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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 976 - (show annotations) (download)
Sat Jun 16 17:53:17 2012 UTC (2 years, 5 months ago) by ph10
File size: 231448 byte(s)
Fix capture problem with repeated, empty-string-matching groups.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.31 02-June-2012
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 33. Variable renamings in the PCRE-JIT compiler. No functionality change.
122
123 34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
124 SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
125 was enabled.
126
127 35. Improve JIT code generation for greedy plus quantifier.
128
129 36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
130 1 to "aa" instead of to an empty string. The bug affected repeated groups
131 that could potentially match an empty string.
132
133
134 Version 8.30 04-February-2012
135 -----------------------------
136
137 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
138 name is defined in ctype.h.
139
140 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
141 only in quite long subpatterns.
142
143 3. Removed the function pcre_info(), which has been obsolete and deprecated
144 since it was replaced by pcre_fullinfo() in February 2000.
145
146 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
147 match a (*MARK), and the match failed at the start of the subject, a
148 reference to memory before the start of the subject could occur. This bug
149 was introduced by fix 17 of release 8.21.
150
151 5. A reference to an unset group with zero minimum repetition was giving
152 totally wrong answers (in non-JavaScript-compatibility mode). For example,
153 /(another)?(\1?)test/ matched against "hello world test". This bug was
154 introduced in release 8.13.
155
156 6. Add support for 16-bit character strings (a large amount of work involving
157 many changes and refactorings).
158
159 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
160 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
161 from a file.
162
163 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
164 rounding is not applied in this particular case).
165
166 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
167 if they appear, or are escaped, in patterns.
168
169 10. Get rid of a number of -Wunused-but-set-variable warnings.
170
171 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
172 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
173 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
174 also returns the mark "x". This bug applied to capturing parentheses,
175 non-capturing parentheses, and atomic parentheses. It also applied to some
176 assertions.
177
178 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
179 information out of configure.ac instead of relying on pcre.h.generic, which
180 is not stored in the repository.
181
182 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
183 -lreadline.
184
185 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
186
187 15. Applied Graycode's patch to put the top-level frame on the stack rather
188 than the heap when not using the stack for recursion. This gives a
189 performance improvement in many cases when recursion is not deep.
190
191 16. Experimental code added to "pcretest -C" to output the stack frame size.
192
193
194 Version 8.21 12-Dec-2011
195 ------------------------
196
197 1. Updating the JIT compiler.
198
199 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
200 are added as well.
201
202 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
203 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
204 calling _pcre_jit_exec. Some extra comments are added.
205
206 4. (*MARK) settings inside atomic groups that do not contain any capturing
207 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
208 was introduced by change 18 for 8.20.
209
210 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
211 ECMA-262 standard.
212
213 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
214 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
215 This bug was probably introduced by change 9 of 8.13.
216
217 7. While fixing 6 above, I noticed that a number of other items were being
218 incorrectly rejected as "not fixed length". This arose partly because newer
219 opcodes had not been added to the fixed-length checking code. I have (a)
220 corrected the bug and added tests for these items, and (b) arranged for an
221 error to occur if an unknown opcode is encountered while checking for fixed
222 length instead of just assuming "not fixed length". The items that were
223 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
224 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
225 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
226
227 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
228 being incorrectly compiled and would have given unpredicatble results.
229
230 9. A possessively repeated subpattern with minimum repeat count greater than
231 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
232 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
233 the first (A) could occur when it should not.
234
235 10. Add a cast and remove a redundant test from the code.
236
237 11. JIT should use pcre_malloc/pcre_free for allocation.
238
239 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
240 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
241 is anything other than /usr, -L is still shown).
242
243 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
244
245 14. Perl does not support \N without a following name in a [] class; PCRE now
246 also gives an error.
247
248 15. If a forward reference was repeated with an upper limit of around 2000,
249 it caused the error "internal error: overran compiling workspace". The
250 maximum number of forward references (including repeats) was limited by the
251 internal workspace, and dependent on the LINK_SIZE. The code has been
252 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
253 the default depends on LINK_SIZE. There is a new upper limit (for safety)
254 of around 200,000 forward references. While doing this, I also speeded up
255 the filling in of repeated forward references.
256
257 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
258 incorrectly expecting the subject to contain another "a" after the start.
259
260 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
261 in the match, the SKIP should be ignored. This was not happening; instead
262 the SKIP was being treated as NOMATCH. For patterns such as
263 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
264 tested.
265
266 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
267 now much more compatible with Perl, in particular in cases where the result
268 is a non-match for a non-anchored pattern. For example, if
269 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
270 "m", where previously it did not return a name. A side effect of this
271 change is that for partial matches, the last encountered mark name is
272 returned, as for non matches. A number of tests that were previously not
273 Perl-compatible have been moved into the Perl-compatible test files. The
274 refactoring has had the pleasing side effect of removing one argument from
275 the match() function, thus reducing its stack requirements.
276
277 19. If the /S+ option was used in pcretest to study a pattern using JIT,
278 subsequent uses of /S (without +) incorrectly behaved like /S+.
279
280 21. Retrieve executable code size support for the JIT compiler and fixing
281 some warnings.
282
283 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
284 not work when the shorter character appeared right at the end of the
285 subject string.
286
287 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
288 systems.
289
290 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
291 output it when the /M option is used in pcretest.
292
293 25. The CheckMan script was not being included in the distribution. Also, added
294 an explicit "perl" to run Perl scripts from the PrepareRelease script
295 because this is reportedly needed in Windows.
296
297 26. If study data was being save in a file and studying had not found a set of
298 "starts with" bytes for the pattern, the data written to the file (though
299 never used) was taken from uninitialized memory and so caused valgrind to
300 complain.
301
302 27. Updated RunTest.bat as provided by Sheri Pierce.
303
304 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
305
306 29. Computation of memory usage for the table of capturing group names was
307 giving an unnecessarily large value.
308
309
310 Version 8.20 21-Oct-2011
311 ------------------------
312
313 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
314 a POSIX class. After further experiments with Perl, which convinced me that
315 Perl has bugs and confusions, a closing square bracket is no longer allowed
316 in a POSIX name. This bug also affected patterns with classes that started
317 with full stops.
318
319 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
320 captured substring, but while checking the failing first alternative,
321 substring 1 is temporarily captured. If the output vector supplied to
322 pcre_exec() was not big enough for this capture, the yield of the function
323 was still zero ("insufficient space for captured substrings"). This cannot
324 be totally fixed without adding another stack variable, which seems a lot
325 of expense for a edge case. However, I have improved the situation in cases
326 such as /(a)(b)x|abc/ matched against "abc", where the return code
327 indicates that fewer than the maximum number of slots in the ovector have
328 been set.
329
330 3. Related to (2) above: when there are more back references in a pattern than
331 slots in the output vector, pcre_exec() uses temporary memory during
332 matching, and copies in the captures as far as possible afterwards. It was
333 using the entire output vector, but this conflicts with the specification
334 that only 2/3 is used for passing back captured substrings. Now it uses
335 only the first 2/3, for compatibility. This is, of course, another edge
336 case.
337
338 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
339 main code base, and can be used by building with --enable-jit. When this is
340 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
341 runtime --no-jit option is given.
342
343 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
344 ovector, the return from the function was zero, implying that there were
345 other matches that did not fit. The correct "exactly full" value is now
346 returned.
347
348 6. If a subpattern that was called recursively or as a subroutine contained
349 (*PRUNE) or any other control that caused it to give a non-standard return,
350 invalid errors such as "Error -26 (nested recursion at the same subject
351 position)" or even infinite loops could occur.
352
353 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
354 computing the minimum length on reaching *ACCEPT, and so ended up with the
355 wrong value of 1 rather than 0. Further investigation indicates that
356 computing a minimum subject length in the presence of *ACCEPT is difficult
357 (think back references, subroutine calls), and so I have changed the code
358 so that no minimum is registered for a pattern that contains *ACCEPT.
359
360 8. If (*THEN) was present in the first (true) branch of a conditional group,
361 it was not handled as intended. [But see 16 below.]
362
363 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
364 Sheri Pierce.
365
366 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
367 the first byte in a match must be "a".
368
369 11. Change 17 for 8.13 increased the recursion depth for patterns like
370 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
371 pattern contains any instances of (*THEN). If it does not, the old
372 optimizations are restored. It would be nice to do this on a per-group
373 basis, but at the moment that is not feasible.
374
375 12. In some environments, the output of pcretest -C is CRLF terminated. This
376 broke RunTest's code that checks for the link size. A single white space
377 character after the value is now allowed for.
378
379 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
380 For "fr", it uses the Windows-specific input and output files.
381
382 14. If (*THEN) appeared in a group that was called recursively or as a
383 subroutine, it did not work as intended. [But see next item.]
384
385 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
386 pattern fragments (but not containing any | characters). If A and B are
387 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
388 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
389 D. In other words, Perl considers parentheses that do not contain any |
390 characters to be part of a surrounding alternative, whereas PCRE was
391 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
392 differently. PCRE now behaves in the same way as Perl, except in the case
393 of subroutine/recursion calls such as (?1) which have in any case always
394 been different (but PCRE had them first :-).
395
396 16. Related to 15 above: Perl does not treat the | in a conditional group as
397 creating alternatives. Such a group is treated in the same way as an
398 ordinary group without any | characters when processing (*THEN). PCRE has
399 been changed to match Perl's behaviour.
400
401 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
402 RunGrepTest script failed.
403
404 18. Change 22 for version 13 caused atomic groups to use more stack. This is
405 inevitable for groups that contain captures, but it can lead to a lot of
406 stack use in large patterns. The old behaviour has been restored for atomic
407 groups that do not contain any capturing parentheses.
408
409 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
410 suppress the check for a minimum subject length at run time. (If it was
411 given to pcre_exec() or pcre_dfa_exec() it did work.)
412
413 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
414 fail to work when decoding hex characters in data strings in EBCDIC
415 environments.
416
417 21. It appears that in at least one Mac OS environment, the isxdigit() function
418 is implemented as a macro that evaluates to its argument more than once,
419 contravening the C 90 Standard (I haven't checked a later standard). There
420 was an instance in pcretest which caused it to go wrong when processing
421 \x{...} escapes in subject strings. The has been rewritten to avoid using
422 things like p++ in the argument of isxdigit().
423
424
425 Version 8.13 16-Aug-2011
426 ------------------------
427
428 1. The Unicode data tables have been updated to Unicode 6.0.0.
429
430 2. Two minor typos in pcre_internal.h have been fixed.
431
432 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
433 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
434 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
435
436 4. There were a number of related bugs in the code for matching backrefences
437 caselessly in UTF-8 mode when codes for the characters concerned were
438 different numbers of bytes. For example, U+023A and U+2C65 are an upper
439 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
440 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
441 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
442 2-byte code at the end of the subject (it thought there wasn't enough data
443 left).
444
445 5. Comprehensive information about what went wrong is now returned by
446 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
447 as the output vector has at least 2 elements. The offset of the start of
448 the failing character and a reason code are placed in the vector.
449
450 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
451 now returned is for the first byte of the failing character, instead of the
452 last byte inspected. This is an incompatible change, but I hope it is small
453 enough not to be a problem. It makes the returned offset consistent with
454 pcre_exec() and pcre_dfa_exec().
455
456 7. pcretest now gives a text phrase as well as the error number when
457 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
458 failure, the offset and reason code are output.
459
460 8. When \R was used with a maximizing quantifier it failed to skip backwards
461 over a \r\n pair if the subsequent match failed. Instead, it just skipped
462 back over a single character (\n). This seems wrong (because it treated the
463 two characters as a single entity when going forwards), conflicts with the
464 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
465 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
466 has been changed.
467
468 9. Some internal refactoring has changed the processing so that the handling
469 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
470 time (the PCRE_DOTALL option was changed this way some time ago: version
471 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
472 which was always a bit of a fudge. It also means that there is one less
473 argument for the match() function, which reduces its stack requirements
474 slightly. This change also fixes an incompatibility with Perl: the pattern
475 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
476
477 10. More internal refactoring has drastically reduced the number of recursive
478 calls to match() for possessively repeated groups such as (abc)++ when
479 using pcre_exec().
480
481 11. While implementing 10, a number of bugs in the handling of groups were
482 discovered and fixed:
483
484 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
485 (a|)*(?1) gave a compile-time internal error.
486 ((a|)+)+ did not notice that the outer group could match an empty string.
487 (^a|^)+ was not marked as anchored.
488 (.*a|.*)+ was not marked as matching at start or after a newline.
489
490 12. Yet more internal refactoring has removed another argument from the match()
491 function. Special calls to this function are now indicated by setting a
492 value in a variable in the "match data" data block.
493
494 13. Be more explicit in pcre_study() instead of relying on "default" for
495 opcodes that mean there is no starting character; this means that when new
496 ones are added and accidentally left out of pcre_study(), testing should
497 pick them up.
498
499 14. The -s option of pcretest has been documented for ages as being an old
500 synonym of -m (show memory usage). I have changed it to mean "force study
501 for every regex", that is, assume /S for every regex. This is similar to -i
502 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
503 using it. It makes it easier to run collections of tests with and without
504 study enabled, and thereby test pcre_study() more easily. All the standard
505 tests are now run with and without -s (but some patterns can be marked as
506 "never study" - see 20 below).
507
508 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
509 restoration of the capturing data to the outer values was not happening
510 correctly.
511
512 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
513 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
514 pattern had matched an empty string, and so incorrectly returned a no
515 match.
516
517 17. There was optimizing code for the last branch of non-capturing parentheses,
518 and also for the obeyed branch of a conditional subexpression, which used
519 tail recursion to cut down on stack usage. Unfortunately, now that there is
520 the possibility of (*THEN) occurring in these branches, tail recursion is
521 no longer possible because the return has to be checked for (*THEN). These
522 two optimizations have therefore been removed. [But see 8.20/11 above.]
523
524 18. If a pattern containing \R was studied, it was assumed that \R always
525 matched two bytes, thus causing the minimum subject length to be
526 incorrectly computed because \R can also match just one byte.
527
528 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
529 was incorrectly computed.
530
531 20. If /S is present twice on a test pattern in pcretest input, it now
532 *disables* studying, thereby overriding the use of -s on the command line
533 (see 14 above). This is necessary for one or two tests to keep the output
534 identical in both cases.
535
536 21. When (*ACCEPT) was used in an assertion that matched an empty string and
537 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
538
539 22. When an atomic group that contained a capturing parenthesis was
540 successfully matched, but the branch in which it appeared failed, the
541 capturing was not being forgotten if a higher numbered group was later
542 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
543 group 1 to "a", when in fact it should be unset. This applied to multi-
544 branched capturing and non-capturing groups, repeated or not, and also to
545 positive assertions (capturing in negative assertions does not happen
546 in PCRE) and also to nested atomic groups.
547
548 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
549 subject after a captured substring, to make it easier to tell which of a
550 number of identical substrings has been captured.
551
552 24. The way atomic groups are processed by pcre_exec() has been changed so that
553 if they are repeated, backtracking one repetition now resets captured
554 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
555 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
556 "aaa". Previously, it would have been "a". As part of this code
557 refactoring, the way recursive calls are handled has also been changed.
558
559 25. If an assertion condition captured any substrings, they were not passed
560 back unless some other capturing happened later. For example, if
561 (?(?=(a))a) was matched against "a", no capturing was returned.
562
563 26. When studying a pattern that contained subroutine calls or assertions,
564 the code for finding the minimum length of a possible match was handling
565 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
566 group 1 called group 2 while simultaneously a separate group 2 called group
567 1). A stack overflow occurred in this case. I have fixed this by limiting
568 the recursion depth to 10.
569
570 27. Updated RunTest.bat in the distribution to the version supplied by Tom
571 Fortmann. This supports explicit test numbers on the command line, and has
572 argument validation and error reporting.
573
574 28. An instance of \X with an unlimited repeat could fail if at any point the
575 first character it looked at was a mark character.
576
577 29. Some minor code refactoring concerning Unicode properties and scripts
578 should reduce the stack requirement of match() slightly.
579
580 30. Added the '=' option to pcretest to check the setting of unused capturing
581 slots at the end of the pattern, which are documented as being -1, but are
582 not included in the return count.
583
584 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
585 compiled something random. Now it gives a compile-time error (as does
586 Perl).
587
588 32. A *MARK encountered during the processing of a positive assertion is now
589 recorded and passed back (compatible with Perl).
590
591 33. If --only-matching or --colour was set on a pcregrep call whose pattern
592 had alternative anchored branches, the search for a second match in a line
593 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
594 matched the line "0102" twice. The same bug affected patterns that started
595 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
596 twice.
597
598 34. Previously, PCRE did not allow quantification of assertions. However, Perl
599 does, and because of capturing effects, quantifying parenthesized
600 assertions may at times be useful. Quantifiers are now allowed for
601 parenthesized assertions.
602
603 35. A minor code tidy in pcre_compile() when checking options for \R usage.
604
605 36. \g was being checked for fancy things in a character class, when it should
606 just be a literal "g".
607
608 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
609 appearance of a nested POSIX class supersedes an apparent external class.
610 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
611 unescaped square brackets may also appear as part of class names. For
612 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
613 more like Perl. (But see 8.20/1 above.)
614
615 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
616 was because it thought it was \N{name}, which is not supported).
617
618 39. Add minix to OS list not supporting the -S option in pcretest.
619
620 40. PCRE tries to detect cases of infinite recursion at compile time, but it
621 cannot analyze patterns in sufficient detail to catch mutual recursions
622 such as ((?1))((?2)). There is now a runtime test that gives an error if a
623 subgroup is called recursively as a subpattern for a second time at the
624 same position in the subject string. In previous releases this might have
625 been caught by the recursion limit, or it might have run out of stack.
626
627 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
628 happen only once. PCRE was, however incorrectly giving a compile time error
629 "recursive call could loop indefinitely" because it cannot analyze the
630 pattern in sufficient detail. The compile time test no longer happens when
631 PCRE is compiling a conditional subpattern, but actual runaway loops are
632 now caught at runtime (see 40 above).
633
634 42. It seems that Perl allows any characters other than a closing parenthesis
635 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
636 has been changed to be the same.
637
638 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
639 as not to get warnings when autogen.sh is called. Also changed
640 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
641
642 44. To help people who use pcregrep to scan files containing exceedingly long
643 lines, the following changes have been made:
644
645 (a) The default value of the buffer size parameter has been increased from
646 8K to 20K. (The actual buffer used is three times this size.)
647
648 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
649 PCRE is built.
650
651 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
652 to be set at run time.
653
654 (d) Numerical values in pcregrep options can be followed by K or M, for
655 example --buffer-size=50K.
656
657 (e) If a line being scanned overflows pcregrep's buffer, an error is now
658 given and the return code is set to 2.
659
660 45. Add a pointer to the latest mark to the callout data block.
661
662 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
663 partial match of an empty string instead of no match. This was specific to
664 the use of ".".
665
666 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
667 complete match instead of a partial match. This bug was dependent on both
668 the PCRE_UTF8 and PCRE_DOTALL options being set.
669
670 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
671 starting byte set, because \b was not being ignored.
672
673
674 Version 8.12 15-Jan-2011
675 ------------------------
676
677 1. Fixed some typos in the markup of the man pages, and wrote a script that
678 checks for such things as part of the documentation building process.
679
680 2. On a big-endian 64-bit system, pcregrep did not correctly process the
681 --match-limit and --recursion-limit options (added for 8.11). In
682 particular, this made one of the standard tests fail. (The integer value
683 went into the wrong half of a long int.)
684
685 3. If the --colour option was given to pcregrep with -v (invert match), it
686 did strange things, either producing crazy output, or crashing. It should,
687 of course, ignore a request for colour when reporting lines that do not
688 match.
689
690 4. Another pcregrep bug caused similar problems if --colour was specified with
691 -M (multiline) and the pattern match finished with a line ending.
692
693 5. In pcregrep, when a pattern that ended with a literal newline sequence was
694 matched in multiline mode, the following line was shown as part of the
695 match. This seems wrong, so I have changed it.
696
697 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
698 the check for further matches in the same line (so they could be coloured)
699 to overrun the end of the current line. If another match was found, it was
700 incorrectly shown (and then shown again when found in the next line).
701
702 7. If pcregrep was compiled under Windows, there was a reference to the
703 function pcregrep_exit() before it was defined. I am assuming this was
704 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
705 reported by a user. I've moved the definition above the reference.
706
707
708 Version 8.11 10-Dec-2010
709 ------------------------
710
711 1. (*THEN) was not working properly if there were untried alternatives prior
712 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
713 backtracked to try for "b" instead of moving to the next alternative branch
714 at the same level (in this case, to look for "c"). The Perl documentation
715 is clear that when (*THEN) is backtracked onto, it goes to the "next
716 alternative in the innermost enclosing group".
717
718 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
719 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
720 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
721 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
722 (*THEN).
723
724 3. If \s appeared in a character class, it removed the VT character from
725 the class, even if it had been included by some previous item, for example
726 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
727 of \s, but is part of the POSIX "space" class.)
728
729 4. A partial match never returns an empty string (because you can always
730 match an empty string at the end of the subject); however the checking for
731 an empty string was starting at the "start of match" point. This has been
732 changed to the "earliest inspected character" point, because the returned
733 data for a partial match starts at this character. This means that, for
734 example, /(?<=abc)def/ gives a partial match for the subject "abc"
735 (previously it gave "no match").
736
737 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
738 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
739 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
740 has an implication that the given string is incomplete (because a partial
741 match is preferred over a full match). For this reason, these items now
742 give a partial match in this situation. [Aside: previously, the one case
743 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
744 match rather than a full match, which was wrong by the old rules, but is
745 now correct.]
746
747 6. There was a bug in the handling of #-introduced comments, recognized when
748 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
749 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
750 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
751 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
752 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
753 places in pcre_compile().
754
755 7. Related to (6) above, when pcre_compile() was skipping #-introduced
756 comments when looking ahead for named forward references to subpatterns,
757 the only newline sequence it recognized was NL. It now handles newlines
758 according to the set newline convention.
759
760 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
761 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
762 cater for a lack of strerror(). These oversights have been fixed.
763
764 9. Added --match-limit and --recursion-limit to pcregrep.
765
766 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
767
768 11. When the -o option was used, pcregrep was setting a return code of 1, even
769 when matches were found, and --line-buffered was not being honoured.
770
771 12. Added an optional parentheses number to the -o and --only-matching options
772 of pcregrep.
773
774 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
775 can match an empty string. The code to do it in pcretest and pcredemo
776 needed fixing:
777
778 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
779 only one byte after an empty string match just before CRLF (this case
780 just got forgotten; "any" and "anycrlf" were OK).
781
782 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
783 mode when an empty string match preceded an ASCII character followed by
784 a non-ASCII character. (The code for advancing by one character rather
785 than one byte was nonsense.)
786
787 (c) The pcredemo.c sample program did not have any code at all to handle
788 the cases when CRLF is a valid newline sequence.
789
790 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
791 as a starting offset was within the subject string. There is now a new
792 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
793 negative or greater than the length of the string. In order to test this,
794 pcretest is extended to allow the setting of negative starting offsets.
795
796 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
797 starting offset points to the beginning of a UTF-8 character was
798 unnecessarily clumsy. I tidied it up.
799
800 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
801 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
802
803 17. Nobody had reported that the --include_dir option, which was added in
804 release 7.7 should have been called --include-dir (hyphen, not underscore)
805 for compatibility with GNU grep. I have changed it to --include-dir, but
806 left --include_dir as an undocumented synonym, and the same for
807 --exclude-dir, though that is not available in GNU grep, at least as of
808 release 2.5.4.
809
810 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
811 characters from a string of bytes) have been redefined so as not to use
812 loops, in order to improve performance in some environments. At the same
813 time, I abstracted some of the common code into auxiliary macros to save
814 repetition (this should not affect the compiled code).
815
816 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
817 compile-time error is now given if \c is not followed by an ASCII
818 character, that is, a byte less than 128. (In EBCDIC mode, the code is
819 different, and any byte value is allowed.)
820
821 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
822 START_OPTIMIZE option, which is now allowed at compile time - but just
823 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
824 to pcregrep and other applications that have no direct access to PCRE
825 options. The new /Y option in pcretest sets this option when calling
826 pcre_compile().
827
828 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
829 back references. Groups containing recursive back references were forced to
830 be atomic by that change, but in the case of named groups, the amount of
831 memory required was incorrectly computed, leading to "Failed: internal
832 error: code overflow". This has been fixed.
833
834 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
835 pcretest.c, to avoid build problems in some Borland environments.
836
837
838 Version 8.10 25-Jun-2010
839 ------------------------
840
841 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
842 THEN.
843
844 2. (*ACCEPT) was not working when inside an atomic group.
845
846 3. Inside a character class, \B is treated as a literal by default, but
847 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
848 causes the error). The code is unchanged, but I tidied the documentation.
849
850 4. Inside a character class, PCRE always treated \R and \X as literals,
851 whereas Perl faults them if its -w option is set. I have changed PCRE so
852 that it faults them when PCRE_EXTRA is set.
853
854 5. Added support for \N, which always matches any character other than
855 newline. (It is the same as "." when PCRE_DOTALL is not set.)
856
857 6. When compiling pcregrep with newer versions of gcc which may have
858 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
859 declared with attribute warn_unused_result" were given. Just casting the
860 result to (void) does not stop the warnings; a more elaborate fudge is
861 needed. I've used a macro to implement this.
862
863 7. Minor change to pcretest.c to avoid a compiler warning.
864
865 8. Added four artifical Unicode properties to help with an option to make
866 \s etc use properties (see next item). The new properties are: Xan
867 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
868
869 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
870 use Unicode properties. (*UCP) at the start of a pattern can be used to set
871 this option. Modified pcretest to add /W to test this facility. Added
872 REG_UCP to make it available via the POSIX interface.
873
874 10. Added --line-buffered to pcregrep.
875
876 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
877 studied, and the match started with a letter with a code point greater than
878 127 whose first byte was different to the first byte of the other case of
879 the letter, the other case of this starting letter was not recognized
880 (#976).
881
882 12. If a pattern that was studied started with a repeated Unicode property
883 test, for example, \p{Nd}+, there was the theoretical possibility of
884 setting up an incorrect bitmap of starting bytes, but fortunately it could
885 not have actually happened in practice until change 8 above was made (it
886 added property types that matched character-matching opcodes).
887
888 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
889 possible starting bytes for non-anchored patterns.
890
891 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
892 \R, and also a number of cases that involve Unicode properties, both
893 explicit and implicit when PCRE_UCP is set.
894
895 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
896 input, it could crash or give wrong results if characters with values
897 greater than 0xc0 were present in the subject string. (Detail: it assumed
898 UTF-8 input when processing these items.)
899
900 16. Added a lot of (int) casts to avoid compiler warnings in systems where
901 size_t is 64-bit (#991).
902
903 17. Added a check for running out of memory when PCRE is compiled with
904 --disable-stack-for-recursion (#990).
905
906 18. If the last data line in a file for pcretest does not have a newline on
907 the end, a newline was missing in the output.
908
909 19. The default pcre_chartables.c file recognizes only ASCII characters (values
910 less than 128) in its various bitmaps. However, there is a facility for
911 generating tables according to the current locale when PCRE is compiled. It
912 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
913 characters, are recognized by isspace() and therefore were getting set in
914 these tables, and indeed these tables seem to approximate to ISO 8859. This
915 caused a problem in UTF-8 mode when pcre_study() was used to create a list
916 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
917 which of course cannot start UTF-8 characters. I have changed the code so
918 that only real ASCII characters (less than 128) and the correct starting
919 bytes for UTF-8 encodings are set for characters greater than 127 when in
920 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
921 altogether.)
922
923 20. Added the /T option to pcretest so as to be able to run tests with non-
924 standard character tables, thus making it possible to include the tests
925 used for 19 above in the standard set of tests.
926
927 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
928 reference to a subpattern the other side of a comment that contains an
929 opening parenthesis caused either an internal compiling error, or a
930 reference to the wrong subpattern.
931
932
933 Version 8.02 19-Mar-2010
934 ------------------------
935
936 1. The Unicode data tables have been updated to Unicode 5.2.0.
937
938 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
939 configured.
940
941 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
942 original author of that file, following a query about its status.
943
944 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
945 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
946
947 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
948 quantifier applied to a forward-referencing subroutine call, could compile
949 incorrect code or give the error "internal error: previously-checked
950 referenced subpattern not found".
951
952 6. Both MS Visual Studio and Symbian OS have problems with initializing
953 variables to point to external functions. For these systems, therefore,
954 pcre_malloc etc. are now initialized to local functions that call the
955 relevant global functions.
956
957 7. There were two entries missing in the vectors called coptable and poptable
958 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
959 I've fixed the data, and added a kludgy way of testing at compile time that
960 the lengths are correct (equal to the number of opcodes).
961
962 8. Following on from 7, I added a similar kludge to check the length of the
963 eint vector in pcreposix.c.
964
965 9. Error texts for pcre_compile() are held as one long string to avoid too
966 much relocation at load time. To find a text, the string is searched,
967 counting zeros. There was no check for running off the end of the string,
968 which could happen if a new error number was added without updating the
969 string.
970
971 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
972
973 11. \K was not working if it appeared in an atomic group or in a group that
974 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
975 \K is "not well defined" if used in an assertion. PCRE now accepts it if
976 the assertion is positive, but not if it is negative.
977
978 12. Change 11 fortuitously reduced the size of the stack frame used in the
979 "match()" function of pcre_exec.c by one pointer. Forthcoming
980 implementation of support for (*MARK) will need an extra pointer on the
981 stack; I have reserved it now, so that the stack frame size does not
982 decrease.
983
984 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
985 item in branch that calls a recursion is a subroutine call - as in the
986 second branch in the above example - was incorrectly given the compile-
987 time error "recursive call could loop indefinitely" because pcre_compile()
988 was not correctly checking the subroutine for matching a non-empty string.
989
990 14. The checks for overrunning compiling workspace could trigger after an
991 overrun had occurred. This is a "should never occur" error, but it can be
992 triggered by pathological patterns such as hundreds of nested parentheses.
993 The checks now trigger 100 bytes before the end of the workspace.
994
995 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
996
997
998 Version 8.01 19-Jan-2010
999 ------------------------
1000
1001 1. If a pattern contained a conditional subpattern with only one branch (in
1002 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1003 computed the wrong minimum data length (which is of course zero for such
1004 subpatterns). This could cause incorrect "no match" results.
1005
1006 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1007 the pattern is reset in the first branch, pcre_compile() failed with
1008 "internal error: code overflow at offset...". This happened only when
1009 the reset was to the original external option setting. (An optimization
1010 abstracts leading options settings into an external setting, which was the
1011 cause of this.)
1012
1013 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1014 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1015 assertion pattern did not match (meaning that the assertion was true), it
1016 was incorrectly treated as false if the SKIP had been reached during the
1017 matching. This also applied to assertions used as conditions.
1018
1019 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
1020 assertion subpattern, including such a pattern used as a condition,
1021 unpredictable results occurred, instead of the error return
1022 PCRE_ERROR_DFA_UITEM.
1023
1024 5. The C++ GlobalReplace function was not working like Perl for the special
1025 situation when an empty string is matched. It now does the fancy magic
1026 stuff that is necessary.
1027
1028 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1029 removed. (These were left over from very, very early versions of PCRE.)
1030
1031 7. Some cosmetic changes to the code to make life easier when compiling it
1032 as part of something else:
1033
1034 (a) Change DEBUG to PCRE_DEBUG.
1035
1036 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1037 called "current" as "current_branch", to prevent a collision with the
1038 Linux macro when compiled as a kernel module.
1039
1040 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1041 prevent a collision with the Linux macro when compiled as a kernel
1042 module.
1043
1044 8. In pcre_compile() there are some checks for integer overflows that used to
1045 cast potentially large values to (double). This has been changed to that
1046 when building, a check for int64_t is made, and if it is found, it is used
1047 instead, thus avoiding the use of floating point arithmetic. (There is no
1048 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1049 double.
1050
1051 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1052 2005 (difference between two addresses compared to an unsigned value).
1053
1054 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1055 custom one, because of the following reported problem in Windows:
1056
1057 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1058 under Win32.
1059 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1060 therefore missing the function definition.
1061 - The compiler thus generates a "C" signature for the test function.
1062 - The linker fails to find the "C" function.
1063 - PCRE fails to configure if asked to do so against libbz2.
1064
1065 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1066 messages were output:
1067
1068 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1069 rerunning libtoolize, to keep the correct libtool macros in-tree.
1070 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1071
1072 I have done both of these things.
1073
1074 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1075 most of the time, it *can* run out if it is given a pattern that contains a
1076 runaway infinite recursion. I updated the discussion in the pcrestack man
1077 page.
1078
1079 13. Now that we have gone to the x.xx style of version numbers, the minor
1080 version may start with zero. Using 08 or 09 is a bad idea because users
1081 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1082 interpreted as invalid octal numbers. I've updated the previous comment in
1083 configure.ac, and also added a check that gives an error if 08 or 09 are
1084 used.
1085
1086 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1087 causing partial matching to fail when the end of the subject matched \W
1088 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1089
1090 15. There were some discrepancies between the declarations in pcre_internal.h
1091 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1092 their definitions. The declarations used "const uschar *" and the
1093 definitions used USPTR. Even though USPTR is normally defined as "const
1094 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1095 reported that: "This difference in casting confuses some C++ compilers, for
1096 example, SunCC recognizes above declarations as different functions and
1097 generates broken code for hbpcre." I have changed the declarations to use
1098 USPTR.
1099
1100 16. GNU libtool is named differently on some systems. The autogen.sh script now
1101 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1102 (FreeBSD).
1103
1104 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1105 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1106 comment: "Figure out how to create a longlong from a string: strtoll and
1107 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1108 instance, but it only takes 2 args instead of 3!"
1109
1110 18. A subtle bug concerned with back references has been fixed by a change of
1111 specification, with a corresponding code fix. A pattern such as
1112 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1113 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1114 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1115 same bug. Such groups have to be quantified to be useful, or contained
1116 inside another quantified group. (If there's no repetition, the reference
1117 can never match.) The problem arises because, having left the group and
1118 moved on to the rest of the pattern, a later failure that backtracks into
1119 the group uses the captured value from the final iteration of the group
1120 rather than the correct earlier one. I have fixed this in PCRE by forcing
1121 any group that contains a reference to itself to be an atomic group; that
1122 is, there cannot be any backtracking into it once it has completed. This is
1123 similar to recursive and subroutine calls.
1124
1125
1126 Version 8.00 19-Oct-09
1127 ----------------------
1128
1129 1. The table for translating pcre_compile() error codes into POSIX error codes
1130 was out-of-date, and there was no check on the pcre_compile() error code
1131 being within the table. This could lead to an OK return being given in
1132 error.
1133
1134 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1135 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1136 in a Windows environment.
1137
1138 3. The pcregrep --count option prints the count for each file even when it is
1139 zero, as does GNU grep. However, pcregrep was also printing all files when
1140 --files-with-matches was added. Now, when both options are given, it prints
1141 counts only for those files that have at least one match. (GNU grep just
1142 prints the file name in this circumstance, but including the count seems
1143 more useful - otherwise, why use --count?) Also ensured that the
1144 combination -clh just lists non-zero counts, with no names.
1145
1146 4. The long form of the pcregrep -F option was incorrectly implemented as
1147 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1148 but it seems right to fix it, and I didn't think it was worth preserving
1149 the old behaviour.
1150
1151 5. The command line items --regex=pattern and --regexp=pattern were not
1152 recognized by pcregrep, which required --regex pattern or --regexp pattern
1153 (with a space rather than an '='). The man page documented the '=' forms,
1154 which are compatible with GNU grep; these now work.
1155
1156 6. No libpcreposix.pc file was created for pkg-config; there was just
1157 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1158
1159 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1160 when UCP support is not needed, by modifying the Python script that
1161 generates it from Unicode data files. This should not matter if the module
1162 is correctly used as a library, but I received one complaint about 50K of
1163 unwanted data. My guess is that the person linked everything into his
1164 program rather than using a library. Anyway, it does no harm.
1165
1166 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1167 was a minimum greater than 1 for a wide character in a possessive
1168 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1169 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1170 character. Chaos in the form of incorrect output or a compiling loop could
1171 result.
1172
1173 9. The restrictions on what a pattern can contain when partial matching is
1174 requested for pcre_exec() have been removed. All patterns can now be
1175 partially matched by this function. In addition, if there are at least two
1176 slots in the offset vector, the offset of the earliest inspected character
1177 for the match and the offset of the end of the subject are set in them when
1178 PCRE_ERROR_PARTIAL is returned.
1179
1180 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1181 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1182 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1183 and may be more useful for multi-segment matching.
1184
1185 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1186 used to be given if ever the end of the subject was reached; now it is
1187 given only if matching could not proceed because another character was
1188 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1189 string "Z", which now yields "no match" instead of "partial match". In the
1190 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1191 final character ended with (*FAIL).
1192
1193 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1194 if the pattern had a "must contain" character that was already found in the
1195 earlier partial match, unless partial matching was again requested. For
1196 example, with the pattern /dog.(body)?/, the "must contain" character is
1197 "g". If the first part-match was for the string "dog", restarting with
1198 "sbody" failed. This bug has been fixed.
1199
1200 13. The string returned by pcre_dfa_exec() after a partial match has been
1201 changed so that it starts at the first inspected character rather than the
1202 first character of the match. This makes a difference only if the pattern
1203 starts with a lookbehind assertion or \b or \B (\K is not supported by
1204 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1205 matching functions compatible, and I think it's the right thing to do.
1206
1207 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1208 so that the demonstration program is easily available in environments where
1209 PCRE has not been installed from source.
1210
1211 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1212 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1213 library.
1214
1215 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1216 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1217 is not the first non-POSIX option to be added. Clearly some people find
1218 these options useful.
1219
1220 17. If a caller to the POSIX matching function regexec() passes a non-zero
1221 value for nmatch with a NULL value for pmatch, the value of
1222 nmatch is forced to zero.
1223
1224 18. RunGrepTest did not have a test for the availability of the -u option of
1225 the diff command, as RunTest does. It now checks in the same way as
1226 RunTest, and also checks for the -b option.
1227
1228 19. If an odd number of negated classes containing just a single character
1229 interposed, within parentheses, between a forward reference to a named
1230 subpattern and the definition of the subpattern, compilation crashed with
1231 an internal error, complaining that it could not find the referenced
1232 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1233 [The bug was that it was starting one character too far in when skipping
1234 over the character class, thus treating the ] as data rather than
1235 terminating the class. This meant it could skip too much.]
1236
1237 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1238 /g option in pcretest when the pattern contains \K, which makes it possible
1239 to have an empty string match not at the start, even when the pattern is
1240 anchored. Updated pcretest and pcredemo to use this option.
1241
1242 21. If the maximum number of capturing subpatterns in a recursion was greater
1243 than the maximum at the outer level, the higher number was returned, but
1244 with unset values at the outer level. The correct (outer level) value is
1245 now given.
1246
1247 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1248 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1249 make it do so. The string so far is captured, making this feature
1250 compatible with Perl.
1251
1252 23. The tests have been re-organized, adding tests 11 and 12, to make it
1253 possible to check the Perl 5.10 features against Perl 5.10.
1254
1255 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1256 pattern matches a fixed length string. PCRE did not allow this; now it
1257 does. Neither allows recursion.
1258
1259 25. I finally figured out how to implement a request to provide the minimum
1260 length of subject string that was needed in order to match a given pattern.
1261 (It was back references and recursion that I had previously got hung up
1262 on.) This code has now been added to pcre_study(); it finds a lower bound
1263 to the length of subject needed. It is not necessarily the greatest lower
1264 bound, but using it to avoid searching strings that are too short does give
1265 some useful speed-ups. The value is available to calling programs via
1266 pcre_fullinfo().
1267
1268 26. While implementing 25, I discovered to my embarrassment that pcretest had
1269 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1270 study optimizations had never been tested with that matching function.
1271 Oops. What is worse, even when it was passed study data, there was a bug in
1272 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1273 were also very few tests of studied patterns with pcre_dfa_exec().
1274
1275 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1276 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1277 on the other side of the coin, they are no longer allowed to have different
1278 names, because these cannot be distinguished in PCRE, and this has caused
1279 confusion. (This is a difference from Perl.)
1280
1281 28. When duplicate subpattern names are present (necessarily with different
1282 numbers, as required by 27 above), and a test is made by name in a
1283 conditional pattern, either for a subpattern having been matched, or for
1284 recursion in such a pattern, all the associated numbered subpatterns are
1285 tested, and the overall condition is true if the condition is true for any
1286 one of them. This is the way Perl works, and is also more like the way
1287 testing by number works.
1288
1289
1290 Version 7.9 11-Apr-09
1291 ---------------------
1292
1293 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1294 (pcretest), all targets were linked against these libraries. This included
1295 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1296 libraries. This caused unwanted dependencies to be created. This problem
1297 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1298 pcretest is linked with readline.
1299
1300 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1301 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1302 moved outside it again, because FALSE and TRUE are already defined in AIX,
1303 but BOOL is not.
1304
1305 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1306 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1307
1308 4. The pcregrep documentation said spaces were inserted as well as colons (or
1309 hyphens) following file names and line numbers when outputting matching
1310 lines. This is not true; no spaces are inserted. I have also clarified the
1311 wording for the --colour (or --color) option.
1312
1313 5. In pcregrep, when --colour was used with -o, the list of matching strings
1314 was not coloured; this is different to GNU grep, so I have changed it to be
1315 the same.
1316
1317 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1318 each matching line was coloured. Now it goes on to look for further matches
1319 of any of the test patterns, which is the same behaviour as GNU grep.
1320
1321 7. A pattern that could match an empty string could cause pcregrep to loop; it
1322 doesn't make sense to accept an empty string match in pcregrep, so I have
1323 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1324 seems to be how GNU grep behaves.
1325
1326 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1327 start or after a newline", because the conditional assertion was not being
1328 correctly handled. The rule now is that both the assertion and what follows
1329 in the first alternative must satisfy the test.
1330
1331 9. If auto-callout was enabled in a pattern with a conditional group whose
1332 condition was an assertion, PCRE could crash during matching, both with
1333 pcre_exec() and pcre_dfa_exec().
1334
1335 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1336 used for matching.
1337
1338 11. Unicode property support in character classes was not working for
1339 characters (bytes) greater than 127 when not in UTF-8 mode.
1340
1341 12. Added the -M command line option to pcretest.
1342
1343 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1344
1345 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1346
1347 16. Added comments and documentation about mis-use of no_arg in the C++
1348 wrapper.
1349
1350 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1351 from Martin Jerabek that uses macro names for all relevant character and
1352 string constants.
1353
1354 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1355 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1356 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1357 these, but not everybody uses configure.
1358
1359 19. A conditional group that had only one branch was not being correctly
1360 recognized as an item that could match an empty string. This meant that an
1361 enclosing group might also not be so recognized, causing infinite looping
1362 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1363 with the subject "ab", where knowledge that the repeated group can match
1364 nothing is needed in order to break the loop.
1365
1366 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1367 exec(), but without supplying a callout function, matching went wrong.
1368
1369 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1370 leak if the size of the offset vector was greater than 30. When the vector
1371 is smaller, the saved offsets during recursion go onto a local stack
1372 vector, but for larger vectors malloc() is used. It was failing to free
1373 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1374 error, in fact).
1375
1376 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1377 heapframe that is used only when UTF-8 support is enabled. This caused no
1378 problem, but was untidy.
1379
1380 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1381 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1382 included within another project.
1383
1384 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1385 slightly modified by me:
1386
1387 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1388 not building pcregrep.
1389
1390 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1391 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1392
1393 25. Forward references, both numeric and by name, in patterns that made use of
1394 duplicate group numbers, could behave incorrectly or give incorrect errors,
1395 because when scanning forward to find the reference group, PCRE was not
1396 taking into account the duplicate group numbers. A pattern such as
1397 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1398
1399 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1400 the feature of a custom pointer more persuasive (as requested by a user).
1401
1402 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1403 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1404 pre-defined.
1405
1406 28. Added support for (*UTF8) at the start of a pattern.
1407
1408 29. Arrange for flags added by the "release type" setting in CMake to be shown
1409 in the configuration summary.
1410
1411
1412 Version 7.8 05-Sep-08
1413 ---------------------
1414
1415 1. Replaced UCP searching code with optimized version as implemented for Ad
1416 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1417 stage table and inline lookup instead of a function, giving speed ups of 2
1418 to 5 times on some simple patterns that I tested. Permission was given to
1419 distribute the MultiStage2.py script that generates the tables (it's not in
1420 the tarball, but is in the Subversion repository).
1421
1422 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1423 scripts.
1424
1425 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1426 a group with a zero qualifier. The result of the study could be incorrect,
1427 or the function might crash, depending on the pattern.
1428
1429 4. Caseless matching was not working for non-ASCII characters in back
1430 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1431 It now works when Unicode Property Support is available.
1432
1433 5. In pcretest, an escape such as \x{de} in the data was always generating
1434 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1435 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1436 truncation.
1437
1438 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1439
1440 7. Added two (int) casts to pcregrep when printing the difference of two
1441 pointers, in case they are 64-bit values.
1442
1443 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1444 test 2 if it fails.
1445
1446 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1447 and a #define of that name to empty if it is not externally set. This is to
1448 allow users of MSVC to set it if necessary.
1449
1450 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1451 the convenience functions in the pcre_get.c source file.
1452
1453 11. An option change at the start of a pattern that had top-level alternatives
1454 could cause overwriting and/or a crash. This command provoked a crash in
1455 some environments:
1456
1457 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1458
1459 This potential security problem was recorded as CVE-2008-2371.
1460
1461 12. For a pattern where the match had to start at the beginning or immediately
1462 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1463 pcre_dfa_exec() could read past the end of the passed subject if there was
1464 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1465 pcretest so that it places the subject at the end of its malloc-ed buffer.
1466
1467 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1468 exec() might read past the end of the data buffer in UTF-8 mode.
1469
1470 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1471 the data contained the byte 0x85 as part of a UTF-8 character within its
1472 first line. This applied both to normal and DFA matching.
1473
1474 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1475 /^[^d]*?$/8 failed to match "abc".
1476
1477 16. Added a missing copyright notice to pcrecpp_internal.h.
1478
1479 17. Make it more clear in the documentation that values returned from
1480 pcre_exec() in ovector are byte offsets, not character counts.
1481
1482 18. Tidied a few places to stop certain compilers from issuing warnings.
1483
1484 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1485 supplied by Stefan Weber. I made a further small update for 7.8 because
1486 there is a change of source arrangements: the pcre_searchfuncs.c module is
1487 replaced by pcre_ucd.c.
1488
1489
1490 Version 7.7 07-May-08
1491 ---------------------
1492
1493 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1494 a string to a long long, pretend we don't even have a long long." This is
1495 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1496
1497 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1498 pre-7.6 versions, which defined a global no_arg variable instead of putting
1499 it in the RE class. (See also #8 below.)
1500
1501 3. Remove a line of dead code, identified by coverity and reported by Nuno
1502 Lopes.
1503
1504 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1505
1506 (1) The include/exclude patterns were being applied to the whole pathnames
1507 of files, instead of just to the final components.
1508
1509 (2) If there was more than one level of directory, the subdirectories were
1510 skipped unless they satisfied the include/exclude conditions. This is
1511 inconsistent with GNU grep (and could even be seen as contrary to the
1512 pcregrep specification - which I improved to make it absolutely clear).
1513 The action now is always to scan all levels of directory, and just
1514 apply the include/exclude patterns to regular files.
1515
1516 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1517 --exclude_dir in the tests to avoid scanning .svn directories.
1518
1519 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1520 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1521 doesn't support NULs in patterns.
1522
1523 7. Added some missing "const"s to declarations of static tables in
1524 pcre_compile.c and pcre_dfa_exec.c.
1525
1526 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1527 caused by fix #2 above. (Subsequently also a second patch to fix the
1528 first patch. And a third patch - this was a messy problem.)
1529
1530 9. Applied Craig's patch to remove the use of push_back().
1531
1532 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1533 matching function regexec().
1534
1535 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1536 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1537 references. PCRE supports relative numbers with this syntax (I don't think
1538 Oniguruma does).
1539
1540 12. Previously, a group with a zero repeat such as (...){0} was completely
1541 omitted from the compiled regex. However, this means that if the group
1542 was called as a subroutine from elsewhere in the pattern, things went wrong
1543 (an internal error was given). Such groups are now left in the compiled
1544 pattern, with a new opcode that causes them to be skipped at execution
1545 time.
1546
1547 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1548 to the way PCRE behaves:
1549
1550 (a) A lone ] character is dis-allowed (Perl treats it as data).
1551
1552 (b) A back reference to an unmatched subpattern matches an empty string
1553 (Perl fails the current match path).
1554
1555 (c) A data ] in a character class must be notated as \] because if the
1556 first data character in a class is ], it defines an empty class. (In
1557 Perl it is not possible to have an empty class.) The empty class []
1558 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1559 The negative empty class [^] matches any one character, independently
1560 of the DOTALL setting.
1561
1562 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1563 non-existent subpattern following a character class starting with ']' and
1564 containing () gave an internal compiling error instead of "reference to
1565 non-existent subpattern". Fortunately, when the pattern did exist, the
1566 compiled code was correct. (When scanning forwards to check for the
1567 existencd of the subpattern, it was treating the data ']' as terminating
1568 the class, so got the count wrong. When actually compiling, the reference
1569 was subsequently set up correctly.)
1570
1571 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1572 it was being rejected as not supported by pcre_dfa_exec(), even though
1573 other assertions are supported. I have made pcre_dfa_exec() support
1574 (*FAIL).
1575
1576 16. The implementation of 13c above involved the invention of a new opcode,
1577 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1578 cannot be changed at match time, I realized I could make a small
1579 improvement to matching performance by compiling OP_ALLANY instead of
1580 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1581 on the OP_ANY path.
1582
1583 17. Compiling pcretest on Windows with readline support failed without the
1584 following two fixes: (1) Make the unistd.h include conditional on
1585 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1586
1587 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1588 ncurses library to be included for pcretest when ReadLine support is
1589 requested, but also to allow for it to be overridden. This patch came from
1590 Daniel Bergström.
1591
1592 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1593 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1594 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1595 spotting this.
1596
1597
1598 Version 7.6 28-Jan-08
1599 ---------------------
1600
1601 1. A character class containing a very large number of characters with
1602 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1603 overflow.
1604
1605 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1606 HAVE_LONG_LONG is not defined.
1607
1608 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1609 bring them up to date and include new features. This patch includes:
1610
1611 - Fixed PH's badly added libz and libbz2 support.
1612 - Fixed a problem with static linking.
1613 - Added pcredemo. [But later removed - see 7 below.]
1614 - Fixed dftables problem and added an option.
1615 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1616 HAVE_LONG_LONG.
1617 - Added readline support for pcretest.
1618 - Added an listing of the option settings after cmake has run.
1619
1620 4. A user submitted a patch to Makefile that makes it easy to create
1621 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1622 Makefile.am that cause it to include this special target, without
1623 affecting anything else. Note that the same mingw target plus all
1624 the other distribution libraries and programs are now supported
1625 when configuring with CMake (see 6 below) instead of with
1626 Configure/Make.
1627
1628 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1629 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1630 exported in the Windows port". It has not yet been confirmed that the patch
1631 solves the problem, but it does no harm.
1632
1633 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1634 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1635 with CMake, and also correct the comment about stack recursion.
1636
1637 7. Remove the automatic building of pcredemo from the ./configure system and
1638 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1639 of a program that users should build themselves after PCRE is installed, so
1640 building it automatically is not really right. What is more, it gave
1641 trouble in some build environments.
1642
1643 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1644
1645
1646 Version 7.5 10-Jan-08
1647 ---------------------
1648
1649 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1650 values in parens when parsing an RE using the C++ wrapper."
1651
1652 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1653 Characters greater than 255 were excluded from the class instead of being
1654 included.
1655
1656 3. The same bug as (2) above applied to negated POSIX classes such as
1657 [:^space:].
1658
1659 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1660 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1661 I have changed it.
1662
1663 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1664 first named subpattern) and a construct such as (?&a) would reference the
1665 first named subpattern whose name started with "a" (in other words, the
1666 length check was missing). Both these problems are fixed. "Subpattern name
1667 expected" is now given for (?&) (a zero-length name), and this patch also
1668 makes it give the same error for \k'' (previously it complained that that
1669 was a reference to a non-existent subpattern).
1670
1671 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1672 this is right because (?- can be followed by option settings as well as by
1673 digits. I have, however, made the messages clearer.
1674
1675 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1676 than the number used in the conditional) now cause a compile-time error.
1677 This is actually not compatible with Perl, which accepts such patterns, but
1678 treats the conditional as always being FALSE (as PCRE used to), but it
1679 seems to me that giving a diagnostic is better.
1680
1681 8. Change "alphameric" to the more common word "alphanumeric" in comments
1682 and messages.
1683
1684 9. Fix two occurrences of "backslash" in comments that should have been
1685 "backspace".
1686
1687 10. Remove two redundant lines of code that can never be obeyed (their function
1688 was moved elsewhere).
1689
1690 11. The program that makes PCRE's Unicode character property table had a bug
1691 which caused it to generate incorrect table entries for sequences of
1692 characters that have the same character type, but are in different scripts.
1693 It amalgamated them into a single range, with the script of the first of
1694 them. In other words, some characters were in the wrong script. There were
1695 thirteen such cases, affecting characters in the following ranges:
1696
1697 U+002b0 - U+002c1
1698 U+0060c - U+0060d
1699 U+0061e - U+00612
1700 U+0064b - U+0065e
1701 U+0074d - U+0076d
1702 U+01800 - U+01805
1703 U+01d00 - U+01d77
1704 U+01d9b - U+01dbf
1705 U+0200b - U+0200f
1706 U+030fc - U+030fe
1707 U+03260 - U+0327f
1708 U+0fb46 - U+0fbb1
1709 U+10450 - U+1049d
1710
1711 12. The -o option (show only the matching part of a line) for pcregrep was not
1712 compatible with GNU grep in that, if there was more than one match in a
1713 line, it showed only the first of them. It now behaves in the same way as
1714 GNU grep.
1715
1716 13. If the -o and -v options were combined for pcregrep, it printed a blank
1717 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1718 does the same. The return code can be used to tell if there were any
1719 non-matching lines.
1720
1721 14. Added --file-offsets and --line-offsets to pcregrep.
1722
1723 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1724 infinitely looping recursion. The bug was that positive lookaheads were not
1725 being skipped when checking for a possible empty match (negative lookaheads
1726 and both kinds of lookbehind were skipped).
1727
1728 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1729 inclusion of <windows.h> to before rather than after the definition of
1730 INVALID_FILE_ATTRIBUTES (patch from David Byron).
1731
1732 17. Specifying a possessive quantifier with a specific limit for a Unicode
1733 character property caused pcre_compile() to compile bad code, which led at
1734 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1735 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1736 caused the error; without that there was no problem.
1737
1738 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1739
1740 19. Added --enable-pcretest-libreadline.
1741
1742 20. In pcrecpp.cc, the variable 'count' was incremented twice in
1743 RE::GlobalReplace(). As a result, the number of replacements returned was
1744 double what it should be. I removed one of the increments, but Craig sent a
1745 later patch that removed the other one (the right fix) and added unit tests
1746 that check the return values (which was not done before).
1747
1748 21. Several CMake things:
1749
1750 (1) Arranged that, when cmake is used on Unix, the libraries end up with
1751 the names libpcre and libpcreposix, not just pcre and pcreposix.
1752
1753 (2) The above change means that pcretest and pcregrep are now correctly
1754 linked with the newly-built libraries, not previously installed ones.
1755
1756 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1757
1758 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1759 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1760 UTF-8 newline character). The key issue is that the pattern starts .*;
1761 this means that the match must be either at the beginning, or after a
1762 newline. The bug was in the code for advancing after a failed match and
1763 checking that the new position followed a newline. It was not taking
1764 account of UTF-8 characters correctly.
1765
1766 23. PCRE was behaving differently from Perl in the way it recognized POSIX
1767 character classes. PCRE was not treating the sequence [:...:] as a
1768 character class unless the ... were all letters. Perl, however, seems to
1769 allow any characters between [: and :], though of course it rejects as
1770 unknown any "names" that contain non-letters, because all the known class
1771 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1772 for example, whereas PCRE did not - it did not recognize a POSIX character
1773 class. This seemed a bit dangerous, so the code has been changed to be
1774 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1775 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1776 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1777 Perl does, and where it didn't before.
1778
1779 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1780 Windows environments %n is disabled by default.
1781
1782
1783 Version 7.4 21-Sep-07
1784 ---------------------
1785
1786 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1787 means that a class such as [\s] counted as "explicit reference to CR or
1788 LF". That isn't really right - the whole point of the change was to try to
1789 help when there was an actual mention of one of the two characters. So now
1790 the change happens only if \r or \n (or a literal CR or LF) character is
1791 encountered.
1792
1793 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1794 of both had grown to the point where there were only 3 bits left.
1795 Fortunately, there was spare space in the data structure, and so I have
1796 moved the internal flags into a new 16-bit field to free up more option
1797 bits.
1798
1799 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1800 but did not set the internal JCHANGED flag - either of these is enough to
1801 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1802 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1803 start sets both bits.
1804
1805 4. Added options (at build time, compile time, exec time) to change \R from
1806 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1807
1808 5. doc/pcresyntax.html was missing from the distribution.
1809
1810 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1811 compatibility, even though it is no longer used.
1812
1813 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1814 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1815 windows.h file is present (where different names are used). [This was
1816 reversed later after testing - see 16 below.]
1817
1818 8. Changed all #include <config.h> to #include "config.h". There were also
1819 some further <pcre.h> cases that I changed to "pcre.h".
1820
1821 9. When pcregrep was used with the --colour option, it missed the line ending
1822 sequence off the lines that it output.
1823
1824 10. It was pointed out to me that arrays of string pointers cause lots of
1825 relocations when a shared library is dynamically loaded. A technique of
1826 using a single long string with a table of offsets can drastically reduce
1827 these. I have refactored PCRE in four places to do this. The result is
1828 dramatic:
1829
1830 Originally: 290
1831 After changing UCP table: 187
1832 After changing error message table: 43
1833 After changing table of "verbs" 36
1834 After changing table of Posix names 22
1835
1836 Thanks to the folks working on Gregex for glib for this insight.
1837
1838 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1839 unicode-properties was also set.
1840
1841 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1842
1843 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1844 checked only for CRLF.
1845
1846 14. Added casts to pcretest.c to avoid compiler warnings.
1847
1848 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1849
1850 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1851 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1852 entirely. This removes changes made in 7 above.
1853
1854 17. The CMake files have been updated, and there is now more information about
1855 building with CMake in the NON-UNIX-USE document.
1856
1857
1858 Version 7.3 28-Aug-07
1859 ---------------------
1860
1861 1. In the rejigging of the build system that eventually resulted in 7.1, the
1862 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1863 brackets there is not right, since it causes compilers to look for an
1864 installed pcre.h, not the version that is in the source that is being
1865 compiled (which of course may be different). I have changed it back to:
1866
1867 #include "pcre.h"
1868
1869 I have a vague recollection that the change was concerned with compiling in
1870 different directories, but in the new build system, that is taken care of
1871 by the VPATH setting the Makefile.
1872
1873 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1874 when the subject happened to end in the byte 0x85 (e.g. if the last
1875 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1876 characters but of course it shouldn't be taken as a newline when it is part
1877 of another character. The bug was that, for an unlimited repeat of . in
1878 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1879 characters when looking for a newline.
1880
1881 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1882
1883 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1884 in debug output.
1885
1886 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1887 long printing in the pcrecpp unittest when running under MinGW.
1888
1889 6. ESC_K was left out of the EBCDIC table.
1890
1891 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1892 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1893 limit also applies to "virtual nesting" when a pattern is recursive, and in
1894 this case 1000 isn't so big. I have been able to remove this limit at the
1895 expense of backing off one optimization in certain circumstances. Normally,
1896 when pcre_exec() would call its internal match() function recursively and
1897 immediately return the result unconditionally, it uses a "tail recursion"
1898 feature to save stack. However, when a subpattern that can match an empty
1899 string has an unlimited repetition quantifier, it no longer makes this
1900 optimization. That gives it a stack frame in which to save the data for
1901 checking that an empty string has been matched. Previously this was taken
1902 from the 1000-entry workspace that had been reserved. So now there is no
1903 explicit limit, but more stack is used.
1904
1905 8. Applied Daniel's patches to solve problems with the import/export magic
1906 syntax that is required for Windows, and which was going wrong for the
1907 pcreposix and pcrecpp parts of the library. These were overlooked when this
1908 problem was solved for the main library.
1909
1910 9. There were some crude static tests to avoid integer overflow when computing
1911 the size of patterns that contain repeated groups with explicit upper
1912 limits. As the maximum quantifier is 65535, the maximum group length was
1913 set at 30,000 so that the product of these two numbers did not overflow a
1914 32-bit integer. However, it turns out that people want to use groups that
1915 are longer than 30,000 bytes (though not repeat them that many times).
1916 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1917 made it possible to implement the integer overflow checks in a much more
1918 dynamic way, which I have now done. The artificial limitation on group
1919 length has been removed - we now have only the limit on the total length of
1920 the compiled pattern, which depends on the LINK_SIZE setting.
1921
1922 10. Fixed a bug in the documentation for get/copy named substring when
1923 duplicate names are permitted. If none of the named substrings are set, the
1924 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1925 empty string.
1926
1927 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1928 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1929 because the ] is interpreted as the first data character and the
1930 terminating ] is not found. PCRE has been made compatible with Perl in this
1931 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1932 cause memory overwriting.
1933
1934 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1935 string has been matched (to stop an infinite loop). It was not recognizing
1936 a conditional subpattern that could match an empty string if that
1937 subpattern was within another subpattern. For example, it looped when
1938 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1939 condition was not nested. This bug has been fixed.
1940
1941 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1942 past the start of the subject in the presence of bytes with the top bit
1943 set, for example "\x8aBCD".
1944
1945 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1946 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1947
1948 14. Optimized (?!) to (*FAIL).
1949
1950 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1951 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1952 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1953 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1954 does: it's just the validity check that is more restrictive.
1955
1956 16. Inserted checks for integer overflows during escape sequence (backslash)
1957 processing, and also fixed erroneous offset values for syntax errors during
1958 backslash processing.
1959
1960 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1961 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1962
1963 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1964 caused an overrun.
1965
1966 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1967 something other than just ASCII characters) inside a group that had an
1968 unlimited repeat caused a loop at compile time (while checking to see
1969 whether the group could match an empty string).
1970
1971 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1972 [\P{Any}] did so. (Error in the code for printing property names.)
1973
1974 21. An orphan \E inside a character class could cause a crash.
1975
1976 22. A repeated capturing bracket such as (A)? could cause a wild memory
1977 reference during compilation.
1978
1979 23. There are several functions in pcre_compile() that scan along a compiled
1980 expression for various reasons (e.g. to see if it's fixed length for look
1981 behind). There were bugs in these functions when a repeated \p or \P was
1982 present in the pattern. These operators have additional parameters compared
1983 with \d, etc, and these were not being taken into account when moving along
1984 the compiled data. Specifically:
1985
1986 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1987 length.
1988
1989 (b) An item such as \pL+ within a repeated group could cause crashes or
1990 loops.
1991
1992 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1993 "reference to non-existent subpattern" error.
1994
1995 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1996
1997 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1998 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1999
2000 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2001
2002 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2003 character were causing crashes (broken optimization).
2004
2005 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2006 \p or \P) caused a compile-time loop.
2007
2008 28. More problems have arisen in unanchored patterns when CRLF is a valid line
2009 break. For example, the unstudied pattern [\r\n]A does not match the string
2010 "\r\nA" because change 7.0/46 below moves the current point on by two
2011 characters after failing to match at the start. However, the pattern \nA
2012 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2013 the same is true. There doesn't seem any very clean way out of this, but
2014 what I have chosen to do makes the common cases work: PCRE now takes note
2015 of whether there can be an explicit match for \r or \n anywhere in the
2016 pattern, and if so, 7.0/46 no longer applies. As part of this change,
2017 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2018 pattern has explicit CR or LF references.
2019
2020 29. Added (*CR) etc for changing newline setting at start of pattern.
2021
2022
2023 Version 7.2 19-Jun-07
2024 ---------------------
2025
2026 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2027 which is apparently normally available under Windows.
2028
2029 2. Re-jig the pcregrep tests with different newline settings in an attempt
2030 to make them independent of the local environment's newline setting.
2031
2032 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2033
2034 4. Some of the "internals" tests were previously cut out when the link size
2035 was not 2, because the output contained actual offsets. The recent new
2036 "Z" feature of pcretest means that these can be cut out, making the tests
2037 usable with all link sizes.
2038
2039 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2040 stack recursion. This gives a massive performance boost under BSD, but just
2041 a small improvement under Linux. However, it saves one field in the frame
2042 in all cases.
2043
2044 6. Added more features from the forthcoming Perl 5.10:
2045
2046 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2047 recursion call. It refers to the nth most recently opened parentheses.
2048
2049 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2050 to be opened parentheses.
2051
2052 (c) Conditions that refer to capturing parentheses can be specified
2053 relatively, for example, (?(-2)... or (?(+3)...
2054
2055 (d) \K resets the start of the current match so that everything before
2056 is not part of it.
2057
2058 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2059
2060 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2061 reference syntax.
2062
2063 (g) (?| introduces a group in which the numbering of parentheses in each
2064 alternative starts with the same number.
2065
2066 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2067
2068 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2069 PCRE_INFO_JCHANGED.
2070
2071 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2072 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2073 for detecting groups that can match an empty string.
2074
2075 9. A pattern with a very large number of alternatives (more than several
2076 hundred) was running out of internal workspace during the pre-compile
2077 phase, where pcre_compile() figures out how much memory will be needed. A
2078 bit of new cunning has reduced the workspace needed for groups with
2079 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2080 workspace instead of running out of the 4096 that are available.
2081
2082 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2083
2084 11. Applied patch from Google to remove an optimization that didn't quite work.
2085 The report of the bug said:
2086
2087 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2088 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2089 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2090
2091 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2092 it matched the wrong number of bytes.
2093
2094
2095 Version 7.1 24-Apr-07
2096 ---------------------
2097
2098 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2099 that is more "standard", making use of automake and other Autotools. There
2100 is some re-arrangement of the files and adjustment of comments consequent
2101 on this.
2102
2103 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2104 for recursive directory scanning broke on some systems because the files
2105 are not scanned in any specific order and on different systems the order
2106 was different. A call to "sort" has been inserted into RunGrepTest for the
2107 approprate test as a short-term fix. In the longer term there may be an
2108 alternative.
2109
2110 3. I had an email from Eric Raymond about problems translating some of PCRE's
2111 man pages to HTML (despite the fact that I distribute HTML pages, some
2112 people do their own conversions for various reasons). The problems
2113 concerned the use of low-level troff macros .br and .in. I have therefore
2114 removed all such uses from the man pages (some were redundant, some could
2115 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2116 HTML has been updated to handle .nf/.fi and to complain if it encounters
2117 .br or .in.
2118
2119 4. Updated comments in configure.ac that get placed in config.h.in and also
2120 arranged for config.h to be included in the distribution, with the name
2121 config.h.generic, for the benefit of those who have to compile without
2122 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2123
2124 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2125 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2126 makevp.bat for the current PCRE, using the additional files
2127 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2128
2129 6. A Windows user reported a minor discrepancy with test 2, which turned out
2130 to be caused by a trailing space on an input line that had got lost in his
2131 copy. The trailing space was an accident, so I've just removed it.
2132
2133 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2134 that is needed.
2135
2136 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2137 as "const" (a) because they are and (b) because it helps the PHP
2138 maintainers who have recently made a script to detect big data structures
2139 in the php code that should be moved to the .rodata section. I remembered
2140 to update Builducptable as well, so it won't revert if ucptable.h is ever
2141 re-created.
2142
2143 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2144 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2145 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2146 support is not required. This saves 1.5-2K of code, which is important in
2147 some applications.
2148
2149 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2150 so as not to refer to the tables, even though these functions will never be
2151 called when UTF-8 support is disabled. Otherwise there are problems with a
2152 shared library.
2153
2154 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2155
2156 (a) It was defining its arguments as char * instead of void *.
2157
2158 (b) It was assuming that all moves were upwards in memory; this was true
2159 a long time ago when I wrote it, but is no longer the case.
2160
2161 The emulated memove() is provided for those environments that have neither
2162 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2163 is clearly not the case, as these two bugs were recently reported.
2164
2165 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2166 and Detrail to create the HTML documentation, the .txt form of the man
2167 pages, and it removes trailing spaces from listed files. It also creates
2168 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2169 case, it wraps all the #defines with #ifndefs. This script should be run
2170 before "make dist".
2171
2172 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2173 with Unicode property support.
2174
2175 (a) For a maximizing quantifier, if the two different cases of the
2176 character were of different lengths in their UTF-8 codings (there are
2177 some cases like this - I found 11), and the matching function had to
2178 back up over a mixture of the two cases, it incorrectly assumed they
2179 were both the same length.
2180
2181 (b) When PCRE was configured to use the heap rather than the stack for
2182 recursion during matching, it was not correctly preserving the data for
2183 the other case of a UTF-8 character when checking ahead for a match
2184 while processing a minimizing repeat. If the check also involved
2185 matching a wide character, but failed, corruption could cause an
2186 erroneous result when trying to check for a repeat of the original
2187 character.
2188
2189 13. Some tidying changes to the testing mechanism:
2190
2191 (a) The RunTest script now detects the internal link size and whether there
2192 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2193 values substituted by "configure". (The RunGrepTest script already did
2194 this for UTF-8.) The configure.ac script no longer substitutes the
2195 relevant variables.
2196
2197 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2198 with length and offset values. This means that the output is different
2199 for different internal link sizes. Test 2 is skipped for link sizes
2200 other than 2 because of this, bypassing the problem. Unfortunately,
2201 there was also a test in test 3 (the locale tests) that used /B and
2202 failed for link sizes other than 2. Rather than cut the whole test out,
2203 I have added a new /Z option to pcretest that replaces the length and
2204 offset values with spaces. This is now used to make test 3 independent
2205 of link size. (Test 2 will be tidied up later.)
2206
2207 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2208 segmentation fault instead of returning the appropriate error message.
2209
2210 15. In multiline mode when the newline sequence was set to "any", the pattern
2211 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2212 This doesn't seem right; it now treats the CRLF combination as the line
2213 ending, and so does not match in that case. It's only a pattern such as ^$
2214 that would hit this one: something like ^ABC$ would have failed after \r
2215 and then tried again after \r\n.
2216
2217 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2218 in an attempt to make files that differ only in their line terminators
2219 compare equal. This works on Linux.
2220
2221 17. Under certain error circumstances pcregrep might try to free random memory
2222 as it exited. This is now fixed, thanks to valgrind.
2223
2224 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2225 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2226 was because its rules for how to advance for /g after matching an empty
2227 string at the end of a line did not allow for this case. They now check for
2228 it specially.
2229
2230 20. pcretest is supposed to handle patterns and data of any length, by
2231 extending its buffers when necessary. It was getting this wrong when the
2232 buffer for a data line had to be extended.
2233
2234 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2235 CRLF as a newline sequence.
2236
2237 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2238 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2239 I have nevertheless tidied it up.
2240
2241 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2242
2243 24. Added a man page for pcre-config.
2244
2245
2246 Version 7.0 19-Dec-06
2247 ---------------------
2248
2249 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2250 moving to gcc 4.1.1.
2251
2252 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2253 sys/time.h, which is documented as needed for this function. It doesn't
2254 seem to matter on Linux, but it showed up on some releases of OS X.
2255
2256 3. It seems that there are systems where bytes whose values are greater than
2257 127 match isprint() in the "C" locale. The "C" locale should be the
2258 default when a C program starts up. In most systems, only ASCII printing
2259 characters match isprint(). This difference caused the output from pcretest
2260 to vary, making some of the tests fail. I have changed pcretest so that:
2261
2262 (a) When it is outputting text in the compiled version of a pattern, bytes
2263 other than 32-126 are always shown as hex escapes.
2264
2265 (b) When it is outputting text that is a matched part of a subject string,
2266 it does the same, unless a different locale has been set for the match
2267 (using the /L modifier). In this case, it uses isprint() to decide.
2268
2269 4. Fixed a major bug that caused incorrect computation of the amount of memory
2270 required for a compiled pattern when options that changed within the
2271 pattern affected the logic of the preliminary scan that determines the
2272 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2273 that the computed length was too small. The symptoms of this bug were
2274 either the PCRE error "internal error: code overflow" from pcre_compile(),
2275 or a glibc crash with a message such as "pcretest: free(): invalid next
2276 size (fast)". Examples of patterns that provoked this bug (shown in
2277 pcretest format) are:
2278
2279 /(?-x: )/x
2280 /(?x)(?-x: \s*#\s*)/
2281 /((?i)[\x{c0}])/8
2282 /(?i:[\x{c0}])/8
2283
2284 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2285 is now done differently.
2286
2287 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2288 wrapper classes; (b) implement a new function in the C++ scanner that is
2289 more efficient than the old way of doing things because it avoids levels of
2290 recursion in the regex matching; (c) add a paragraph to the documentation
2291 for the FullMatch() function.
2292
2293 6. The escape sequence \n was being treated as whatever was defined as
2294 "newline". Not only was this contrary to the documentation, which states
2295 that \n is character 10 (hex 0A), but it also went horribly wrong when
2296 "newline" was defined as CRLF. This has been fixed.
2297
2298 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2299 was being set to -1 for the "end of line" case (supposedly a value that no
2300 character can have). Though this value is never used (the check for end of
2301 line is "zero bytes in current character"), it caused compiler complaints.
2302 I've changed it to 0xffffffff.
2303
2304 8. In pcre_version.c, the version string was being built by a sequence of
2305 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2306 string (as it is for production releases) called a macro with an empty
2307 argument. The C standard says the result of this is undefined. The gcc
2308 compiler treats it as an empty string (which was what was wanted) but it is
2309 reported that Visual C gives an error. The source has been hacked around to
2310 avoid this problem.
2311
2312 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2313 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2314 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2315 of them did).
2316
2317 10. Originally, pcretest opened its input and output without "b"; then I was
2318 told that "b" was needed in some environments, so it was added for release
2319 5.0 to both the input and output. (It makes no difference on Unix-like
2320 systems.) Later I was told that it is wrong for the input on Windows. I've
2321 now abstracted the modes into two macros, to make it easier to fiddle with
2322 them, and removed "b" from the input mode under Windows.
2323
2324 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2325
2326 12. Added -help and --help to pcretest as an official way of being reminded
2327 of the options.
2328
2329 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2330 and pcrecpp.cc because they annoy compilers at high warning levels.
2331
2332 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2333
2334 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2335 scripts are not C programs :-) and which was not noticed because it works
2336 on Linux.
2337
2338 16. pcretest is supposed to handle any length of pattern and data line (as one
2339 line or as a continued sequence of lines) by extending its input buffer if
2340 necessary. This feature was broken for very long pattern lines, leading to
2341 a string of junk being passed to pcre_compile() if the pattern was longer
2342 than about 50K.
2343
2344 17. I have done a major re-factoring of the way pcre_compile() computes the
2345 amount of memory needed for a compiled pattern. Previously, there was code
2346 that made a preliminary scan of the pattern in order to do this. That was
2347 OK when PCRE was new, but as the facilities have expanded, it has become
2348 harder and harder to keep it in step with the real compile phase, and there
2349 have been a number of bugs (see for example, 4 above). I have now found a
2350 cunning way of running the real compile function in a "fake" mode that
2351 enables it to compute how much memory it would need, while actually only
2352 ever using a few hundred bytes of working memory and without too many
2353 tests of the mode. This should make future maintenance and development
2354 easier. A side effect of this work is that the limit of 200 on the nesting
2355 depth of parentheses has been removed (though this was never a serious
2356 limitation, I suspect). However, there is a downside: pcre_compile() now
2357 runs more slowly than before (30% or more, depending on the pattern). I
2358 hope this isn't a big issue. There is no effect on runtime performance.
2359
2360 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2361 newline (only possible for the last line of a file) and it was a
2362 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2363
2364 19. Added additional timing features to pcretest. (1) The -tm option now times
2365 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2366 separate command line item, by a number that specifies the number of
2367 repeats to use when timing. The default is 50000; this gives better
2368 precision, but takes uncomfortably long for very large patterns.
2369
2370 20. Extended pcre_study() to be more clever in cases where a branch of a
2371 subpattern has no definite first character. For example, (a*|b*)[cd] would
2372 previously give no result from pcre_study(). Now it recognizes that the
2373 first character must be a, b, c, or d.
2374
2375 21. There was an incorrect error "recursive call could loop indefinitely" if
2376 a subpattern (or the entire pattern) that was being tested for matching an
2377 empty string contained only one non-empty item after a nested subpattern.
2378 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2379 incorrectly, because the \d was being skipped in the check.
2380
2381 22. The pcretest program now has a new pattern option /B and a command line
2382 option -b, which is equivalent to adding /B to every pattern. This causes
2383 it to show the compiled bytecode, without the additional information that
2384 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2385 is the same as /B/I).
2386
2387 23. A new optimization is now able automatically to treat some sequences such
2388 as a*b as a*+b. More specifically, if something simple (such as a character
2389 or a simple class like \d) has an unlimited quantifier, and is followed by
2390 something that cannot possibly match the quantified thing, the quantifier
2391 is automatically "possessified".
2392
2393 24. A recursive reference to a subpattern whose number was greater than 39
2394 went wrong under certain circumstances in UTF-8 mode. This bug could also
2395 have affected the operation of pcre_study().
2396
2397 25. Realized that a little bit of performance could be had by replacing
2398 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2399
2400 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2401
2402 27. Possessive quantifiers such as a++ were previously implemented by turning
2403 them into atomic groups such as ($>a+). Now they have their own opcodes,
2404 which improves performance. This includes the automatically created ones
2405 from 23 above.
2406
2407 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2408 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2409 the first matched character to be a colon. This applied both to named and
2410 numbered groups.
2411
2412 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2413
2414 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2415 building PCRE on Windows easier, so I have included it in the distribution.
2416
2417 31. There is now a check in pcretest against a ridiculously large number being
2418 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2419 loop, the loop is abandoned.
2420
2421 32. Forward references to subpatterns in conditions such as (?(2)...) where
2422 subpattern 2 is defined later cause pcre_compile() to search forwards in
2423 the pattern for the relevant set of parentheses. This search went wrong
2424 when there were unescaped parentheses in a character class, parentheses
2425 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2426
2427 33. "Subroutine" calls and backreferences were previously restricted to
2428 referencing subpatterns earlier in the regex. This restriction has now
2429 been removed.
2430
2431 34. Added a number of extra features that are going to be in Perl 5.10. On the
2432 whole, these are just syntactic alternatives for features that PCRE had
2433 previously implemented using the Python syntax or my own invention. The
2434 other formats are all retained for compatibility.
2435
2436 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2437 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2438 also .NET compatible.
2439
2440 (b) A recursion or subroutine call to a named group can now be defined as
2441 (?&name) as well as (?P>name).
2442
2443 (c) A backreference to a named group can now be defined as \k<name> or
2444 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2445 5.10, are also .NET compatible.
2446
2447 (d) A conditional reference to a named group can now use the syntax
2448 (?(<name>) or (?('name') as well as (?(name).
2449
2450 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2451 groups (named and numbered) that are never evaluated inline, but can be
2452 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2453 is always false. There may be only one alternative in such a group.
2454
2455 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2456 as the simple (?(R). The condition is true only if the most recent
2457 recursion is that of the given number or name. It does not search out
2458 through the entire recursion stack.
2459
2460 (g) The escape \gN or \g{N} has been added, where N is a positive or
2461 negative number, specifying an absolute or relative reference.
2462
2463 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2464 some "unreachable code" warnings.
2465
2466 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2467 things, this adds five new scripts.
2468
2469 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2470 There were also incompatibilities regarding the handling of \Q..\E inside
2471 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2472 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2473
2474 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2475 matches an empty string, and forcibly breaks the loop. There were bugs in
2476 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2477 against aaaa the result was just "a" rather than "aaaa", for example. Two
2478 separate and independent bugs (that affected different cases) have been
2479 fixed.
2480
2481 39. Refactored the code to abolish the use of different opcodes for small
2482 capturing bracket numbers. This is a tidy that I avoided doing when I
2483 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2484 The new approach is not only tidier, it makes it possible to reduce the
2485 memory needed to fix the previous bug (38).
2486
2487 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2488 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2489 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2490 mode.
2491
2492 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2493 report.
2494
2495 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2496 copy construction and assignment in the C++ wrapper.
2497
2498 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2499 couple of bugs that could have given wrong results in the "--newline=crlf"
2500 case.
2501
2502 44. Added a number of casts and did some reorganization of signed/unsigned int
2503 variables following suggestions from Dair Grant. Also renamed the variable
2504 "this" as "item" because it is a C++ keyword.
2505
2506 45. Arranged for dftables to add
2507
2508 #include "pcre_internal.h"
2509
2510 to pcre_chartables.c because without it, gcc 4.x may remove the array
2511 definition from the final binary if PCRE is built into a static library and
2512 dead code stripping is activated.
2513
2514 46. For an unanchored pattern, if a match attempt fails at the start of a
2515 newline sequence, and the newline setting is CRLF or ANY, and the next two
2516 characters are CRLF, advance by two characters instead of one.
2517
2518
2519 Version 6.7 04-Jul-06
2520 ---------------------
2521
2522 1. In order to handle tests when input lines are enormously long, pcretest has
2523 been re-factored so that it automatically extends its buffers when
2524 necessary. The code is crude, but this _is_ just a test program. The
2525 default size has been increased from 32K to 50K.
2526
2527 2. The code in pcre_study() was using the value of the re argument before
2528 testing it for NULL. (Of course, in any sensible call of the function, it
2529 won't be NULL.)
2530
2531 3. The memmove() emulation function in pcre_internal.h, which is used on
2532 systems that lack both memmove() and bcopy() - that is, hardly ever -
2533 was missing a "static" storage class specifier.
2534
2535 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2536 containing an extended class (one that cannot be represented by a bitmap
2537 because it contains high-valued characters or Unicode property items, e.g.
2538 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2539 pattern, but PCRE should not loop if you do not (it no longer does).
2540 [Detail: two cases were found: (a) a repeated subpattern containing an
2541 extended class; (b) a recursive reference to a subpattern that followed a
2542 previous extended class. It wasn't skipping over the extended class
2543 correctly when UTF-8 mode was not set.]
2544
2545 5. A negated single-character class was not being recognized as fixed-length
2546 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2547 compile error "lookbehind assertion is not fixed length".
2548
2549 6. The RunPerlTest auxiliary script was showing an unexpected difference
2550 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2551 write a Perl script that can interpret lines of an input file either as
2552 byte characters or as UTF-8, which is what "perltest" was being required to
2553 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2554 can't do is switch easily at run time between having the "use utf8;" pragma
2555 or not. In the end, I fudged it by using the RunPerlTest script to insert
2556 "use utf8;" explicitly for the UTF-8 tests.
2557
2558 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2559 the end of the subject string, contrary to the documentation and to what
2560 Perl does. This was true of both matching functions. Now it matches only at
2561 the start of the subject and immediately after *internal* newlines.
2562
2563 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2564 a pointer to an int instead of a pointer to an unsigned long int. This
2565 caused problems on 64-bit systems.
2566
2567 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2568 instance of the 'standard' template library not being so standard".
2569
2570 10. There was no check on the number of named subpatterns nor the maximum
2571 length of a subpattern name. The product of these values is used to compute
2572 the size of the memory block for a compiled pattern. By supplying a very
2573 long subpattern name and a large number of named subpatterns, the size
2574 computation could be caused to overflow. This is now prevented by limiting
2575 the length of names to 32 characters, and the number of named subpatterns
2576 to 10,000.
2577
2578 11. Subpatterns that are repeated with specific counts have to be replicated in
2579 the compiled pattern. The size of memory for this was computed from the
2580 length of the subpattern and the repeat count. The latter is limited to
2581 65535, but there was no limit on the former, meaning that integer overflow
2582 could in principle occur. The compiled length of a repeated subpattern is
2583 now limited to 30,000 bytes in order to prevent this.
2584
2585 12. Added the optional facility to have named substrings with the same name.
2586
2587 13. Added the ability to use a named substring as a condition, using the
2588 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2589 are numbers (not recommended). Forward references are permitted.
2590
2591 14. Added forward references in named backreferences (if you see what I mean).
2592
2593 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2594 pattern could run off the end of the subject. For example, the pattern
2595 "(?s)(.{1,5})"8 did this with the subject "ab".
2596
2597 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2598 PCRE_CASELESS was set when matching characters that were quantified with ?
2599 or *.
2600
2601 17. A character class other than a single negated character that had a minimum
2602 but no maximum quantifier - for example [ab]{6,} - was not handled
2603 correctly by pce_dfa_exec(). It would match only one character.
2604
2605 18. A valid (though odd) pattern that looked like a POSIX character
2606 class but used an invalid character after [ (for example [[,abc,]]) caused
2607 pcre_compile() to give the error "Failed: internal error: code overflow" or
2608 in some cases to crash with a glibc free() error. This could even happen if
2609 the pattern terminated after [[ but there just happened to be a sequence of
2610 letters, a binary zero, and a closing ] in the memory that followed.
2611
2612 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2613 over the years. Originally (before any Unicode support), just the bottom 8
2614 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2615 output from "man perlunicode" includes this:
2616
2617 The regular expression compiler produces polymorphic opcodes. That
2618 is, the pattern adapts to the data and automatically switches to
2619 the Unicode character scheme when presented with Unicode data--or
2620 instead uses a traditional byte scheme when presented with byte
2621 data.
2622
2623 Sadly, a wide octal escape does not cause a switch, and in a string with
2624 no other multibyte characters, these octal escapes are treated as before.
2625 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2626 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2627 Unicode string.
2628
2629 I have not perpetrated such confusion in PCRE. Up till now, it took just
2630 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2631 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2632 translate to the appropriate multibyte character.
2633
2634 29. Applied some refactoring to reduce the number of warnings from Microsoft
2635 and Borland compilers. This has included removing the fudge introduced
2636 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2637 a warning about an unused variable.
2638
2639 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2640 characters since release 4.0, because Perl (from release 5.004) does not.
2641 [Or at least, is documented not to: some releases seem to be in conflict
2642 with the documentation.] However, when a pattern was studied with
2643 pcre_study() and all its branches started with \s, PCRE still included VT
2644 as a possible starting character. Of course, this did no harm; it just
2645 caused an unnecessary match attempt.
2646
2647 22. Removed a now-redundant internal flag bit that recorded the fact that case
2648 dependency changed within the pattern. This was once needed for "required
2649 byte" processing, but is no longer used. This recovers a now-scarce options
2650 bit. Also moved the least significant internal flag bit to the most-
2651 significant bit of the word, which was not previously used (hangover from
2652 the days when it was an int rather than a uint) to free up another bit for
2653 the future.
2654
2655 23. Added support for CRLF line endings as well as CR and LF. As well as the
2656 default being selectable at build time, it can now be changed at runtime
2657 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2658 specify that it is scanning data with non-default line endings.
2659
2660 24. Changed the definition of CXXLINK to make it agree with the definition of
2661 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2662
2663 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2664 recursions. This makes a big different to stack usage for some patterns.
2665
2666 26. If a subpattern containing a named recursion or subroutine reference such
2667 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2668 the space required for the compiled pattern went wrong and gave too small a
2669 value. Depending on the environment, this could lead to "Failed: internal
2670 error: code overflow at offset 49" or "glibc detected double free or
2671 corruption" errors.
2672
2673 27. Applied patches from Google (a) to support the new newline modes and (b) to
2674 advance over multibyte UTF-8 characters in GlobalReplace.
2675
2676 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2677 difference for some implementation of PCRE in some Windows version.
2678
2679 29. Added some extra testing facilities to pcretest:
2680
2681 \q<number> in a data line sets the "match limit" value
2682 \Q<number> in a data line sets the "match recursion limt" value
2683 -S <number> sets the stack size, where <number> is in megabytes
2684
2685 The -S option isn't available for Windows.
2686
2687
2688 Version 6.6 06-Feb-06
2689 ---------------------
2690
2691 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2692 in pcreposix.h. I have copied the definition from pcre.h.
2693
2694 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2695 because pcre.h is no longer a built file.
2696
2697 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2698 not normally included in the compiled code.
2699
2700
2701 Version 6.5 01-Feb-06
2702 ---------------------
2703
2704 1. When using the partial match feature with pcre_dfa_exec(), it was not
2705 anchoring the second and subsequent partial matches at the new starting
2706 point. This could lead to incorrect results. For example, with the pattern
2707 /1234/, partially matching against "123" and then "a4" gave a match.
2708
2709 2. Changes to pcregrep:
2710
2711 (a) All non-match returns from pcre_exec() were being treated as failures
2712 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2713 error message is output. Some extra information is given for the
2714 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2715 probably the only errors that are likely to be caused by users (by
2716 specifying a regex that has nested indefinite repeats, for instance).
2717 If there are more than 20 of these errors, pcregrep is abandoned.
2718
2719 (b) A binary zero was treated as data while matching, but terminated the
2720 output line if it was written out. This has been fixed: binary zeroes
2721 are now no different to any other data bytes.
2722
2723 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2724 used to set a locale for matching. The --locale=xxxx long option has
2725 been added (no short equivalent) to specify a locale explicitly on the
2726 pcregrep command, overriding the environment variables.
2727
2728 (d) When -B was used with -n, some line numbers in the output were one less
2729 than they should have been.
2730
2731 (e) Added the -o (--only-matching) option.
2732
2733 (f) If -A or -C was used with -c (count only), some lines of context were
2734 accidentally printed for the final match.
2735
2736 (g) Added the -H (--with-filename) option.
2737
2738 (h) The combination of options -rh failed to suppress file names for files
2739 that were found from directory arguments.
2740
2741 (i) Added the -D (--devices) and -d (--directories) options.
2742
2743 (j) Added the -F (--fixed-strings) option.
2744
2745 (k) Allow "-" to be used as a file name for -f as well as for a data file.
2746
2747 (l) Added the --colo(u)r option.
2748
2749 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2750 is not present by default.
2751
2752 3. A nasty bug was discovered in the handling of recursive patterns, that is,
2753 items such as (?R) or (?1), when the recursion could match a number of
2754 alternatives. If it matched one of the alternatives, but subsequently,
2755 outside the recursion, there was a failure, the code tried to back up into
2756 the recursion. However, because of the way PCRE is implemented, this is not
2757 possible, and the result was an incorrect result from the match.
2758
2759 In order to prevent this happening, the specification of recursion has
2760 been changed so that all such subpatterns are automatically treated as
2761 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2762
2763 4. I had overlooked the fact that, in some locales, there are characters for
2764 which isalpha() is true but neither isupper() nor islower() are true. In
2765 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2766 and ordfeminine) are like this. This affected the treatment of \w and \W
2767 when they appeared in character classes, but not when they appeared outside
2768 a character class. The bit map for "word" characters is now created
2769 separately from the results of isalnum() instead of just taking it from the
2770 upper, lower, and digit maps. (Plus the underscore character, of course.)
2771
2772 5. The above bug also affected the handling of POSIX character classes such as
2773 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2774 permanent tables. Instead, the bit maps for such a class were previously
2775 created as the appropriate unions of the upper, lower, and digit bitmaps.
2776 Now they are created by subtraction from the [[:word:]] class, which has
2777 its own bitmap.
2778
2779 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2780 It is created by subtracting the vertical space characters (\x09, \x0a,
2781 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2782 subtraction was done in the overall bitmap for a character class, meaning
2783 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2784 be recognized. This bug has been fixed.
2785
2786 7. Patches from the folks at Google:
2787
2788 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2789 real life, but is still worth protecting against".
2790
2791 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2792 regular expressions".
2793
2794 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2795 have it.
2796
2797 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2798 "configure" and the latter not, in order to fix a problem somebody had
2799 with compiling the Arg class on HP-UX.
2800
2801 (e) Improve the error-handling of the C++ wrapper a little bit.
2802
2803 (f) New tests for checking recursion limiting.
2804
2805 8. The pcre_memmove() function, which is used only if the environment does not
2806 have a standard memmove() function (and is therefore rarely compiled),
2807 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2808 returning a result (though PCRE never actually uses the result).
2809
2810 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2811 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2812 returned instead of calling malloc() with an overflowing number that would
2813 most likely cause subsequent chaos.
2814
2815 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2816
2817 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2818 with this option is matched, the nmatch and pmatch options of regexec() are
2819 ignored.
2820
2821 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2822 provided in case anyone wants to the the POSIX interface with UTF-8
2823 strings.
2824
2825 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2826 C++ linking (needed for some HP-UX environments).
2827
2828 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2829 (unused parameter) and in the pcre_printint() function (omitted "default"
2830 switch label when the default is to do nothing).
2831
2832 15. Added some code to make it possible, when PCRE is compiled as a C++
2833 library, to replace subject pointers for pcre_exec() with a smart pointer
2834 class, thus making it possible to process discontinuous strings.
2835
2836 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2837 much the same function. They were added by different people who were trying
2838 to make PCRE easy to compile on non-Unix systems. It has been suggested
2839 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2840 for compiling on Windows systems. I have therefore replaced it with
2841 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2842 defaults to "extern" for C or "extern C" for C++, which works fine on
2843 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2844 SCOPE with something explicit in config.h. In addition:
2845
2846 (a) pcreposix.h still had just "extern" instead of either of these macros;
2847 I have replaced it with PCRE_DATA_SCOPE.
2848
2849 (b) Functions such as _pcre_xclass(), which are internal to the library,
2850 but external in the C sense, all had PCRE_EXPORT in their definitions.
2851 This is apparently wrong for the Windows case, so I have removed it.
2852 (It makes no difference on Unix-like systems.)
2853
2854 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2855 of recursive calls to match(). This is different to MATCH_LIMIT because
2856 that limits the total number of calls to match(), not all of which increase
2857 the depth of recursion. Limiting the recursion depth limits the amount of
2858 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2859 when PCRE is compiled, and changed at run time. A patch from Google adds
2860 this functionality to the C++ interface.
2861
2862 18. Changes to the handling of Unicode character properties:
2863
2864 (a) Updated the table to Unicode 4.1.0.
2865
2866 (b) Recognize characters that are not in the table as "Cn" (undefined).
2867
2868 (c) I revised the way the table is implemented to a much improved format
2869 which includes recognition of ranges. It now supports the ranges that
2870 are defined in UnicodeData.txt, and it also amalgamates other
2871 characters into ranges. This has reduced the number of entries in the
2872 table from around 16,000 to around 3,000, thus reducing its size
2873 considerably. I realized I did not need to use a tree structure after
2874 all - a binary chop search is just as efficient. Having reduced the
2875 number of entries, I extended their size from 6 bytes to 8 bytes to
2876 allow for more data.
2877
2878 (d) Added support for Unicode script names via properties such as \p{Han}.
2879
2880 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2881 matching that character.
2882
2883 20. When matching a repeated Unicode property with a minimum greater than zero,
2884 (for example \pL{2,}), PCRE could look past the end of the subject if it
2885 reached it while seeking the minimum number of characters. This could
2886 happen only if some of the characters were more than one byte long, because
2887 there is a check for at least the minimum number of bytes.
2888
2889 21. Refactored the implementation of \p and \P so as to be more general, to
2890 allow for more different types of property in future. This has changed the
2891 compiled form incompatibly. Anybody with saved compiled patterns that use
2892 \p or \P will have to recompile them.
2893
2894 22. Added "Any" and "L&" to the supported property types.
2895
2896 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2897 but give a compile time error if the value is greater than 0xff.
2898
2899 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2900 accidentally not being installed or uninstalled.
2901
2902 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2903 made were to insert the current release number. This seemed silly, because
2904 it made things harder for people building PCRE on systems that don't run
2905 "configure". I have turned pcre.h into a distributed file, no longer built
2906 by "configure", with the version identification directly included. There is
2907 no longer a pcre.h.in file.
2908
2909 However, this change necessitated a change to the pcre-config script as
2910 well. It is built from pcre-config.in, and one of the substitutions was the
2911 release number. I have updated configure.ac so that ./configure now finds
2912 the release number by grepping pcre.h.
2913
2914 26. Added the ability to run the tests under valgrind.
2915
2916
2917 Version 6.4 05-Sep-05
2918 ---------------------
2919
2920 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2921 "--" to be printed when multiple files were scanned, even when none of the
2922 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2923 consider it to be a bug, and have restored the previous behaviour.
2924
2925 2. A couple of code tidies to get rid of compiler warnings.
2926
2927 3. The pcretest program used to cheat by referring to symbols in the library
2928 whose names begin with _pcre_. These are internal symbols that are not
2929 really supposed to be visible externally, and in some environments it is
2930 possible to suppress them. The cheating is now confined to including
2931 certain files from the library's source, which is a bit cleaner.
2932
2933 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2934 file's purpose clearer.
2935
2936 5. Reorganized pcre_ucp_findchar().
2937
2938
2939 Version 6.3 15-Aug-05
2940 ---------------------
2941
2942 1. The file libpcre.pc.in did not have general read permission in the tarball.
2943
2944 2. There were some problems when building without C++ support:
2945
2946 (a) If C++ support was not built, "make install" and "make test" still
2947 tried to test it.
2948
2949 (b) There were problems when the value of CXX was explicitly set. Some
2950 changes have been made to try to fix these, and ...
2951
2952 (c) --disable-cpp can now be used to explicitly disable C++ support.
2953
2954 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2955 backslash in a target when C++ was disabled. This confuses some
2956 versions of "make", apparently. Using an intermediate variable solves
2957 this. (Same for CPP_LOBJ.)
2958
2959 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2960 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2961 necessary on certain architectures.
2962
2963 4. Added a setting of -export-symbols-regex to the link command to remove
2964 those symbols that are exported in the C sense, but actually are local
2965 within the library, and not documented. Their names all begin with
2966 "_pcre_". This is not a perfect job, because (a) we have to except some
2967 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2968 available (and never for static libraries). I have made a note to try to
2969 find a way round (a) in the future.
2970
2971
2972 Version 6.2 01-Aug-05
2973 ---------------------
2974
2975 1. There was no test for integer overflow of quantifier values. A construction
2976 such as {1111111111111111} would give undefined results. What is worse, if
2977 a minimum quantifier for a parenthesized subpattern overflowed and became
2978 negative, the calculation of the memory size went wrong. This could have
2979 led to memory overwriting.
2980
2981 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2982
2983 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2984 operating environments where this matters.
2985
2986 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2987 PCRE options from within the C++ wrapper.
2988
2989 5. Named capturing subpatterns were not being correctly counted when a pattern
2990 was compiled. This caused two problems: (a) If there were more than 100
2991 such subpatterns, the calculation of the memory needed for the whole
2992 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2993 back references of the form \12, where the number was greater than 9, were
2994 not recognized as back references, even though there were sufficient
2995 previous subpatterns.
2996
2997 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2998 versions of gcc, e.g. 2.95.4.
2999
3000
3001 Version 6.1 21-Jun-05
3002 ---------------------
3003
3004 1. There was one reference to the variable "posix" in pcretest.c that was not
3005 surrounded by "#if !defined NOPOSIX".
3006
3007 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3008 the cross-check on the old pcre_info() function, for the benefit of the
3009 cut-down version of PCRE that is currently imported into Exim.
3010
3011 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3012 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3013 patterns that start (?i) but I don't think that matters. The use of (?i) is
3014 just an example; this all applies to the other options as well.
3015
3016 4. Since libtool seems to echo the compile commands it is issuing, the output
3017 from "make" can be reduced a bit by putting "@" in front of each libtool
3018 compile command.
3019
3020 5. Patch from the folks at Google for configure.in to be a bit more thorough
3021 in checking for a suitable C++ installation before trying to compile the
3022 C++ stuff. This should fix a reported problem when a compiler was present,
3023 but no suitable headers.
3024
3025 6. The man pages all had just "PCRE" as their title. I have changed them to
3026 be the relevant file name. I have also arranged that these names are
3027 retained in the file doc/pcre.txt, which is a concatenation in text format
3028 of all the man pages except the little individual ones for each function.
3029
3030 7. The NON-UNIX-USE file had not been updated for the different set of source
3031 files that come with release 6. I also added a few comments about the C++
3032 wrapper.
3033
3034
3035 Version 6.0 07-Jun-05
3036 ---------------------
3037
3038 1. Some minor internal re-organization to help with my DFA experiments.
3039
3040 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3041 didn't matter for the library itself when fully configured, but did matter
3042 when compiling without UCP support, or within Exim, where the ucp files are
3043 not imported.
3044
3045 3. Refactoring of the library code to split up the various functions into
3046 different source modules. The addition of the new DFA matching code (see
3047 below) to a single monolithic source would have made it really too
3048 unwieldy, quite apart from causing all the code to be include in a
3049 statically linked application, when only some functions are used. This is
3050 relevant even without the DFA addition now that patterns can be compiled in
3051 one application and matched in another.
3052
3053 The downside of splitting up is that there have to be some external
3054 functions and data tables that are used internally in different modules of
3055 the library but which are not part of the API. These have all had their
3056 names changed to start with "_pcre_" so that they are unlikely to clash
3057 with other external names.
3058
3059 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3060 a different (DFA) algorithm. Although it is slower than the original
3061 function, it does have some advantages for certain types of matching
3062 problem.
3063
3064 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3065 including restarting after a partial match.
3066
3067 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3068 defined when compiling for Windows was sent to me. I have put it into the
3069 code, though I have no means of testing or verifying it.
3070
3071 7. Added the pcre_refcount() auxiliary function.
3072
3073 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3074 match before or at the first newline in the subject string. In pcretest,
3075 the /f option on a pattern can be used to set this.
3076
3077 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3078 would behave wrongly. This has been present in PCRE since release 4.0.
3079
3080 10. A number of changes to the pcregrep command:
3081
3082 (a) Refactored how -x works; insert ^(...)$ instead of setting
3083 PCRE_ANCHORED and checking the length, in preparation for adding
3084 something similar for -w.
3085
3086 (b) Added the -w (match as a word) option.
3087
3088 (c) Refactored the way lines are read and buffered so as to have more
3089 than one at a time available.
3090
3091 (d) Implemented a pcregrep test script.
3092
3093 (e) Added the -M (multiline match) option. This allows patterns to match
3094 over several lines of the subject. The buffering ensures that at least
3095 8K, or the rest of the document (whichever is the shorter) is available
3096 for matching (and similarly the previous 8K for lookbehind assertions).
3097
3098 (f) Changed the --help output so that it now says
3099
3100 -w, --word-regex(p)
3101
3102 instead of two lines, one with "regex" and the other with "regexp"
3103 because that confused at least one person since the short forms are the
3104 same. (This required a bit of code, as the output is generated
3105 automatically from a table. It wasn't just a text change.)
3106
3107 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3108 option but starts with a hyphen. Could be a pattern or a path name
3109 starting with a hyphen, for instance.
3110
3111 (h) "-" can be given as a file name to represent stdin.
3112
3113 (i) When file names are being printed, "(standard input)" is used for
3114 the standard input, for compatibility with GNU grep. Previously
3115 "<stdin>" was used.
3116
3117 (j) The option --label=xxx can be used to supply a name to be used for
3118 stdin when file names are being printed. There is no short form.
3119
3120 (k) Re-factored the options decoding logic because we are going to add
3121 two more options that take data. Such options can now be given in four
3122 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3123
3124 (l) Added the -A, -B, and -C options for requesting that lines of context
3125 around matches be printed.
3126
3127 (m) Added the -L option to print the names of files that do not contain
3128 any matching lines, that is, the complement of -l.
3129
3130 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3131 continue to scan other files.
3132
3133 (o) The -s option was incorrectly implemented. For compatibility with other
3134 greps, it now suppresses the error message for a non-existent or non-
3135 accessible file (but not the return code). There is a new option called
3136 -q that suppresses the output of matching lines, which was what -s was
3137 previously doing.
3138
3139 (p) Added --include and --exclude options to specify files for inclusion
3140 and exclusion when recursing.
3141
3142 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3143 Hopefully, it now does.
3144
3145 12. Missing cast in pcre_study().
3146
3147 13. Added an "uninstall" target to the makefile.
3148
3149 14. Replaced "extern" in the function prototypes in Makefile.in with
3150 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3151 world, but is set differently for Windows.
3152
3153 15. Added a second compiling function called pcre_compile2(). The only
3154 difference is that it has an extra argument, which is a pointer to an
3155 integer error code. When there is a compile-time failure, this is set
3156 non-zero, in addition to the error test pointer being set to point to an
3157 error message. The new argument may be NULL if no error number is required
3158 (but then you may as well call pcre_compile(), which is now just a
3159 wrapper). This facility is provided because some applications need a
3160 numeric error indication, but it has also enabled me to tidy up the way
3161 compile-time errors are handled in the POSIX wrapper.
3162
3163 16. Added VPATH=.libs to the makefile; this should help when building with one
3164 prefix path and installing with another. (Or so I'm told by someone who
3165 knows more about this stuff than I do.)
3166
3167 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3168 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3169 match everything, including newlines. This is not POSIX-compatible, but
3170 somebody wanted the feature. From pcretest it can be activated by using
3171 both the P and the s flags.
3172
3173 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3174
3175 19. libpcre.pc was being incorrectly installed as executable.
3176
3177 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3178 it now also looks for '\r' so that it will work unmodified on Windows.
3179
3180 21. Added Google's contributed C++ wrapper to the distribution.
3181
3182 22. Added some untidy missing memory free() calls in pcretest, to keep
3183 Electric Fence happy when testing.
3184
3185
3186
3187 Version 5.0 13-Sep-04
3188 ---------------------
3189
3190 1. Internal change: literal characters are no longer packed up into items
3191 containing multiple characters in a single byte-string. Each character
3192 is now matched using a separate opcode. However, there may be more than one
3193 byte in the character in UTF-8 mode.
3194
3195 2. The pcre_callout_block structure has two new fields: pattern_position and
3196 next_item_length. These contain the offset in the pattern to the next match
3197 item, and its length, respectively.
3198
3199 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3200 insertion of callouts before each pattern item. Added the /C option to
3201 pcretest to make use of this.
3202
3203 4. On the advice of a Windows user, the lines
3204
3205 #if defined(_WIN32) || defined(WIN32)
3206 _setmode( _fileno( stdout ), 0x8000 );
3207 #endif /* defined(_WIN32) || defined(WIN32) */
3208
3209 have been added to the source of pcretest. This apparently does useful
3210 magic in relation to line terminators.
3211
3212 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3213 for the benefit of those environments where the "b" makes a difference.
3214
3215 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3216 to know about it. I have put a hack into configure.in that adds in code
3217 to set GCC=yes if CC=icc. This seems to end up at a point in the
3218 generated configure script that is early enough to affect the setting of
3219 compiler options, which is what is needed, but I have no means of testing
3220 whether it really works. (The user who reported this had patched the
3221 generated configure script, which of course I cannot do.)
3222
3223 LATER: After change 22 below (new libtool files), the configure script
3224 seems to know about icc (and also ecc). Therefore, I have commented out
3225 this hack in configure.in.
3226
3227 7. Added support for pkg-config (2 patches were sent in).
3228
3229 8. Negated POSIX character classes that used a combination of internal tables
3230 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3231 [[:^ascii]]. Typically, they would match almost any characters. The other
3232 POSIX classes were not broken in this way.
3233
3234 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3235 to find the match, as PCRE was deluded into thinking that the match had to
3236 start at the start point or following a newline. The same bug applied to
3237 patterns with negative forward assertions or any backward assertions
3238 preceding ".*" at the start, unless the pattern required a fixed first
3239 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3240
3241 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3242 starting at the last subject character, bytes beyond the end of the subject
3243 string were read.
3244
3245 11. Renamed the variable "class" as "classbits" to make life easier for C++
3246 users. (Previously there was a macro definition, but it apparently wasn't
3247 enough.)
3248
3249 12. Added the new field "tables" to the extra data so that tables can be passed
3250 in at exec time, or the internal tables can be re-selected. This allows
3251 a compiled regex to be saved and re-used at a later time by a different
3252 program that might have everything at different addresses.
3253
3254 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3255 -R library as well as a -L library.
3256
3257 14. The debugging options of pcretest (-d on the command line or D on a
3258 pattern) showed incorrect output for anything following an extended class
3259 that contained multibyte characters and which was followed by a quantifier.
3260
3261 15. Added optional support for general category Unicode character properties
3262 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3263 support. It adds about 90K to the size of the library. The meanings of the
3264 inbuilt class escapes such as \d and \s have NOT been changed.
3265
3266 16. Updated pcredemo.c to include calls to free() to release the memory for the
3267 compiled pattern.
3268
3269 17. The generated file chartables.c was being created in the source directory
3270 instead of in the building directory. This caused the build to fail if the
3271 source directory was different from the building directory, and was
3272 read-only.
3273
3274 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3275 file. No doubt somebody will tell me if they don't make sense... Also added
3276 Dan Mooney's comments about building on OpenVMS.
3277
3278 19. Added support for partial matching via the PCRE_PARTIAL option for
3279 pcre_exec() and the \P data escape in pcretest.
3280
3281 20. Extended pcretest with 3 new pattern features:
3282
3283 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3284 write the compiled pattern to the file whose name is "rest-of-line".
3285 This is a straight binary dump of the data, with the saved pointer to
3286 the character tables forced to be NULL. The study data, if any, is
3287 written too. After writing, pcretest reads a new pattern.
3288
3289 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3290 compiled pattern from the given file. There must not be any
3291 occurrences of "<" in the file name (pretty unlikely); if there are,
3292 pcretest will instead treat the initial "<" as a pattern delimiter.
3293 After reading in the pattern, pcretest goes on to read data lines as
3294 usual.
3295
3296 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3297 and 16-bit fields in a compiled pattern, to simulate a pattern that
3298 was compiled on a host of opposite endianness.
3299
3300 21. The pcre-exec() function can now cope with patterns that were compiled on
3301 hosts of opposite endianness, with this restriction:
3302
3303 As for any compiled expression that is saved and used later, the tables
3304 pointer field cannot be preserved; the extra_data field in the arguments
3305 to pcre_exec() should be used to pass in a tables address if a value
3306 other than the default internal tables were used at compile time.
3307
3308 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3309 now diagnosed as an error. Previously, most of the time, a negative number
3310 would have been treated as zero, but if in addition "ovector" was passed as
3311 NULL, a crash could occur.
3312
3313 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3314 new versions from the libtool 1.5 distribution (the last one is a copy of
3315 a file called libtool.m4). This seems to have fixed the need to patch
3316 "configure" to support Darwin 1.3 (which I used to do). However, I still
3317 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3318 workstation).
3319
3320 24. Changed the PCRE licence to be the more standard "BSD" licence.
3321
3322
3323 Version 4.5 01-Dec-03
3324 ---------------------
3325
3326 1. There has been some re-arrangement of the code for the match() function so
3327 that it can be compiled in a version that does not call itself recursively.
3328 Instead, it keeps those local variables that need separate instances for
3329 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3330 needs to "recurse". Keeping track of where control must go is done by means
3331 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3332 hide most of the details from the main code, and operates only if
3333 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3334 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3335 operating.
3336
3337 To make it easier for callers to provide specially tailored get/free
3338 functions for this usage, two new functions, pcre_stack_malloc, and
3339 pcre_stack_free, are used. They are always called in strict stacking order,
3340 and the size of block requested is always the same.
3341
3342 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3343 PCRE has been compiled to use the stack or the heap for recursion. The
3344 -C option of pcretest uses this to show which version is compiled.
3345
3346 A new data escape \S, is added to pcretest; it causes the amounts of store
3347 obtained and freed by both kinds of malloc/free at match time to be added
3348 to the output.
3349
3350 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3351 what's available on my current Linux desktop machine.
3352
3353 3. When matching a UTF-8 string, the test for a valid string at the start has
3354 been extended. If start_offset is not zero, PCRE now checks that it points
3355 to a byte that is the start of a UTF-8 character. If not, it returns
3356 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3357 this is necessary because there may be backward assertions in the pattern.
3358 When matching the same subject several times, it may save resources to use
3359 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3360
3361 4. The code for checking the validity of UTF-8 strings has been tightened so
3362 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3363 containing "overlong sequences".
3364
3365 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3366 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3367 should have been "&", but it just so happened that all the cases this let
3368 through by mistake were picked up later in the function.
3369
3370 6. I had used a variable called "isblank" - this is a C99 function, causing
3371 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3372
3373 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3374 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3375 so the version is shown; (c) stop "make test" echoing "./RunTest".
3376
3377 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3378
3379 9. The prototype for memmove() for systems that don't have it was using
3380 size_t, but the inclusion of the header that defines size_t was later. I've
3381 moved the #includes for the C headers earlier to avoid this.
3382
3383 10. Added some adjustments to the code to make it easier to compiler on certain
3384 special systems:
3385
3386 (a) Some "const" qualifiers were missing.
3387 (b) Added the macro EXPORT before all exported functions; by default this
3388 is defined to be empty.
3389 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3390 that it reads its output file name as an argument instead of writing
3391 to the standard output and assuming this can be redirected.
3392
3393 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3394 class containing characters with values greater than 255, PCRE compilation
3395 went into a loop.
3396
3397 12. A recursive reference to a subpattern that was within another subpattern
3398 that had a minimum quantifier of zero caused PCRE to crash. For example,
3399 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3400 recursion. If the recursively-called subpattern itself had a zero repeat,
3401 that was OK.
3402
3403 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3404 buffer into which it was copied (for escape processing) was still set at
3405 1024, so long lines caused crashes.
3406
3407 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3408 "internal error: code overflow...". This applied to any character class
3409 that was followed by a possessive quantifier.
3410
3411 15. Modified the Makefile to add libpcre.la as a prerequisite for
3412 libpcreposix.la because I was told this is needed for a parallel build to
3413 work.
3414
3415 16. If a pattern that contained .* following optional items at the start was
3416 studied, the wrong optimizing data was generated, leading to matching
3417 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3418 matching string must start with a or b or c. The correct conclusion for
3419 this pattern is that a match can start with any character.
3420
3421
3422 Version 4.4 13-Aug-03
3423 ---------------------
3424
3425 1. In UTF-8 mode, a character class containing characters with values between
3426 127 and 255 was not handled correctly if the compiled pattern was studied.
3427 In fixing this, I have also improved the studying algorithm for such
3428 classes (slightly).
3429
3430 2. Three internal functions had redundant arguments passed to them. Removal
3431 might give a very teeny performance improvement.
3432
3433 3. Documentation bug: the value of the capture_top field in a callout is *one
3434 more than* the number of the hightest numbered captured substring.
3435
3436 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3437 in incorrectly linking with a previously installed version. They now link
3438 explicitly with libpcre.la.
3439
3440 5. configure.in no longer needs to recognize Cygwin specially.
3441
3442 6. A problem in pcre.in for Windows platforms is fixed.
3443
3444 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3445 pcretest, it used to include the size of the study block as part of its
3446 output. Unfortunately, the structure contains a field that has a different
3447 size on different hardware architectures. This meant that the tests that
3448 showed this size failed. As the block is currently always of a fixed size,
3449 this information isn't actually particularly useful in pcretest output, so
3450 I have just removed it.
3451
3452 8. Three pre-processor statements accidentally did not start in column 1.
3453 Sadly, there are *still* compilers around that complain, even though
3454 standard C has not required this for well over a decade. Sigh.
3455
3456 9. In pcretest, the code for checking callouts passed small integers in the
3457 callout_data field, which is a void * field. However, some picky compilers
3458 complained about the casts involved for this on 64-bit systems. Now
3459 pcretest passes the address of the small integer instead, which should get
3460 rid of the warnings.
3461
3462 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3463 both compile and run time, and gives an error if an invalid UTF-8 sequence
3464 is found. There is a option for disabling this check in cases where the
3465 string is known to be correct and/or the maximum performance is wanted.
3466
3467 11. In response to a bug report, I changed one line in Makefile.in from
3468
3469 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3470 to
3471 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3472
3473 to look similar to other lines, but I have no way of telling whether this
3474 is the right thing to do, as I do not use Windows. No doubt I'll get told
3475 if it's wrong...
3476
3477
3478 Version 4.3 21-May-03
3479 ---------------------
3480
3481 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3482 Makefile.
3483
3484 2. Some refactoring to improve the quality of the code:
3485
3486 (i) The utf8_table... variables are now declared "const".
3487
3488 (ii) The code for \cx, which used the "case flipping" table to upper case
3489 lower case letters, now just substracts 32. This is ASCII-specific,
3490 but the whole concept of \cx is ASCII-specific, so it seems
3491 reasonable.
3492
3493 (iii) PCRE was using its character types table to recognize decimal and
3494 hexadecimal digits in the pattern. This is silly, because it handles
3495 only 0-9, a-f, and A-F, but the character types table is locale-
3496 specific, which means strange things might happen. A private
3497 table is now used for this - though it costs 256 bytes, a table is
3498 much faster than multiple explicit tests. Of course, the standard
3499 character types table is still used for matching digits in subject
3500 strings against \d.
3501
3502 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3503 ending in _t are). So I've renamed it as ESC_tee.
3504
3505 3. The first argument for regexec() in the POSIX wrapper should have been
3506 defined as "const".
3507
3508 4. Changed pcretest to use malloc() for its buffers so that they can be
3509 Electric Fenced for debugging.
3510
3511 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3512 to read one or more bytes before the start of the subject string. Often this
3513 had no effect on PCRE's behaviour, but in some circumstances it could
3514 provoke a segmentation fault.
3515
3516 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3517 to try to read one or more bytes before the start of the subject string.
3518
3519 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3520 UTF-8 support could misbehave in various ways if the subject string
3521 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3522 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3523 back over UTF-8 characters.)
3524
3525
3526 Version 4.2 14-Apr-03
3527 ---------------------
3528
3529 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3530
3531 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3532 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3533 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3534 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3535 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3536 and BUILD_EXEEXT
3537 Note: automatic setting of the BUILD variables is not yet working
3538 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3539 compile-time but not at link-time
3540 [LINK]: use for linking executables only
3541 make different versions for Windows and non-Windows
3542 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3543 libraries
3544 [LINK_FOR_BUILD]: new variable
3545 [OBJEXT]: use throughout
3546 [EXEEXT]: use throughout
3547 <winshared>: new target
3548 <wininstall>: new target
3549 <dftables.o>: use native compiler
3550 <dftables>: use native linker
3551 <install>: handle Windows platform correctly
3552 <clean>: ditto
3553 <check>: ditto
3554 copy DLL to top builddir before testing
3555
3556 As part of these changes, -no-undefined was removed again. This was reported
3557 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3558 in any case.
3559
3560 3. Some tidies to get rid of compiler warnings:
3561
3562 . In the match_data structure, match_limit was an unsigned long int, whereas
3563 match_call_count was an int. I've made them both unsigned long ints.
3564
3565 . In pcretest the fact that a const uschar * doesn't automatically cast to
3566 a void * provoked a warning.
3567
3568 . Turning on some more compiler warnings threw up some "shadow" variables
3569 and a few more missing casts.
3570
3571 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3572 option, a class that contained a single character with a value between 128
3573 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3574
3575 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3576 option, a class that contained several characters, but with at least one
3577 whose value was between 128 and 255 caused PCRE to crash.
3578
3579
3580 Version 4.1 12-Mar-03
3581 ---------------------
3582
3583 1. Compiling with gcc -pedantic found a couple of places where casts were
3584 needed, and a string in dftables.c that was longer than standard compilers are
3585 required to support.
3586
3587 2. Compiling with Sun's compiler found a few more places where the code could
3588 be tidied up in order to avoid warnings.
3589
3590 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3591 first of these names is deprecated in the latest Autoconf in favour of the name
3592 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3593 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3594 analogy I have changed it to CFLAGS_FOR_BUILD.
3595
3596 4. Added -no-undefined to the linking command in the Makefile, because this is
3597 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3598 linking step for the pcreposix library.
3599
3600 5. PCRE was failing to diagnose the case of two named groups with the same
3601 name.
3602
3603 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3604 literal character that is needed in the subject for a match, and scans along to
3605 ensure that it is present before embarking on the full matching process. This
3606 saves time in cases of nested unlimited repeats that are never going to match.
3607 Problem: the scan can take a lot of time if the subject is very long (e.g.
3608 megabytes), thus penalizing straightforward matches. It is now done only if the
3609 amount of subject to be scanned is less than 1000 bytes.
3610
3611 7. A lesser problem with the same optimization is that it was recording the
3612 first character of an anchored pattern as "needed", thus provoking a search
3613 right along the subject, even when the first match of the pattern was going to
3614 fail. The "needed" character is now not set for anchored patterns, unless it
3615 follows something in the pattern that is of non-fixed length. Thus, it still
3616 fulfils its original purpose of finding quick non-matches in cases of nested
3617 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3618
3619
3620 Version 4.0 17-Feb-03
3621 ---------------------
3622
3623 1. If a comment in an extended regex that started immediately after a meta-item
3624 extended to the end of string, PCRE compiled incorrect data. This could lead to
3625 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3626
3627 2. Moved to autoconf 2.53 and libtool 1.4.2.
3628
3629 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3630 the special perltest8 script is no longer needed - all the tests can be run
3631 from a single perltest script.
3632
3633 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3634 by \s. It has now been removed in PCRE. This means it isn't recognized as
3635 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3636 class [:space:] *does* include VT, thereby creating a mess.
3637
3638 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3639 space and tab.
3640
3641 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3642 its new features into the main test script, reducing the number of scripts.
3643
3644 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3645 were backward compatible, and made the (?i) apply to the whole pattern, as if
3646 /i were given. Now it behaves more logically, and applies the option setting
3647 only to what follows. PCRE has been changed to follow suit. However, if it
3648 finds options settings right at the start of the pattern, it extracts them into
3649 the global options, as before. Thus, they show up in the info data.
3650
3651 8. Added support for the \Q...\E escape sequence. Characters in between are
3652 treated as literals. This is slightly different from Perl in that $ and @ are
3653 also handled as literals inside the quotes. In Perl, they will cause variable
3654 interpolation. Note the following examples:
3655
3656 Pattern PCRE matches Perl matches
3657
3658 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3659 \Qabc\$xyz\E abc\$xyz abc\$xyz
3660 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3661
3662 For compatibility with Perl, \Q...\E sequences are recognized inside character
3663 classes as well as outside them.
3664
3665 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3666 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3667 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3668 signed/unsigned warnings.
3669
3670 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3671 option for pcretest, so I've replaced it by a simple function that does just
3672 that job.
3673
3674 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3675 "pcregrep -".
3676
3677 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3678 Java package. This provides some syntactic sugar for simple cases of what my
3679 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3680 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3681 item, you can use this simplified notation. Note that only makes sense with
3682 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3683 greediness, whatever the setting of the PCRE_UNGREEDY option.
3684
3685 13. A change of greediness default within a pattern was not taking effect at
3686 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3687 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3688 was abstracted outside.
3689
3690 14. PCRE now supports the \G assertion. It is true when the current matching
3691 position is at the start point of the match. This differs from \A when the
3692 starting offset is non-zero. Used with the /g option of pcretest (or similar
3693 code), it works in the same way as it does for Perl's /g option. If all
3694 alternatives of a regex begin with \G, the expression is anchored to the start
3695 match position, and the "anchored" flag is set in the compiled expression.
3696
3697 15. Some bugs concerning the handling of certain option changes within patterns
3698 have been fixed. These applied to options other than (?ims). For example,
3699 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3700 been the other way round. Some of this was related to change 7 above.
3701
3702 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3703 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3704 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3705 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3706
3707 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3708 mode. Unlike ".", it always matches newline, whatever the setting of
3709 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3710 assertions. Perl allows it, but it doesn't (in general) work because it can't
3711 calculate the length of the lookbehind. At least, that's the case for Perl
3712 5.8.0 - I've been told they are going to document that it doesn't work in
3713 future.
3714
3715 18. Added an error diagnosis for escapes that PCRE does not support: these are
3716 \L, \l, \N, \P, \p, \U, \u, and \X.
3717
3718 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3719 reading past the end of the pattern in cases such as /[abcd/.
3720
3721 20. PCRE was getting more memory than necessary for patterns with classes that
3722 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3723
3724 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3725 compiling PCRE for use with Virtual Pascal.
3726
3727 22. Small fix to the Makefile to make it work properly if the build is done
3728 outside the source tree.
3729
3730 23. Added a new extension: a condition to go with recursion. If a conditional
3731 subpattern starts with (?(R) the "true" branch is used if recursion has
3732 happened, whereas the "false" branch is used only at the top level.
3733
3734 24. When there was a very long string of literal characters (over 255 bytes
3735 without UTF support, over 250 bytes with UTF support), the computation of how
3736 much memory was required could be incorrect, leading to segfaults or other
3737 strange effects.
3738
3739 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3740 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3741 there was a subsequent back reference to those brackets. This meant that, for
3742 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3743 possible to check for precisely this case. All we can do is abandon the
3744 optimization if .* occurs inside capturing brackets when there are any back
3745 references whatsoever. (See below for a better fix that came later.)
3746
3747 26. The handling of the optimization for finding the first character of a
3748 non-anchored pattern, and for finding a character that is required later in the
3749 match were failing in some cases. This didn't break the matching; it just
3750 failed to optimize when it could. The way this is done has been re-implemented.
3751
3752 27. Fixed typo in error message for invalid (?R item (it said "(?p").
3753
3754 28. Added a new feature that provides some of the functionality that Perl
3755 provides with (?{...}). The facility is termed a "callout". The way it is done
3756 in PCRE is for the caller to provide an optional function, by setting
3757 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3758 global variable. By default it is unset, which disables all calling out. To get
3759 the function called, the regex must include (?C) at appropriate points. This
3760 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3761 This provides a means of identifying different callout points. When PCRE
3762 reaches such a point in the regex, if pcre_callout has been set, the external
3763 function is called. It is provided with data in a structure called
3764 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3765 matching continues; if it returns a non-zero value, the match at the current
3766 point fails. However, backtracking will occur if possible. [This was changed
3767 later and other features added - see item 49 below.]
3768
3769 29. pcretest is upgraded to test the callout functionality. It provides a
3770 callout function that displays information. By default, it shows the start of
3771 the match and the current position in the text. There are some new data escapes
3772 to vary what happens:
3773
3774 \C+ in addition, show current contents of captured substrings
3775 \C- do not supply a callout function
3776 \C!n return 1 when callout number n is reached
3777 \C!n!m return 1 when callout number n is reached for the mth time
3778
3779 30. If pcregrep was called with the -l option and just a single file name, it
3780 output "<stdin>" if a match was found, instead of the file name.
3781
3782 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3783 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3784 pcre_exec(). This saves a malloc/free per call. The default value of
3785 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3786 when configuring.
3787
3788 32. The default maximum size of a compiled pattern is 64K. There have been a
3789 few cases of people hitting this limit. The code now uses macros to handle the
3790 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3791 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3792 configuring. Tests 2 and 5 work only with 2-byte links because they output
3793 debugging information about compiled patterns.
3794
3795 33. Internal code re-arrangements:
3796
3797 (a) Moved the debugging function for printing out a compiled regex into
3798 its own source file (printint.c) and used #include to pull it into
3799 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3800 separate copies.
3801
3802 (b) Defined the list of op-code names for debugging as a macro in
3803 internal.h so that it is next to the definition of the opcodes.
3804
3805 (c) Defined a table of op-code lengths for simpler skipping along compiled
3806 code. This is again a macro in internal.h so that it is next to the
3807 definition of the opcodes.
3808
3809 34. Added support for recursive calls to individual subpatterns, along the
3810 lines of Robin Houston's patch (but implemented somewhat differently).
3811
3812 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3813 allow it to read and process whole directories in Win32. This code was
3814 contributed by Lionel Fourquaux; it has not been tested by me.
3815
3816 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3817 used to name a group. Names consist of alphanumerics and underscores, and must
3818 be unique. Back references use the syntax (?P=name) and recursive calls use
3819 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3820 numbers. The function pcre_fullinfo() can be used after compilation to extract
3821 a name/number map. There are three relevant calls:
3822
3823 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3824 PCRE_INFO_NAMECOUNT yields the number of entries
3825 PCRE_INFO_NAMETABLE yields a pointer to the map.
3826
3827 The map is a vector of fixed-size entries. The size of each entry depends on
3828 the length of the longest name used. The first two bytes of each entry are the
3829 group number, most significant byte first. There follows the corresponding
3830 name, zero terminated. The names are in alphabetical order.
3831
3832 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3833 case instead of 255. Making it the same both with and without UTF-8 support
3834 means that the same test output works with both.
3835
3836 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3837 calling malloc() with a zero argument.
3838
3839 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3840 optimization. I've improved things by keeping a bitmap of backreferences with
3841 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3842 fact referenced, the optimization can be applied. It is unlikely that a
3843 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3844 the match to follow \n) will appear inside brackets with a number greater than
3845 31, but if it does, any back reference > 31 suppresses the optimization.
3846
3847 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3848 of disabling numbered capturing parentheses. Any opening parenthesis that is
3849 not followed by ? behaves as if it were followed by ?: but named parentheses
3850 can still be used for capturing (and they will acquire numbers in the usual
3851 way).
3852
3853 41. Redesigned the return codes from the match() function into yes/no/error so
3854 that errors can be passed back from deep inside the nested calls. A malloc
3855 failure while inside a recursive subpattern call now causes the
3856 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3857
3858 42. It is now possible to set a limit on the number of times the match()
3859 function is called in a call to pcre_exec(). This facility makes it possible to
3860 limit the amount of recursion and backtracking, though not in a directly
3861 obvious way, because the match() function is used in a number of different
3862 circumstances. The count starts from zero for each position in the subject
3863 string (for non-anchored patterns). The default limit is, for compatibility, a
3864 large number, namely 10 000 000. You can change this in two ways:
3865
3866 (a) When configuring PCRE before making, you can use --with-match-limit=n
3867 to set a default value for the compiled library.
3868
3869 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3870 a different value is set. See 45 below.
3871
3872 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3873
3874 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3875 of things that can be changed at compile time. The first argument specifies
3876 what is wanted and the second points to where the information is to be placed.
3877 The current list of available information is:
3878
3879 PCRE_CONFIG_UTF8
3880
3881 The output is an integer that is set to one if UTF-8 support is available;
3882 otherwise it is set to zero.
3883
3884 PCRE_CONFIG_NEWLINE
3885
3886 The output is an integer that it set to the value of the code that is used for
3887 newline. It is either LF (10) or CR (13).
3888
3889 PCRE_CONFIG_LINK_SIZE
3890
3891 The output is an integer that contains the number of bytes used for internal
3892 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3893
3894 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3895
3896 The output is an integer that contains the threshold above which the POSIX
3897 interface uses malloc() for output vectors. See item 31 above.
3898
3899 PCRE_CONFIG_MATCH_LIMIT
3900
3901 The output is an unsigned integer that contains the default limit of the number
3902 of match() calls in a pcre_exec() execution. See 42 above.
3903
3904 44. pcretest has been upgraded by the addition of the -C option. This causes it
3905 to extract all the available output from the new pcre_config() function, and to
3906 output it. The program then exits immediately.
3907
3908 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3909 order to support additional features. One way would have been to define
3910 pcre_exec2() (for example) with extra arguments, but this would not have been
3911 extensible, and would also have required all calls to the original function to
3912 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3913 is used for passing in "extra" data from pcre_study().
3914
3915 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3916 contains the following fields:
3917
3918 flags a bitmap indicating which of the following fields are set
3919 study_data opaque data from pcre_study()
3920 match_limit a way of specifying a limit on match() calls for a specific
3921 call to pcre_exec()
3922 callout_data data for callouts (see 49 below)
3923
3924 The flag bits are also defined in pcre.h, and are
3925
3926 PCRE_EXTRA_STUDY_DATA
3927 PCRE_EXTRA_MATCH_LIMIT
3928 PCRE_EXTRA_CALLOUT_DATA
3929
3930 The pcre_study() function now returns one of these new pcre_extra blocks, with
3931 the actual study data pointed to by the study_data field, and the
3932 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3933 before. That is, this change is entirely upwards-compatible and requires no
3934 change to existing code.
3935
3936 If you want to pass in additional data to pcre_exec(), you can either place it
3937 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3938 block.
3939
3940 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3941 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3942 times with different match limits, until it finds the minimum value needed for
3943 pcre_exec() to complete. The value is then output. This can be instructive; for
3944 most simple matches the number is quite small, but for pathological cases it
3945 gets very large very quickly.
3946
3947 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3948 returns the size of the data block pointed to by the study_data field in a
3949 pcre_extra block, that is, the value that was passed as the argument to
3950 pcre_malloc() when PCRE was getting memory in which to place the information
3951 created by pcre_study(). The fourth argument should point to a size_t variable.
3952 pcretest has been extended so that this information is shown after a successful
3953 pcre_study() call when information about the compiled regex is being displayed.
3954
3955 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3956 because what follows is always an absolute path. (Later: it turns out that this
3957 is more than cosmetic for MinGW, because it doesn't like empty path
3958 components.)
3959
3960 49. Some changes have been made to the callout feature (see 28 above):
3961
3962 (i) A callout function now has three choices for what it returns:
3963
3964 0 => success, carry on matching
3965 > 0 => failure at this point, but backtrack if possible
3966 < 0 => serious error, return this value from pcre_exec()
3967
3968 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3969 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3970 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3971 use by callout functions. It will never be used by PCRE itself.
3972
3973 (ii) The pcre_extra structure (see 45 above) has a void * field called
3974 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3975 pcre_callout_block structure has a field of the same name. The contents of
3976 the field passed in the pcre_extra structure are passed to the callout
3977 function in the corresponding field in the callout block. This makes it
3978 easier to use the same callout-containing regex from multiple threads. For
3979 testing, the pcretest program has a new data escape
3980
3981 \C*n pass the number n (may be negative) as callout_data
3982
3983 If the callout function in pcretest receives a non-zero value as
3984 callout_data, it returns that value.
3985
3986 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3987 there were some redundant $(CFLAGS) in commands that are now specified as
3988 $(LINK), which already includes $(CFLAGS).
3989
3990 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3991 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3992 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3993 one-byte characters throughout. Note that case-insensitive matching applies
3994 only to characters whose values are less than 256. PCRE doesn't support the
3995 notion of cases for higher-valued characters.
3996
3997 (i) A character class whose characters are all within 0-255 is handled as
3998 a bit map, and the map is inverted for negative classes. Previously, a
3999 character > 255 always failed to match such a class; however it should
4000 match if the class was a negative one (e.g. [^ab]). This has been fixed.
4001
4002 (ii) A negated character class with a single character < 255 is coded as
4003 "not this character" (OP_NOT). This wasn't working properly when the test
4004 character was multibyte, either singly or repeated.
4005
4006 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
4007 mode, for example: \x{100}{2,3}.
4008
4009 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
4010 singly or repeated) now correctly test multibyte characters. However,
4011 PCRE doesn't recognize any characters with values greater than 255 as
4012 digits, spaces, or word characters. Such characters always match \D, \S,
4013 and \W, and never match \d, \s, or \w.
4014
4015 (v) Classes may now contain characters and character ranges with values
4016 greater than 255. For example: [ab\x{100}-\x{400}].
4017
4018 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
4019 PCRE in UTF-8 mode.
4020
4021 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
4022 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
4023 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
4024 value.)
4025
4026 53. The single man page has become too large. I have therefore split it up into
4027 a number of separate man pages. These also give rise to individual HTML pages;
4028 these are now put in a separate directory, and there is an index.html page that
4029 lists them all. Some hyperlinking between the pages has been installed.
4030
4031 54. Added convenience functions for handling named capturing parentheses.
4032
4033 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4034 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4035 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4036 are faulted.
4037
4038 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4039 calling configure. These values are used when compiling the dftables.c program
4040 which is run to generate the source of the default character tables. They
4041 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4042 you will need to set these values.
4043
4044 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4045
4046
4047 Version 3.9 02-Jan-02
4048 ---------------------
4049
4050 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4051
4052 2. If --disable-static was given, the building process failed when trying to
4053 build pcretest and pcregrep. (For some reason it was using libtool to compile
4054 them, which is not right, as they aren't part of the library.)
4055
4056
4057 Version 3.8 18-Dec-01
4058 ---------------------
4059
4060 1. The experimental UTF-8 code was completely screwed up. It was packing the
4061 bytes in the wrong order. How dumb can you get?
4062
4063
4064 Version 3.7 29-Oct-01
4065 ---------------------
4066
4067 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4068 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4069 this didn't happen under Solaris 8, where I normally test things.
4070
4071 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4072 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4073 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4074 this fix an hour or so after the initial 3.7 release.)
4075
4076
4077 Version 3.6 23-Oct-01
4078 ---------------------
4079
4080 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4081 offsets passed as NULL with zero offset count.
4082
4083 2. The config.guess and config.sub files had not been updated when I moved to
4084 the latest autoconf.
4085
4086
4087 Version 3.5 15-Aug-01
4088 ---------------------
4089
4090 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4091 had been forgotten.
4092
4093 2. By using declared but undefined structures, we can avoid using "void"
4094 definitions in pcre.h while keeping the internal definitions of the structures
4095 private.
4096
4097 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4098 user point of view, this means that both static and shared libraries are built
4099 by default, but this can be individually controlled. More of the work of
4100 handling this static/shared cases is now inside libtool instead of PCRE's make
4101 file.
4102
4103 4. The pcretest utility is now installed along with pcregrep because it is
4104 useful for users (to test regexs) and by doing this, it automatically gets
4105 relinked by libtool. The documentation has been turned into a man page, so
4106 there are now .1, .txt, and .html versions in /doc.
4107
4108 5. Upgrades to pcregrep:
4109 (i) Added long-form option names like gnu grep.
4110 (ii) Added --help to list all options with an explanatory phrase.
4111 (iii) Added -r, --recursive to recurse into sub-directories.
4112 (iv) Added -f, --file to read patterns from a file.
4113
4114 6. pcre_exec() was referring to its "code" argument before testing that
4115 argument for NULL (and giving an error if it was NULL).
4116
4117 7. Upgraded Makefile.in to allow for compiling in a different directory from
4118 the source directory.
4119
4120 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4121 options bits, the pointer it was passed was to an int instead of to an unsigned
4122 long int. This mattered only on 64-bit systems.
4123
4124 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4125 generated) instead of pcre.in, which it its source. Also made the same change
4126 in several of the .c files.
4127
4128 10. A new release of gcc defines printf() as a macro, which broke pcretest
4129 because it had an ifdef in the middle of a string argument for printf(). Fixed
4130 by using separate calls to printf().
4131
4132 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4133 script, to force use of CR or LF instead of \n in the source. On non-Unix
4134 systems, the value can be set in config.h.
4135
4136 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4137 absolute limit. Changed the text of the error message to make this clear, and
4138 likewise updated the man page.
4139
4140 13. The limit of 99 on the number of capturing subpatterns has been removed.
4141 The new limit is 65535, which I hope will not be a "real" limit.
4142
4143
4144 Version 3.4 22-Aug-00
4145 ---------------------
4146
4147 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4148
4149 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4150
4151
4152 Version 3.3 01-Aug-00
4153 ---------------------
4154
4155 1. If an octal character was given, but the value was greater than \377, it
4156 was not getting masked to the least significant bits, as documented. This could
4157 lead to crashes in some systems.
4158
4159 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4160 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4161
4162 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4163 These just pass their arguments on to (pcre_free)(), but they are provided
4164 because some uses of PCRE bind it to non-C systems that can call its functions,
4165 but cannot call free() or pcre_free() directly.
4166
4167 4. Add "make test" as a synonym for "make check". Corrected some comments in
4168 the Makefile.
4169
4170 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4171 Makefile.
4172
4173 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4174 command called pgrep for grepping around the active processes.
4175
4176 7. Added the beginnings of support for UTF-8 character strings.
4177
4178 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4179 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4180 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4181 out for the ar command.)
4182
4183
4184 Version 3.2 12-May-00
4185 ---------------------
4186
4187 This is purely a bug fixing release.
4188
4189 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4190 of ZA. This was just one example of several cases that could provoke this bug,
4191 which was introduced by change 9 of version 2.00. The code for breaking
4192 infinite loops after an iteration that matches an empty string was't working
4193 correctly.
4194
4195 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4196 when matched against abbab (for example). After matching an empty string, it
4197 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4198 caused it to match further down the string than it should.
4199
4200 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4201 was there because it doesn't seem to be needed, and it causes trouble on some
4202 systems, as it is not a Standard C header. It has been removed.
4203
4204 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4205 were reported on the Macintosh. The changes were from
4206
4207 while ((c = *(++ptr)) != 0 && c != '\n');
4208 to
4209 while ((c = *(++ptr)) != 0 && c != '\n') ;
4210
4211 Totally extraordinary, but if that's what it takes...
4212
4213 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4214 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4215 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4216 assumes the way PCRE uses memmove() (always moving upwards).
4217
4218 6. PCRE is being used in one environment where strchr() is not available. There
4219 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4220 faster code anyway.
4221
4222
4223 Version 3.1 09-Feb-00
4224 ---------------------
4225
4226 The only change in this release is the fixing of some bugs in Makefile.in for
4227 the "install" target:
4228
4229 (1) It was failing to install pcreposix.h.
4230
4231 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4232
4233
4234 Version 3.0 01-Feb-00
4235 ---------------------
4236
4237 1. Add support for the /+ modifier to perltest (to output $` like it does in
4238 pcretest).
4239
4240 2. Add support for the /g modifier to perltest.
4241
4242 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4243 matches null strings.
4244
4245 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4246 pattern. Perl treats empty patterns specially - it reuses the most recent
4247 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4248 effect.
4249
4250 5. The POSIX interface was broken in that it was just handing over the POSIX
4251 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4252 required a bigger vector, with some working space on the end. This means that
4253 the POSIX wrapper now has to get and free some memory, and copy the results.
4254
4255 6. Added some simple autoconf support, placing the test data and the
4256 documentation in separate directories, re-organizing some of the
4257 information files, and making it build pcre-config (a GNU standard). Also added
4258 libtool support for building PCRE as a shared library, which is now the
4259 default.
4260
4261 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4262 09 are not valid octal constants. Single digits will be used for minor values
4263 less than 10.
4264
4265 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4266 existing programs that set these in the POSIX interface can use PCRE without
4267 modification.
4268
4269 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4270 return all that pcre_info() returns, plus additional data. The pcre_info()
4271 function is retained for compatibility, but is considered to be obsolete.
4272
4273 10. Added experimental recursion feature (?R) to handle one common case that
4274 Perl 5.6 will be able to do with (?p{...}).
4275
4276 11. Added support for POSIX character classes like [:alpha:], which Perl is
4277 adopting.
4278
4279
4280 Version 2.08 31-Aug-99
4281 ----------------------
4282
4283 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4284 trying to match at the startoffset position, but instead was moving forward to
4285 the next newline as if a previous match had failed.
4286
4287 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4288 and could get into a loop if a null string was matched other than at the start
4289 of the subject.
4290
4291 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4292 be distinguished at compile time, and for completeness also added PCRE_DATE.
4293
4294 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4295 in GnuWin32 environments.
4296
4297
4298 Version 2.07 29-Jul-99
4299 ----------------------
4300
4301 1. The documentation is now supplied in plain text form and HTML as well as in
4302 the form of man page sources.
4303
4304 2. C++ compilers don't like assigning (void *) values to other pointer types.
4305 In particular this affects malloc(). Although there is no problem in Standard
4306 C, I've put in casts to keep C++ compilers happy.
4307
4308 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4309 should be (const char *).
4310
4311 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4312 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4313 However, I haven't made this a standard facility. The documentation doesn't
4314 mention it, and the Makefile doesn't support it.
4315
4316 5. The Makefile now contains an "install" target, with editable destinations at
4317 the top of the file. The pcretest program is not installed.
4318
4319 6. pgrep -V now gives the PCRE version number and date.
4320
4321 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4322 causing the entire string to be ignored, instead of just the last character.
4323
4324 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4325 non-matching string, it can take a very, very long time, even for strings of
4326 quite modest length, because of the nested recursion. PCRE now does better in
4327 some of these cases. It does this by remembering the last required literal
4328 character in the pattern, and pre-searching the subject to ensure it is present
4329 before running the real match. In other words, it applies a heuristic to detect
4330 some types of certain failure quickly, and in the above example, if presented
4331 with a string that has no trailing " it gives "no match" very quickly.
4332
4333 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4334 other alternatives are tried instead.
4335
4336
4337 Version 2.06 09-Jun-99
4338 ----------------------
4339
4340 1. Change pcretest's output for amount of store used to show just the code
4341 space, because the remainder (the data block) varies in size between 32-bit and
4342 64-bit systems.
4343
4344 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4345 start matching at. This allows lookbehinds to work when searching for multiple
4346 occurrences in a string.
4347
4348 3. Added additional options to pcretest for testing multiple occurrences:
4349
4350 /+ outputs the rest of the string that follows a match
4351 /g loops for multiple occurrences, using the new startoffset argument
4352 /G loops for multiple occurrences by passing an incremented pointer
4353
4354 4. PCRE wasn't doing the "first character" optimization for patterns starting
4355 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4356 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4357 the letter 'x'. On long subject strings, this gives a significant speed-up.
4358
4359
4360 Version 2.05 21-Apr-99
4361 ----------------------
4362
4363 1. Changed the type of magic_number from int to long int so that it works
4364 properly on 16-bit systems.
4365
4366 2. Fixed a bug which caused patterns starting with .* not to work correctly
4367 when the subject string contained newline characters. PCRE was assuming
4368 anchoring for such patterns in all cases, which is not correct because .* will
4369 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4370 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4371 must be retried after every newline in the subject.
4372
4373
4374 Version 2.04 18-Feb-99
4375 ----------------------
4376
4377 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4378 computation of the store needed to hold the pattern was incorrect (too large).
4379 If such patterns were nested a few deep, this could multiply and become a real
4380 problem.
4381
4382 2. Added /M option to pcretest to show the memory requirement of a specific
4383 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4384
4385 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4386 compiled in such a way that the backtracking after subsequent failure was
4387 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4388 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4389
4390
4391 Version 2.03 02-Feb-99
4392 ----------------------
4393
4394 1. Fixed typo and small mistake in man page.
4395
4396 2. Added 4th condition (GPL supersedes if conflict) and created separate
4397 LICENCE file containing the conditions.
4398
4399 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4400 Perl, that is the internal \ allows the delimiter to be included in the
4401 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4402 the final delimiter, add \ to the end of the pattern (to test the error).
4403
4404 4. Added the convenience functions for extracting substrings after a successful
4405 match. Updated pcretest to make it able to test these functions.
4406
4407
4408 Version 2.02 14-Jan-99
4409 ----------------------
4410
4411 1. Initialized the working variables associated with each extraction so that
4412 their saving and restoring doesn't refer to uninitialized store.
4413
4414 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4415 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4416 fix the problem.
4417
4418 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4419 calls, and wasn't printing the correct value for compiling calls. Increased the
4420 default value of LOOPREPEAT, and the number of significant figures in the
4421 times.
4422
4423 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4424
4425 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4426 a building problem on Windows NT with a FAT file system.
4427
4428
4429 Version 2.01 21-Oct-98
4430 ----------------------
4431
4432 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4433 to character tables built by pcre_maketables() in the current locale. If NULL
4434 is passed, the default tables are used.
4435
4436
4437 Version 2.00 24-Sep-98
4438 ----------------------
4439
4440 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4441 it any more.
4442
4443 2. Allow quantification of (?>) groups, and make it work correctly.
4444
4445 3. The first character computation wasn't working for (?>) groups.
4446
4447 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4448 end of the subject) and add 5.005's \z, which really does match only at the
4449 very end of the subject.
4450
4451 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4452
4453 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4454 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4455 localized options. All options to pcre_study() were also removed.
4456
4457 7. Add other new features from 5.005:
4458
4459 $(?<= positive lookbehind
4460 $(?<! negative lookbehind
4461 (?imsx-imsx) added the unsetting capability
4462 such a setting is global if at outer level; local otherwise
4463 (?imsx-imsx:) non-capturing groups with option setting
4464 (?(cond)re|re) conditional pattern matching
4465
4466 A backreference to itself in a repeated group matches the previous
4467 captured string.
4468
4469 8. General tidying up of studying (both automatic and via "study")
4470 consequential on the addition of new assertions.
4471
4472 9. As in 5.005, unlimited repeated groups that could match an empty substring
4473 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4474 runtime if any iteration does actually match an empty substring.
4475
4476 10. Include the RunTest script in the distribution.
4477
4478 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4479 discrepancies, some of which were old and were also with respect to 5.004. They
4480 have now been fixed.
4481
4482
4483 Version 1.09 28-Apr-98
4484 ----------------------
4485
4486 1. A negated single character class followed by a quantifier with a minimum
4487 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4488 program crashes, or just wrong answers. This did not apply to negated classes
4489 containing more than one character, or to minima other than one.
4490
4491
4492 Version 1.08 27-Mar-98
4493 ----------------------
4494
4495 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4496
4497 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4498 latter must appear before anything that relies on it in the pattern.
4499
4500
4501 Version 1.07 16-Feb-98
4502 ----------------------
4503
4504 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4505 repeat of a potentially empty string).
4506
4507
4508 Version 1.06 23-Jan-98
4509 ----------------------
4510
4511 1. Added Markus Oberhumer's little patches for C++.
4512
4513 2. Literal strings longer than 255 characters were broken.
4514
4515
4516 Version 1.05 23-Dec-97
4517 ----------------------
4518
4519 1. Negated character classes containing more than one character were failing if
4520 PCRE_CASELESS was set at run time.
4521
4522
4523 Version 1.04 19-Dec-97
4524 ----------------------
4525
4526 1. Corrected the man page, where some "const" qualifiers had been omitted.
4527
4528 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4529 input syntax.
4530
4531 3. Fixed memory leak which occurred when a regex with back references was
4532 matched with an offsets vector that wasn't big enough. The temporary memory
4533 that is used in this case wasn't being freed if the match failed.
4534
4535 4. Tidied pcretest to ensure it frees memory that it gets.
4536
4537 5. Temporary memory was being obtained in the case where the passed offsets
4538 vector was exactly big enough.
4539
4540 6. Corrected definition of offsetof() from change 5 below.
4541
4542 7. I had screwed up change 6 below and broken the rules for the use of
4543 setjmp(). Now fixed.
4544
4545
4546 Version 1.03 18-Dec-97
4547 ----------------------
4548
4549 1. A erroneous regex with a missing opening parenthesis was correctly
4550 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4551 on some systems.
4552
4553 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4554 it was reported that one broken compiler failed on the former because "code" is
4555 also an independent variable.
4556
4557 3. The erroneous regex a[]b caused an array overrun reference.
4558
4559 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4560 fail on data ending with that character. (It was going on too far, and checking
4561 the next character, typically a binary zero.) This was specific to the
4562 optimized code for single-character negative classes.
4563
4564 5. Added a contributed patch from the TIN world which does the following:
4565
4566 + Add an undef for memmove, in case the the system defines a macro for it.
4567
4568 + Add a definition of offsetof(), in case there isn't one. (I don't know
4569 the reason behind this - offsetof() is part of the ANSI standard - but
4570 it does no harm).
4571
4572 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4573 most of the places where whitespace preceded '#'. I have given up and
4574 allowed the remaining 2 cases to be at the margin.
4575
4576 + Rename some variables in pcre to eliminate shadowing. This seems very
4577 pedantic, but does no harm, of course.
4578
4579 6. Moved the call to setjmp() into its own function, to get rid of warnings
4580 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4581
4582 7. Constructs such as \d{8,} were compiling into the equivalent of
4583 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4584 outcome, but in this particular case used more store than had been allocated,
4585 which caused the bug to be discovered because it threw up an internal error.
4586
4587 8. The debugging code in both pcre and pcretest for outputting the compiled
4588 form of a regex was going wrong in the case of back references followed by
4589 curly-bracketed repeats.
4590
4591
4592 Version 1.02 12-Dec-97
4593 ----------------------
4594
4595 1. Typos in pcre.3 and comments in the source fixed.
4596
4597 2. Applied a contributed patch to get rid of places where it used to remove
4598 'const' from variables, and fixed some signed/unsigned and uninitialized
4599 variable warnings.
4600
4601 3. Added the "runtest" target to Makefile.
4602
4603 4. Set default compiler flag to -O2 rather than just -O.
4604
4605
4606 Version 1.01 19-Nov-97
4607 ----------------------
4608
4609 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4610 like /([ab]*)*/, that is, for classes with more than one character in them.
4611
4612 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4613 as /((?>a*))*/ (a PCRE_EXTRA facility).
4614
4615
4616 Version 1.00 18-Nov-97
4617 ----------------------
4618
4619 1. Added compile-time macros to support systems such as SunOS4 which don't have
4620 memmove() or strerror() but have other things that can be used instead.
4621
4622 2. Arranged that "make clean" removes the executables.
4623
4624
4625 Version 0.99 27-Oct-97
4626 ----------------------
4627
4628 1. Fixed bug in code for optimizing classes with only one character. It was
4629 initializing a 32-byte map regardless, which could cause it to run off the end
4630 of the memory it had got.
4631
4632 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4633
4634
4635 Version 0.98 22-Oct-97
4636 ----------------------
4637
4638 1. Fixed bug in code for handling temporary memory usage when there are more
4639 back references than supplied space in the ovector. This could cause segfaults.
4640
4641
4642 Version 0.97 21-Oct-97
4643 ----------------------
4644
4645 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4646
4647 2. Optimized negated single characters not to use a bit map.
4648
4649 3. Brought error texts together as macro definitions; clarified some of them;
4650 fixed one that was wrong - it said "range out of order" when it meant "invalid
4651 escape sequence".
4652
4653 4. Changed some char * arguments to const char *.
4654
4655 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4656
4657 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4658 pcretest.
4659
4660
4661 Version 0.96 16-Oct-97
4662 ----------------------
4663
4664 1. Added a simple "pgrep" utility to the distribution.
4665
4666 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4667 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4668 where "ddd" means "one or more decimal digits".
4669
4670 3. Fixed serious bug. If a pattern had a back reference, but the call to
4671 pcre_exec() didn't supply a large enough ovector to record the related
4672 identifying subpattern, the match always failed. PCRE now remembers the number
4673 of the largest back reference, and gets some temporary memory in which to save
4674 the offsets during matching if necessary, in order to ensure that
4675 backreferences always work.
4676
4677 4. Increased the compatibility with Perl in a number of ways:
4678
4679 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4680 to request this handling. The option can be set at compile or exec time.
4681
4682 (b) $ matches before a terminating newline by default; an option
4683 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4684 mode). The option can be set at compile or exec time.
4685
4686 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4687 the same as Perl's. If the decimal number it represents is less than 10
4688 or there aren't that many previous left capturing parentheses, an octal
4689 escape is read. Inside a character class, it's always an octal escape,
4690 even if it is a single digit.
4691
4692 (d) An escaped but undefined alphabetic character is taken as a literal,
4693 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4694 escapes.
4695
4696 (e) {0} is now permitted. (The previous item is removed from the compiled
4697 pattern).
4698
4699 5. Changed all the names of code files so that the basic parts are no longer
4700 than 10 characters, and abolished the teeny "globals.c" file.
4701
4702 6. Changed the handling of character classes; they are now done with a 32-byte
4703 bit map always.
4704
4705 7. Added the -d and /D options to pcretest to make it possible to look at the
4706 internals of compilation without having to recompile pcre.
4707
4708
4709 Version 0.95 23-Sep-97
4710 ----------------------
4711
4712 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4713 \x20 at the start of a run of normal characters. These were being treated as
4714 real characters, instead of the source characters being re-checked.
4715
4716
4717 Version 0.94 18-Sep-97
4718 ----------------------
4719
4720 1. The functions are now thread-safe, with the caveat that the global variables
4721 containing pointers to malloc() and free() or alternative functions are the
4722 same for all threads.
4723
4724 2. Get pcre_study() to generate a bitmap of initial characters for non-
4725 anchored patterns when this is possible, and use it if passed to pcre_exec().
4726
4727
4728 Version 0.93 15-Sep-97
4729 ----------------------
4730
4731 1. /(b)|(:+)/ was computing an incorrect first character.
4732
4733 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
4734 but not actually doing anything yet.
4735
4736 3. Treat "-" characters in classes that cannot be part of ranges as literals,
4737 as Perl does (e.g. [-az] or [az-]).
4738
4739 4. Set the anchored flag if a branch starts with .* or .*? because that tests
4740 all possible positions.
4741
4742 5. Split up into different modules to avoid including unneeded functions in a
4743 compiled binary. However, compile and exec are still in one module. The "study"
4744 function is split off.
4745
4746 6. The character tables are now in a separate module whose source is generated
4747 by an auxiliary program - but can then be edited by hand if required. There are
4748 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
4749 toupper() in the code.
4750
4751 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
4752 make them global. Abolish the function for setting them, as the caller can now
4753 set them directly.
4754
4755
4756 Version 0.92 11-Sep-97
4757 ----------------------
4758
4759 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
4760 (e.g. /a{1,3}/) was broken (I mis-optimized it).
4761
4762 2. Caseless matching was not working in character classes if the characters in
4763 the pattern were in upper case.
4764
4765 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
4766
4767 4. Make PCRE_ANCHORED public and accept as a compile option.
4768
4769 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4770 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4771 pass them.
4772
4773 6. Give an error if bad option bits passed at compile or run time.
4774
4775 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4776 pcretest to cause it to pass that flag.
4777
4778 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4779 options, and the first character, if set.
4780
4781 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4782
4783
4784 Version 0.91 10-Sep-97
4785 ----------------------
4786
4787 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4788 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4789
4790 2. PCRE was looping on encountering an indefinitely repeated back reference to
4791 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4792 Perl does - treats the match as successful.
4793
4794 ****

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12