2. The config.guess and config.sub files had not been updated when I moved to
4072 the latest autoconf.
4073
4074
4075 Version 3.5 15-Aug-01
4076 ---------------------
4077
4078 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4079 had been forgotten.
4080
4081 2. By using declared but undefined structures, we can avoid using "void"
4082 definitions in pcre.h while keeping the internal definitions of the structures
4083 private.
4084
4085 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4086 user point of view, this means that both static and shared libraries are built
4087 by default, but this can be individually controlled. More of the work of
4088 handling this static/shared cases is now inside libtool instead of PCRE's make
4089 file.
4090
4091 4. The pcretest utility is now installed along with pcregrep because it is
4092 useful for users (to test regexs) and by doing this, it automatically gets
4093 relinked by libtool. The documentation has been turned into a man page, so
4094 there are now .1, .txt, and .html versions in /doc.
4095
4096 5. Upgrades to pcregrep:
4097 (i) Added long-form option names like gnu grep.
4098 (ii) Added --help to list all options with an explanatory phrase.
4099 (iii) Added -r, --recursive to recurse into sub-directories.
4100 (iv) Added -f, --file to read patterns from a file.
4101
4102 6. pcre_exec() was referring to its "code" argument before testing that
4103 argument for NULL (and giving an error if it was NULL).
4104
4105 7. Upgraded Makefile.in to allow for compiling in a different directory from
4106 the source directory.
4107
4108 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4109 options bits, the pointer it was passed was to an int instead of to an unsigned
4110 long int. This mattered only on 64-bit systems.
4111
4112 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4113 generated) instead of pcre.in, which it its source. Also made the same change
4114 in several of the .c files.
4115
4116 10. A new release of gcc defines printf() as a macro, which broke pcretest
4117 because it had an ifdef in the middle of a string argument for printf(). Fixed
4118 by using separate calls to printf().
4119
4120 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4121 script, to force use of CR or LF instead of \n in the source. On non-Unix
4122 systems, the value can be set in config.h.
4123
4124 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4125 absolute limit. Changed the text of the error message to make this clear, and
4126 likewise updated the man page.
4127
4128 13. The limit of 99 on the number of capturing subpatterns has been removed.
4129 The new limit is 65535, which I hope will not be a "real" limit.
4130
4131
4132 Version 3.4 22-Aug-00
4133 ---------------------
4134
4135 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4136
4137 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4138
4139
4140 Version 3.3 01-Aug-00
4141 ---------------------
4142
4143 1. If an octal character was given, but the value was greater than \377, it
4144 was not getting masked to the least significant bits, as documented. This could
4145 lead to crashes in some systems.
4146
4147 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4148 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4149
4150 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4151 These just pass their arguments on to (pcre_free)(), but they are provided
4152 because some uses of PCRE bind it to non-C systems that can call its functions,
4153 but cannot call free() or pcre_free() directly.
4154
4155 4. Add "make test" as a synonym for "make check". Corrected some comments in
4156 the Makefile.
4157
4158 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4159 Makefile.
4160
4161 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4162 command called pgrep for grepping around the active processes.
4163
4164 7. Added the beginnings of support for UTF-8 character strings.
4165
4166 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4167 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4168 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4169 out for the ar command.)
4170
4171
4172 Version 3.2 12-May-00
4173 ---------------------
4174
4175 This is purely a bug fixing release.
4176
4177 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4178 of ZA. This was just one example of several cases that could provoke this bug,
4179 which was introduced by change 9 of version 2.00. The code for breaking
4180 infinite loops after an iteration that matches an empty string was't working
4181 correctly.
4182
4183 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4184 when matched against abbab (for example). After matching an empty string, it
4185 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4186 caused it to match further down the string than it should.
4187
4188 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4189 was there because it doesn't seem to be needed, and it causes trouble on some
4190 systems, as it is not a Standard C header. It has been removed.
4191
4192 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4193 were reported on the Macintosh. The changes were from
4194
4195 while ((c = *(++ptr)) != 0 && c != '\n');
4196 to
4197 while ((c = *(++ptr)) != 0 && c != '\n') ;
4198
4199 Totally extraordinary, but if that's what it takes...
4200
4201 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4202 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4203 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4204 assumes the way PCRE uses memmove() (always moving upwards).
4205
4206 6. PCRE is being used in one environment where strchr() is not available. There
4207 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4208 faster code anyway.
4209
4210
4211 Version 3.1 09-Feb-00
4212 ---------------------
4213
4214 The only change in this release is the fixing of some bugs in Makefile.in for
4215 the "install" target:
4216
4217 (1) It was failing to install pcreposix.h.
4218
4219 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4220
4221
4222 Version 3.0 01-Feb-00
4223 ---------------------
4224
4225 1. Add support for the /+ modifier to perltest (to output $` like it does in
4226 pcretest).
4227
4228 2. Add support for the /g modifier to perltest.
4229
4230 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4231 matches null strings.
4232
4233 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4234 pattern. Perl treats empty patterns specially - it reuses the most recent
4235 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4236 effect.
4237
4238 5. The POSIX interface was broken in that it was just handing over the POSIX
4239 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4240 required a bigger vector, with some working space on the end. This means that
4241 the POSIX wrapper now has to get and free some memory, and copy the results.
4242
4243 6. Added some simple autoconf support, placing the test data and the
4244 documentation in separate directories, re-organizing some of the
4245 information files, and making it build pcre-config (a GNU standard). Also added
4246 libtool support for building PCRE as a shared library, which is now the
4247 default.
4248
4249 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4250 09 are not valid octal constants. Single digits will be used for minor values
4251 less than 10.
4252
4253 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4254 existing programs that set these in the POSIX interface can use PCRE without
4255 modification.
4256
4257 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4258 return all that pcre_info() returns, plus additional data. The pcre_info()
4259 function is retained for compatibility, but is considered to be obsolete.
4260
4261 10. Added experimental recursion feature (?R) to handle one common case that
4262 Perl 5.6 will be able to do with (?p{...}).
4263
4264 11. Added support for POSIX character classes like [:alpha:], which Perl is
4265 adopting.
4266
4267
4268 Version 2.08 31-Aug-99
4269 ----------------------
4270
4271 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4272 trying to match at the startoffset position, but instead was moving forward to
4273 the next newline as if a previous match had failed.
4274
4275 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4276 and could get into a loop if a null string was matched other than at the start
4277 of the subject.
4278
4279 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4280 be distinguished at compile time, and for completeness also added PCRE_DATE.
4281
4282 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4283 in GnuWin32 environments.
4284
4285
4286 Version 2.07 29-Jul-99
4287 ----------------------
4288
4289 1. The documentation is now supplied in plain text form and HTML as well as in
4290 the form of man page sources.
4291
4292 2. C++ compilers don't like assigning (void *) values to other pointer types.
4293 In particular this affects malloc(). Although there is no problem in Standard
4294 C, I've put in casts to keep C++ compilers happy.
4295
4296 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4297 should be (const char *).
4298
4299 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4300 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4301 However, I haven't made this a standard facility. The documentation doesn't
4302 mention it, and the Makefile doesn't support it.
4303
4304 5. The Makefile now contains an "install" target, with editable destinations at
4305 the top of the file. The pcretest program is not installed.
4306
4307 6. pgrep -V now gives the PCRE version number and date.
4308
4309 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4310 causing the entire string to be ignored, instead of just the last character.
4311
4312 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4313 non-matching string, it can take a very, very long time, even for strings of
4314 quite modest length, because of the nested recursion. PCRE now does better in
4315 some of these cases. It does this by remembering the last required literal
4316 character in the pattern, and pre-searching the subject to ensure it is present
4317 before running the real match. In other words, it applies a heuristic to detect
4318 some types of certain failure quickly, and in the above example, if presented
4319 with a string that has no trailing " it gives "no match" very quickly.
4320
4321 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4322 other alternatives are tried instead.
4323
4324
4325 Version 2.06 09-Jun-99
4326 ----------------------
4327
4328 1. Change pcretest's output for amount of store used to show just the code
4329 space, because the remainder (the data block) varies in size between 32-bit and
4330 64-bit systems.
4331
4332 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4333 start matching at. This allows lookbehinds to work when searching for multiple
4334 occurrences in a string.
4335
4336 3. Added additional options to pcretest for testing multiple occurrences:
4337
4338 /+ outputs the rest of the string that follows a match
4339 /g loops for multiple occurrences, using the new startoffset argument
4340 /G loops for multiple occurrences by passing an incremented pointer
4341
4342 4. PCRE wasn't doing the "first character" optimization for patterns starting
4343 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4344 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4345 the letter 'x'. On long subject strings, this gives a significant speed-up.
4346
4347
4348 Version 2.05 21-Apr-99
4349 ----------------------
4350
4351 1. Changed the type of magic_number from int to long int so that it works
4352 properly on 16-bit systems.
4353
4354 2. Fixed a bug which caused patterns starting with .* not to work correctly
4355 when the subject string contained newline characters. PCRE was assuming
4356 anchoring for such patterns in all cases, which is not correct because .* will
4357 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4358 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4359 must be retried after every newline in the subject.
4360
4361
4362 Version 2.04 18-Feb-99
4363 ----------------------
4364
4365 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4366 computation of the store needed to hold the pattern was incorrect (too large).
4367 If such patterns were nested a few deep, this could multiply and become a real
4368 problem.
4369
4370 2. Added /M option to pcretest to show the memory requirement of a specific
4371 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4372
4373 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4374 compiled in such a way that the backtracking after subsequent failure was
4375 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4376 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4377
4378
4379 Version 2.03 02-Feb-99
4380 ----------------------
4381
4382 1. Fixed typo and small mistake in man page.
4383
4384 2. Added 4th condition (GPL supersedes if conflict) and created separate
4385 LICENCE file containing the conditions.
4386
4387 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4388 Perl, that is the internal \ allows the delimiter to be included in the
4389 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4390 the final delimiter, add \ to the end of the pattern (to test the error).
4391
4392 4. Added the convenience functions for extracting substrings after a successful
4393 match. Updated pcretest to make it able to test these functions.
4394
4395
4396 Version 2.02 14-Jan-99
4397 ----------------------
4398
4399 1. Initialized the working variables associated with each extraction so that
4400 their saving and restoring doesn't refer to uninitialized store.
4401
4402 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4403 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4404 fix the problem.
4405
4406 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4407 calls, and wasn't printing the correct value for compiling calls. Increased the
4408 default value of LOOPREPEAT, and the number of significant figures in the
4409 times.
4410
4411 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4412
4413 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4414 a building problem on Windows NT with a FAT file system.
4415
4416
4417 Version 2.01 21-Oct-98
4418 ----------------------
4419
4420 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4421 to character tables built by pcre_maketables() in the current locale. If NULL
4422 is passed, the default tables are used.
4423
4424
4425 Version 2.00 24-Sep-98
4426 ----------------------
4427
4428 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4429 it any more.
4430
4431 2. Allow quantification of (?>) groups, and make it work correctly.
4432
4433 3. The first character computation wasn't working for (?>) groups.
4434
4435 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4436 end of the subject) and add 5.005's \z, which really does match only at the
4437 very end of the subject.
4438
4439 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4440
4441 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4442 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4443 localized options. All options to pcre_study() were also removed.
4444
4445 7. Add other new features from 5.005:
4446
4447 $(?<= positive lookbehind
4448 $(?<! negative lookbehind
4449 (?imsx-imsx) added the unsetting capability
4450 such a setting is global if at outer level; local otherwise
4451 (?imsx-imsx:) non-capturing groups with option setting
4452 (?(cond)re|re) conditional pattern matching
4453
4454 A backreference to itself in a repeated group matches the previous
4455 captured string.
4456
4457 8. General tidying up of studying (both automatic and via "study")
4458 consequential on the addition of new assertions.
4459
4460 9. As in 5.005, unlimited repeated groups that could match an empty substring
4461 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4462 runtime if any iteration does actually match an empty substring.
4463
4464 10. Include the RunTest script in the distribution.
4465
4466 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4467 discrepancies, some of which were old and were also with respect to 5.004. They
4468 have now been fixed.
4469
4470
4471 Version 1.09 28-Apr-98
4472 ----------------------
4473
4474 1. A negated single character class followed by a quantifier with a minimum
4475 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4476 program crashes, or just wrong answers. This did not apply to negated classes
4477 containing more than one character, or to minima other than one.
4478
4479
4480 Version 1.08 27-Mar-98
4481 ----------------------
4482
4483 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4484
4485 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4486 latter must appear before anything that relies on it in the pattern.
4487
4488
4489 Version 1.07 16-Feb-98
4490 ----------------------
4491
4492 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4493 repeat of a potentially empty string).
4494
4495
4496 Version 1.06 23-Jan-98
4497 ----------------------
4498
4499 1. Added Markus Oberhumer's little patches for C++.
4500
4501 2. Literal strings longer than 255 characters were broken.
4502
4503
4504 Version 1.05 23-Dec-97
4505 ----------------------
4506
4507 1. Negated character classes containing more than one character were failing if
4508 PCRE_CASELESS was set at run time.
4509
4510
4511 Version 1.04 19-Dec-97
4512 ----------------------
4513
4514 1. Corrected the man page, where some "const" qualifiers had been omitted.
4515
4516 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4517 input syntax.
4518
4519 3. Fixed memory leak which occurred when a regex with back references was
4520 matched with an offsets vector that wasn't big enough. The temporary memory
4521 that is used in this case wasn't being freed if the match failed.
4522
4523 4. Tidied pcretest to ensure it frees memory that it gets.
4524
4525 5. Temporary memory was being obtained in the case where the passed offsets
4526 vector was exactly big enough.
4527
4528 6. Corrected definition of offsetof() from change 5 below.
4529
4530 7. I had screwed up change 6 below and broken the rules for the use of
4531 setjmp(). Now fixed.
4532
4533
4534 Version 1.03 18-Dec-97
4535 ----------------------
4536
4537 1. A erroneous regex with a missing opening parenthesis was correctly
4538 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4539 on some systems.
4540
4541 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4542 it was reported that one broken compiler failed on the former because "code" is
4543 also an independent variable.
4544
4545 3. The erroneous regex a[]b caused an array overrun reference.
4546
4547 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4548 fail on data ending with that character. (It was going on too far, and checking
4549 the next character, typically a binary zero.) This was specific to the
4550 optimized code for single-character negative classes.
4551
4552 5. Added a contributed patch from the TIN world which does the following:
4553
4554 + Add an undef for memmove, in case the the system defines a macro for it.
4555
4556 + Add a definition of offsetof(), in case there isn't one. (I don't know
4557 the reason behind this - offsetof() is part of the ANSI standard - but
4558 it does no harm).
4559
4560 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4561 most of the places where whitespace preceded '#'. I have given up and
4562 allowed the remaining 2 cases to be at the margin.
4563
4564 + Rename some variables in pcre to eliminate shadowing. This seems very
4565 pedantic, but does no harm, of course.
4566
4567 6. Moved the call to setjmp() into its own function, to get rid of warnings
4568 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4569
4570 7. Constructs such as \d{8,} were compiling into the equivalent of
4571 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4572 outcome, but in this particular case used more store than had been allocated,
4573 which caused the bug to be discovered because it threw up an internal error.
4574
4575 8. The debugging code in both pcre and pcretest for outputting the compiled
4576 form of a regex was going wrong in the case of back references followed by
4577 curly-bracketed repeats.
4578
4579
4580 Version 1.02 12-Dec-97
4581 ----------------------
4582
4583 1. Typos in pcre.3 and comments in the source fixed.
4584
4585 2. Applied a contributed patch to get rid of places where it used to remove
4586 'const' from variables, and fixed some signed/unsigned and uninitialized
4587 variable warnings.
4588
4589 3. Added the "runtest" target to Makefile.
4590
4591 4. Set default compiler flag to -O2 rather than just -O.
4592
4593
4594 Version 1.01 19-Nov-97
4595 ----------------------
4596
4597 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4598 like /([ab]*)*/, that is, for classes with more than one character in them.
4599
4600 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4601 as /((?>a*))*/ (a PCRE_EXTRA facility).
4602
4603
4604 Version 1.00 18-Nov-97
4605 ----------------------
4606
4607 1. Added compile-time macros to support systems such as SunOS4 which don't have
4608 memmove() or strerror() but have other things that can be used instead.
4609
4610 2. Arranged that "make clean" removes the executables.
4611
4612
4613 Version 0.99 27-Oct-97
4614 ----------------------
4615
4616 1. Fixed bug in code for optimizing classes with only one character. It was
4617 initializing a 32-byte map regardless, which could cause it to run off the end
4618 of the memory it had got.
4619
4620 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4621
4622
4623 Version 0.98 22-Oct-97
4624 ----------------------
4625
4626 1. Fixed bug in code for handling temporary memory usage when there are more
4627 back references than supplied space in the ovector. This could cause segfaults.
4628
4629
4630 Version 0.97 21-Oct-97
4631 ----------------------
4632
4633 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4634
4635 2. Optimized negated single characters not to use a bit map.
4636
4637 3. Brought error texts together as macro definitions; clarified some of them;
4638 fixed one that was wrong - it said "range out of order" when it meant "invalid
4639 escape sequence".
4640
4641 4. Changed some char * arguments to const char *.
4642
4643 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4644
4645 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4646 pcretest.
4647
4648
4649 Version 0.96 16-Oct-97
4650 ----------------------
4651
4652 1. Added a simple "pgrep" utility to the distribution.
4653
4654 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4655 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4656 where "ddd" means "one or more decimal digits".
4657
4658 3. Fixed serious bug. If a pattern had a back reference, but the call to
4659 pcre_exec() didn't supply a large enough ovector to record the related
4660 identifying subpattern, the match always failed. PCRE now remembers the number
4661 of the largest back reference, and gets some temporary memory in which to save
4662 the offsets during matching if necessary, in order to ensure that
4663 backreferences always work.
4664
4665 4. Increased the compatibility with Perl in a number of ways:
4666
4667 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4668 to request this handling. The option can be set at compile or exec time.
4669
4670 (b) $ matches before a terminating newline by default; an option
4671 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4672 mode). The option can be set at compile or exec time.
4673
4674 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4675 the same as Perl's. If the decimal number it represents is less than 10
4676 or there aren't that many previous left capturing parentheses, an octal
4677 escape is read. Inside a character class, it's always an octal escape,
4678 even if it is a single digit.
4679
4680 (d) An escaped but undefined alphabetic character is taken as a literal,
4681 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4682 escapes.
4683
4684 (e) {0} is now permitted. (The previous item is removed from the compiled
4685 pattern).
4686
4687 5. Changed all the names of code files so that the basic parts are no longer
4688 than 10 characters, and abolished the teeny "globals.c" file.
4689
4690 6. Changed the handling of character classes; they are now done with a 32-byte
4691 bit map always.
4692
4693 7. Added the -d and /D options to pcretest to make it possible to look at the
4694 internals of compilation without having to recompile pcre.
4695
4696
4697 Version 0.95 23-Sep-97
4698 ----------------------
4699
4700 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4701 \x20 at the start of a run of normal characters. These were being treated as
4702 real characters, instead of the source characters being re-checked.
4703
4704
4705 Version 0.94 18-Sep-97
4706 ----------------------
4707
4708 1. The functions are now thread-safe, with the caveat that the global variables
4709 containing pointers to malloc() and free() or alternative functions are the
4710 same for all threads.
4711
4712 2. Get pcre_study() to generate a bitmap of initial characters for non-
4713 anchored patterns when this is possible, and use it if passed to pcre_exec().
4714
4715
4716 Version 0.93 15-Sep-97
4717 ----------------------
4718
4719 1. /(b)|(:+)/ was computing an incorrect first character.
4720
4721 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4722 but not actually doing anything yet.
4723
4724 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4725 as Perl does (e.g. [-az] or [az-]).
4726
4727 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4728 all possible positions.
4729
4730 5. Split up into different modules to avoid including unneeded functions in a
4731 compiled binary. However, compile and exec are still in one module. The "study"
4732 function is split off.
4733
4734 6. The character tables are now in a separate module whose source is generated
4735 by an auxiliary program - but can then be edited by hand if required. There are
4736 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4737 toupper() in the code.
4738
4739 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4740 make them global. Abolish the function for setting them, as the caller can now
4741 set them directly.
4742
4743
4744 Version 0.92 11-Sep-97
4745 ----------------------
4746
4747 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4748 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4749
4750 2. Caseless matching was not working in character classes if the characters in
4751 the pattern were in upper case.
4752
4753 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4754
4755 4. Make PCRE_ANCHORED public and accept as a compile option.
4756
4757 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4758 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4759 pass them.
4760
4761 6. Give an error if bad option bits passed at compile or run time.
4762
4763 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4764 pcretest to cause it to pass that flag.
4765
4766 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4767 options, and the first character, if set.
4768
4769 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4770
4771
4772 Version 0.91 10-Sep-97
4773 ----------------------
4774
4775 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4776 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4777
4778 2. PCRE was looping on encountering an indefinitely repeated back reference to
4779 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4780 Perl does - treats the match as successful.
4781
4782 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5