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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1266 - (show annotations)
Sun Mar 3 11:14:26 2013 UTC (6 years, 8 months ago) by ph10
File size: 244210 byte(s)
Error occurred while calculating annotation data.
Allow callout before assertion condition in a conditional group.
1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.33 xx-xxxx-201x
5 -------------------------
6
7 1. Added 'U' to some constants that are compared to unsigned integers, to
8 avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
9 variables that are added to signed variables, to ensure the result is
10 signed and can be negated.
11
12 2. Applied patch by Daniel Richard G for quashing MSVC warnings to the
13 CMake config files.
14
15 3. Revise the creation of config.h.generic so that all boolean macros are
16 #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
17 overriding via -D on the command line possible.
18
19 4. Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
20 to unsigned int is reported to make a quite noticeable speed difference in
21 a specific Windows environment. Testing on Linux did also appear to show
22 some benefit (and it is clearly not harmful). Also fixed the definition of
23 Xop which should be unsigned.
24
25 5. Related to (4), changing the definition of the intermediate variable cc
26 in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
27 improvements.
28
29 6. Fix forward search in JIT when link size is 3 or greater. Also removed some
30 unnecessary spaces.
31
32 7. Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
33 and later.
34
35 8. Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
36
37 9. Optimizing fast_forward_start_bits in JIT.
38
39 10. Adding experimental support for callouts in JIT, and fixing some
40 issues revealed during this work. Namely:
41
42 (a) Unoptimized capturing brackets incorrectly reset on backtrack.
43
44 (b) Minimum length was not checked before the matching is started.
45
46 11. The value of capture_last that is passed to callouts was incorrect in some
47 cases when there was a capture on one path that was subsequently abandoned
48 after a backtrack. Also, the capture_last value is now reset after a
49 recursion, since all captures are also reset in this case.
50
51 12. The interpreter no longer returns the "too many substrings" error in the
52 case when an overflowing capture is in a branch that is subsequently
53 abandoned after a backtrack.
54
55 13. In the pathological case when an offset vector of size 2 is used, pcretest
56 now prints out the matched string after a yield of 0 or 1.
57
58 14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
59 Only supported by the JIT compiler at the moment.
60
61 15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
62
63 16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
64 offset of the starting point of the matching process, provided the offsets
65 vector is large enough.
66
67 17. The \A escape now records a lookbehind value of 1, though its execution
68 does not actually inspect the previous character. This is to ensure that,
69 in partial multi-segment matching, at least one character from the old
70 segment is retained when a new segment is processed. Otherwise, if there
71 are no lookbehinds in the pattern, \A might match incorrectly at the start
72 of a new segment.
73
74 18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
75
76 19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
77 gives some modest performance improvement in 8-bit mode.
78
79 20. Added the PCRE-specific property \p{Xuc} for matching characters that can
80 be expressed in certain programming languages using Universal Character
81 Names.
82
83 21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
84 which points out that "non characters" are not "characters that may not
85 appear in Unicode strings" but rather "characters that are reserved for
86 internal use and have only local meaning".
87
88 22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
89 there was a conditional group that depended on an assertion, if the
90 assertion was false, the callout that immediately followed the alternation
91 in the condition was skipped when pcre_exec() was used for matching.
92
93 23. Allow an explicit callout to be inserted before an assertion that is the
94 condition for a conditional group, for compatibility with automatic
95 callouts, which always insert a callout at this point.
96
97
98 Version 8.32 30-November-2012
99 -----------------------------
100
101 1. Improved JIT compiler optimizations for first character search and single
102 character iterators.
103
104 2. Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
105 Patch by Daniel Richard G.
106
107 3. Single character iterator optimizations in the JIT compiler.
108
109 4. Improved JIT compiler optimizations for character ranges.
110
111 5. Rename the "leave" variable names to "quit" to improve WinCE compatibility.
112 Reported by Giuseppe D'Angelo.
113
114 6. The PCRE_STARTLINE bit, indicating that a match can occur only at the start
115 of a line, was being set incorrectly in cases where .* appeared inside
116 atomic brackets at the start of a pattern, or where there was a subsequent
117 *PRUNE or *SKIP.
118
119 7. Improved instruction cache flush for POWER/PowerPC.
120 Patch by Daniel Richard G.
121
122 8. Fixed a number of issues in pcregrep, making it more compatible with GNU
123 grep:
124
125 (a) There is now no limit to the number of patterns to be matched.
126
127 (b) An error is given if a pattern is too long.
128
129 (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
130 are now supported.
131
132 (d) --exclude-from and --include-from (multiple use) have been added.
133
134 (e) Exclusions and inclusions now apply to all files and directories, not
135 just to those obtained from scanning a directory recursively.
136
137 (f) Multiple uses of -f and --file-list are now supported.
138
139 (g) In a Windows environment, the default for -d has been changed from
140 "read" (the GNU grep default) to "skip", because otherwise the presence
141 of a directory in the file list provokes an error.
142
143 (h) The documentation has been revised and clarified in places.
144
145 9. Improve the matching speed of capturing brackets.
146
147 10. Changed the meaning of \X so that it now matches a Unicode extended
148 grapheme cluster.
149
150 11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
151 out POSIX threads when JIT support is configured.
152
153 12. Added support for PCRE_STUDY_EXTRA_NEEDED.
154
155 13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
156 structure could go wrong in environments where size_t is not the same size
157 as int.
158
159 14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
160 set.
161
162 15. The EBCDIC support had decayed; later updates to the code had included
163 explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
164 general tidy up of EBCDIC-related issues, and the documentation was also
165 not quite right. There is now a test that can be run on ASCII systems to
166 check some of the EBCDIC-related things (but is it not a full test).
167
168 16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
169 in a small tidy to the code.
170
171 17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
172
173 18. If the --only-matching (-o) option in pcregrep is specified multiple
174 times, each one causes appropriate output. For example, -o1 -o2 outputs the
175 substrings matched by the 1st and 2nd capturing parentheses. A separating
176 string can be specified by --om-separator (default empty).
177
178 19. Improving the first n character searches.
179
180 20. Turn case lists for horizontal and vertical white space into macros so that
181 they are defined only once.
182
183 21. This set of changes together give more compatible Unicode case-folding
184 behaviour for characters that have more than one other case when UCP
185 support is available.
186
187 (a) The Unicode property table now has offsets into a new table of sets of
188 three or more characters that are case-equivalent. The MultiStage2.py
189 script that generates these tables (the pcre_ucd.c file) now scans
190 CaseFolding.txt instead of UnicodeData.txt for character case
191 information.
192
193 (b) The code for adding characters or ranges of characters to a character
194 class has been abstracted into a generalized function that also handles
195 case-independence. In UTF-mode with UCP support, this uses the new data
196 to handle characters with more than one other case.
197
198 (c) A bug that is fixed as a result of (b) is that codepoints less than 256
199 whose other case is greater than 256 are now correctly matched
200 caselessly. Previously, the high codepoint matched the low one, but not
201 vice versa.
202
203 (d) The processing of \h, \H, \v, and \ in character classes now makes use
204 of the new class addition function, using character lists defined as
205 macros alongside the case definitions of 20 above.
206
207 (e) Caseless back references now work with characters that have more than
208 one other case.
209
210 (f) General caseless matching of characters with more than one other case
211 is supported.
212
213 22. Unicode character properties were updated from Unicode 6.2.0
214
215 23. Improved CMake support under Windows. Patch by Daniel Richard G.
216
217 24. Add support for 32-bit character strings, and UTF-32
218
219 25. Major JIT compiler update (code refactoring and bugfixing).
220 Experimental Sparc 32 support is added.
221
222 26. Applied a modified version of Daniel Richard G's patch to create
223 pcre.h.generic and config.h.generic by "make" instead of in the
224 PrepareRelease script.
225
226 27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
227 pcre_compile.c when checking for a zero character.
228
229 28. Introducing a native interface for JIT. Through this interface, the compiled
230 machine code can be directly executed. The purpose of this interface is to
231 provide fast pattern matching, so several sanity checks are not performed.
232 However, feature tests are still performed. The new interface provides
233 1.4x speedup compared to the old one.
234
235 29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
236 the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
237 was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
238
239 30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
240 greater than 0x7fffffff (which cannot be represented in UTF-8, even under
241 the "old" RFC 2279). Instead, it ended up passing a negative length to
242 pcre_exec().
243
244 31. Add support for GCC's visibility feature to hide internal functions.
245
246 32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
247 "unknown -C option" after outputting 0 or 1.
248
249 33. There is now support for generating a code coverage report for the test
250 suite in environments where gcc is the compiler and lcov is installed. This
251 is mainly for the benefit of the developers.
252
253 34. If PCRE is built with --enable-valgrind, certain memory regions are marked
254 unaddressable using valgrind annotations, allowing valgrind to detect
255 invalid memory accesses. This is mainly for the benefit of the developers.
256
257 25. (*UTF) can now be used to start a pattern in any of the three libraries.
258
259 26. Give configure error if --enable-cpp but no C++ compiler found.
260
261
262 Version 8.31 06-July-2012
263 -------------------------
264
265 1. Fixing a wrong JIT test case and some compiler warnings.
266
267 2. Removed a bashism from the RunTest script.
268
269 3. Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
270 to unsigned type, result still unsigned" that was given by an MS compiler
271 on encountering the code "-sizeof(xxx)".
272
273 4. Partial matching support is added to the JIT compiler.
274
275 5. Fixed several bugs concerned with partial matching of items that consist
276 of more than one character:
277
278 (a) /^(..)\1/ did not partially match "aba" because checking references was
279 done on an "all or nothing" basis. This also applied to repeated
280 references.
281
282 (b) \R did not give a hard partial match if \r was found at the end of the
283 subject.
284
285 (c) \X did not give a hard partial match after matching one or more
286 characters at the end of the subject.
287
288 (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
289 a partial match for the string "\r".
290
291 (e) When newline was set to CRLF, the metacharacter "." did not recognize
292 a partial match for a CR character at the end of the subject string.
293
294 6. If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
295 running pcretest, the text "(JIT)" added to the output whenever JIT is
296 actually used to run the match.
297
298 7. Individual JIT compile options can be set in pcretest by following -s+[+]
299 or /S+[+] with a digit between 1 and 7.
300
301 8. OP_NOT now supports any UTF character not just single-byte ones.
302
303 9. (*MARK) control verb is now supported by the JIT compiler.
304
305 10. The command "./RunTest list" lists the available tests without actually
306 running any of them. (Because I keep forgetting what they all are.)
307
308 11. Add PCRE_INFO_MAXLOOKBEHIND.
309
310 12. Applied a (slightly modified) user-supplied patch that improves performance
311 when the heap is used for recursion (compiled with --disable-stack-for-
312 recursion). Instead of malloc and free for each heap frame each time a
313 logical recursion happens, frames are retained on a chain and re-used where
314 possible. This sometimes gives as much as 30% improvement.
315
316 13. As documented, (*COMMIT) is now confined to within a recursive subpattern
317 call.
318
319 14. As documented, (*COMMIT) is now confined to within a positive assertion.
320
321 15. It is now possible to link pcretest with libedit as an alternative to
322 libreadline.
323
324 16. (*COMMIT) control verb is now supported by the JIT compiler.
325
326 17. The Unicode data tables have been updated to Unicode 6.1.0.
327
328 18. Added --file-list option to pcregrep.
329
330 19. Added binary file support to pcregrep, including the -a, --binary-files,
331 -I, and --text options.
332
333 20. The madvise function is renamed for posix_madvise for QNX compatibility
334 reasons. Fixed by Giuseppe D'Angelo.
335
336 21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
337
338 22. Changed the option for creating symbolic links for 16-bit man pages from
339 -s to -sf so that re-installing does not cause issues.
340
341 23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
342
343 24. Fixed a very old bug in pcretest that caused errors with restarted DFA
344 matches in certain environments (the workspace was not being correctly
345 retained). Also added to pcre_dfa_exec() a simple plausibility check on
346 some of the workspace data at the beginning of a restart.
347
348 25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
349 was not doing so when it should - probably a typo introduced by SVN 528
350 (change 8.10/14).
351
352 26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
353 \w+ when the character tables indicated that \x{c4} was a word character.
354 There were several related cases, all because the tests for doing a table
355 lookup were testing for characters less than 127 instead of 255.
356
357 27. If a pattern contains capturing parentheses that are not used in a match,
358 their slots in the ovector are set to -1. For those that are higher than
359 any matched groups, this happens at the end of processing. In the case when
360 there were back references that the ovector was too small to contain
361 (causing temporary malloc'd memory to be used during matching), and the
362 highest capturing number was not used, memory off the end of the ovector
363 was incorrectly being set to -1. (It was using the size of the temporary
364 memory instead of the true size.)
365
366 28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
367 ovector size, it uses memory at the end of the block that it has got.
368
369 29. Check for an overlong MARK name and give an error at compile time. The
370 limit is 255 for the 8-bit library and 65535 for the 16-bit library.
371
372 30. JIT compiler update.
373
374 31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
375 Rill for the patch.
376
377 32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
378
379 33. Variable renamings in the PCRE-JIT compiler. No functionality change.
380
381 34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
382 SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
383 was enabled.
384
385 35. Improve JIT code generation for greedy plus quantifier.
386
387 36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
388 1 to "aa" instead of to an empty string. The bug affected repeated groups
389 that could potentially match an empty string.
390
391 37. Optimizing single character iterators in JIT.
392
393 38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
394 the same checks as \x{...} characters in non-JavaScript mode. Specifically,
395 codepoints that are too big for the mode are faulted, and in a UTF mode,
396 disallowed codepoints are also faulted.
397
398 39. If PCRE was compiled with UTF support, in three places in the DFA
399 matcher there was code that should only have been obeyed in UTF mode, but
400 was being obeyed unconditionally. In 8-bit mode this could cause incorrect
401 processing when bytes with values greater than 127 were present. In 16-bit
402 mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
403 both cases the values are those that cannot be the first data item in a UTF
404 character. The three items that might have provoked this were recursions,
405 possessively repeated groups, and atomic groups.
406
407 40. Ensure that libpcre is explicitly listed in the link commands for pcretest
408 and pcregrep, because some OS require shared objects to be explicitly
409 passed to ld, causing the link step to fail if they are not.
410
411 41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
412 mode, patterns that started with \h* or \R* might be incorrectly matched.
413
414
415 Version 8.30 04-February-2012
416 -----------------------------
417
418 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
419 name is defined in ctype.h.
420
421 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
422 only in quite long subpatterns.
423
424 3. Removed the function pcre_info(), which has been obsolete and deprecated
425 since it was replaced by pcre_fullinfo() in February 2000.
426
427 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
428 match a (*MARK), and the match failed at the start of the subject, a
429 reference to memory before the start of the subject could occur. This bug
430 was introduced by fix 17 of release 8.21.
431
432 5. A reference to an unset group with zero minimum repetition was giving
433 totally wrong answers (in non-JavaScript-compatibility mode). For example,
434 /(another)?(\1?)test/ matched against "hello world test". This bug was
435 introduced in release 8.13.
436
437 6. Add support for 16-bit character strings (a large amount of work involving
438 many changes and refactorings).
439
440 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
441 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
442 from a file.
443
444 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
445 rounding is not applied in this particular case).
446
447 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
448 if they appear, or are escaped, in patterns.
449
450 10. Get rid of a number of -Wunused-but-set-variable warnings.
451
452 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
453 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
454 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
455 also returns the mark "x". This bug applied to capturing parentheses,
456 non-capturing parentheses, and atomic parentheses. It also applied to some
457 assertions.
458
459 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
460 information out of configure.ac instead of relying on pcre.h.generic, which
461 is not stored in the repository.
462
463 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
464 -lreadline.
465
466 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
467
468 15. Applied Graycode's patch to put the top-level frame on the stack rather
469 than the heap when not using the stack for recursion. This gives a
470 performance improvement in many cases when recursion is not deep.
471
472 16. Experimental code added to "pcretest -C" to output the stack frame size.
473
474
475 Version 8.21 12-Dec-2011
476 ------------------------
477
478 1. Updating the JIT compiler.
479
480 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
481 are added as well.
482
483 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
484 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
485 calling _pcre_jit_exec. Some extra comments are added.
486
487 4. (*MARK) settings inside atomic groups that do not contain any capturing
488 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
489 was introduced by change 18 for 8.20.
490
491 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
492 ECMA-262 standard.
493
494 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
495 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
496 This bug was probably introduced by change 9 of 8.13.
497
498 7. While fixing 6 above, I noticed that a number of other items were being
499 incorrectly rejected as "not fixed length". This arose partly because newer
500 opcodes had not been added to the fixed-length checking code. I have (a)
501 corrected the bug and added tests for these items, and (b) arranged for an
502 error to occur if an unknown opcode is encountered while checking for fixed
503 length instead of just assuming "not fixed length". The items that were
504 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
505 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
506 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
507
508 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
509 being incorrectly compiled and would have given unpredicatble results.
510
511 9. A possessively repeated subpattern with minimum repeat count greater than
512 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
513 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
514 the first (A) could occur when it should not.
515
516 10. Add a cast and remove a redundant test from the code.
517
518 11. JIT should use pcre_malloc/pcre_free for allocation.
519
520 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
521 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
522 is anything other than /usr, -L is still shown).
523
524 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
525
526 14. Perl does not support \N without a following name in a [] class; PCRE now
527 also gives an error.
528
529 15. If a forward reference was repeated with an upper limit of around 2000,
530 it caused the error "internal error: overran compiling workspace". The
531 maximum number of forward references (including repeats) was limited by the
532 internal workspace, and dependent on the LINK_SIZE. The code has been
533 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
534 the default depends on LINK_SIZE. There is a new upper limit (for safety)
535 of around 200,000 forward references. While doing this, I also speeded up
536 the filling in of repeated forward references.
537
538 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
539 incorrectly expecting the subject to contain another "a" after the start.
540
541 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
542 in the match, the SKIP should be ignored. This was not happening; instead
543 the SKIP was being treated as NOMATCH. For patterns such as
544 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
545 tested.
546
547 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
548 now much more compatible with Perl, in particular in cases where the result
549 is a non-match for a non-anchored pattern. For example, if
550 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
551 "m", where previously it did not return a name. A side effect of this
552 change is that for partial matches, the last encountered mark name is
553 returned, as for non matches. A number of tests that were previously not
554 Perl-compatible have been moved into the Perl-compatible test files. The
555 refactoring has had the pleasing side effect of removing one argument from
556 the match() function, thus reducing its stack requirements.
557
558 19. If the /S+ option was used in pcretest to study a pattern using JIT,
559 subsequent uses of /S (without +) incorrectly behaved like /S+.
560
561 21. Retrieve executable code size support for the JIT compiler and fixing
562 some warnings.
563
564 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
565 not work when the shorter character appeared right at the end of the
566 subject string.
567
568 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
569 systems.
570
571 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
572 output it when the /M option is used in pcretest.
573
574 25. The CheckMan script was not being included in the distribution. Also, added
575 an explicit "perl" to run Perl scripts from the PrepareRelease script
576 because this is reportedly needed in Windows.
577
578 26. If study data was being save in a file and studying had not found a set of
579 "starts with" bytes for the pattern, the data written to the file (though
580 never used) was taken from uninitialized memory and so caused valgrind to
581 complain.
582
583 27. Updated RunTest.bat as provided by Sheri Pierce.
584
585 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
586
587 29. Computation of memory usage for the table of capturing group names was
588 giving an unnecessarily large value.
589
590
591 Version 8.20 21-Oct-2011
592 ------------------------
593
594 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
595 a POSIX class. After further experiments with Perl, which convinced me that
596 Perl has bugs and confusions, a closing square bracket is no longer allowed
597 in a POSIX name. This bug also affected patterns with classes that started
598 with full stops.
599
600 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
601 captured substring, but while checking the failing first alternative,
602 substring 1 is temporarily captured. If the output vector supplied to
603 pcre_exec() was not big enough for this capture, the yield of the function
604 was still zero ("insufficient space for captured substrings"). This cannot
605 be totally fixed without adding another stack variable, which seems a lot
606 of expense for a edge case. However, I have improved the situation in cases
607 such as /(a)(b)x|abc/ matched against "abc", where the return code
608 indicates that fewer than the maximum number of slots in the ovector have
609 been set.
610
611 3. Related to (2) above: when there are more back references in a pattern than
612 slots in the output vector, pcre_exec() uses temporary memory during
613 matching, and copies in the captures as far as possible afterwards. It was
614 using the entire output vector, but this conflicts with the specification
615 that only 2/3 is used for passing back captured substrings. Now it uses
616 only the first 2/3, for compatibility. This is, of course, another edge
617 case.
618
619 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
620 main code base, and can be used by building with --enable-jit. When this is
621 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
622 runtime --no-jit option is given.
623
624 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
625 ovector, the return from the function was zero, implying that there were
626 other matches that did not fit. The correct "exactly full" value is now
627 returned.
628
629 6. If a subpattern that was called recursively or as a subroutine contained
630 (*PRUNE) or any other control that caused it to give a non-standard return,
631 invalid errors such as "Error -26 (nested recursion at the same subject
632 position)" or even infinite loops could occur.
633
634 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
635 computing the minimum length on reaching *ACCEPT, and so ended up with the
636 wrong value of 1 rather than 0. Further investigation indicates that
637 computing a minimum subject length in the presence of *ACCEPT is difficult
638 (think back references, subroutine calls), and so I have changed the code
639 so that no minimum is registered for a pattern that contains *ACCEPT.
640
641 8. If (*THEN) was present in the first (true) branch of a conditional group,
642 it was not handled as intended. [But see 16 below.]
643
644 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
645 Sheri Pierce.
646
647 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
648 the first byte in a match must be "a".
649
650 11. Change 17 for 8.13 increased the recursion depth for patterns like
651 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
652 pattern contains any instances of (*THEN). If it does not, the old
653 optimizations are restored. It would be nice to do this on a per-group
654 basis, but at the moment that is not feasible.
655
656 12. In some environments, the output of pcretest -C is CRLF terminated. This
657 broke RunTest's code that checks for the link size. A single white space
658 character after the value is now allowed for.
659
660 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
661 For "fr", it uses the Windows-specific input and output files.
662
663 14. If (*THEN) appeared in a group that was called recursively or as a
664 subroutine, it did not work as intended. [But see next item.]
665
666 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
667 pattern fragments (but not containing any | characters). If A and B are
668 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
669 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
670 D. In other words, Perl considers parentheses that do not contain any |
671 characters to be part of a surrounding alternative, whereas PCRE was
672 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
673 differently. PCRE now behaves in the same way as Perl, except in the case
674 of subroutine/recursion calls such as (?1) which have in any case always
675 been different (but PCRE had them first :-).
676
677 16. Related to 15 above: Perl does not treat the | in a conditional group as
678 creating alternatives. Such a group is treated in the same way as an
679 ordinary group without any | characters when processing (*THEN). PCRE has
680 been changed to match Perl's behaviour.
681
682 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
683 RunGrepTest script failed.
684
685 18. Change 22 for version 13 caused atomic groups to use more stack. This is
686 inevitable for groups that contain captures, but it can lead to a lot of
687 stack use in large patterns. The old behaviour has been restored for atomic
688 groups that do not contain any capturing parentheses.
689
690 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
691 suppress the check for a minimum subject length at run time. (If it was
692 given to pcre_exec() or pcre_dfa_exec() it did work.)
693
694 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
695 fail to work when decoding hex characters in data strings in EBCDIC
696 environments.
697
698 21. It appears that in at least one Mac OS environment, the isxdigit() function
699 is implemented as a macro that evaluates to its argument more than once,
700 contravening the C 90 Standard (I haven't checked a later standard). There
701 was an instance in pcretest which caused it to go wrong when processing
702 \x{...} escapes in subject strings. The has been rewritten to avoid using
703 things like p++ in the argument of isxdigit().
704
705
706 Version 8.13 16-Aug-2011
707 ------------------------
708
709 1. The Unicode data tables have been updated to Unicode 6.0.0.
710
711 2. Two minor typos in pcre_internal.h have been fixed.
712
713 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
714 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
715 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
716
717 4. There were a number of related bugs in the code for matching backrefences
718 caselessly in UTF-8 mode when codes for the characters concerned were
719 different numbers of bytes. For example, U+023A and U+2C65 are an upper
720 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
721 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
722 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
723 2-byte code at the end of the subject (it thought there wasn't enough data
724 left).
725
726 5. Comprehensive information about what went wrong is now returned by
727 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
728 as the output vector has at least 2 elements. The offset of the start of
729 the failing character and a reason code are placed in the vector.
730
731 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
732 now returned is for the first byte of the failing character, instead of the
733 last byte inspected. This is an incompatible change, but I hope it is small
734 enough not to be a problem. It makes the returned offset consistent with
735 pcre_exec() and pcre_dfa_exec().
736
737 7. pcretest now gives a text phrase as well as the error number when
738 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
739 failure, the offset and reason code are output.
740
741 8. When \R was used with a maximizing quantifier it failed to skip backwards
742 over a \r\n pair if the subsequent match failed. Instead, it just skipped
743 back over a single character (\n). This seems wrong (because it treated the
744 two characters as a single entity when going forwards), conflicts with the
745 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
746 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
747 has been changed.
748
749 9. Some internal refactoring has changed the processing so that the handling
750 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
751 time (the PCRE_DOTALL option was changed this way some time ago: version
752 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
753 which was always a bit of a fudge. It also means that there is one less
754 argument for the match() function, which reduces its stack requirements
755 slightly. This change also fixes an incompatibility with Perl: the pattern
756 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
757
758 10. More internal refactoring has drastically reduced the number of recursive
759 calls to match() for possessively repeated groups such as (abc)++ when
760 using pcre_exec().
761
762 11. While implementing 10, a number of bugs in the handling of groups were
763 discovered and fixed:
764
765 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
766 (a|)*(?1) gave a compile-time internal error.
767 ((a|)+)+ did not notice that the outer group could match an empty string.
768 (^a|^)+ was not marked as anchored.
769 (.*a|.*)+ was not marked as matching at start or after a newline.
770
771 12. Yet more internal refactoring has removed another argument from the match()
772 function. Special calls to this function are now indicated by setting a
773 value in a variable in the "match data" data block.
774
775 13. Be more explicit in pcre_study() instead of relying on "default" for
776 opcodes that mean there is no starting character; this means that when new
777 ones are added and accidentally left out of pcre_study(), testing should
778 pick them up.
779
780 14. The -s option of pcretest has been documented for ages as being an old
781 synonym of -m (show memory usage). I have changed it to mean "force study
782 for every regex", that is, assume /S for every regex. This is similar to -i
783 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
784 using it. It makes it easier to run collections of tests with and without
785 study enabled, and thereby test pcre_study() more easily. All the standard
786 tests are now run with and without -s (but some patterns can be marked as
787 "never study" - see 20 below).
788
789 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
790 restoration of the capturing data to the outer values was not happening
791 correctly.
792
793 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
794 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
795 pattern had matched an empty string, and so incorrectly returned a no
796 match.
797
798 17. There was optimizing code for the last branch of non-capturing parentheses,
799 and also for the obeyed branch of a conditional subexpression, which used
800 tail recursion to cut down on stack usage. Unfortunately, now that there is
801 the possibility of (*THEN) occurring in these branches, tail recursion is
802 no longer possible because the return has to be checked for (*THEN). These
803 two optimizations have therefore been removed. [But see 8.20/11 above.]
804
805 18. If a pattern containing \R was studied, it was assumed that \R always
806 matched two bytes, thus causing the minimum subject length to be
807 incorrectly computed because \R can also match just one byte.
808
809 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
810 was incorrectly computed.
811
812 20. If /S is present twice on a test pattern in pcretest input, it now
813 *disables* studying, thereby overriding the use of -s on the command line
814 (see 14 above). This is necessary for one or two tests to keep the output
815 identical in both cases.
816
817 21. When (*ACCEPT) was used in an assertion that matched an empty string and
818 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
819
820 22. When an atomic group that contained a capturing parenthesis was
821 successfully matched, but the branch in which it appeared failed, the
822 capturing was not being forgotten if a higher numbered group was later
823 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
824 group 1 to "a", when in fact it should be unset. This applied to multi-
825 branched capturing and non-capturing groups, repeated or not, and also to
826 positive assertions (capturing in negative assertions does not happen
827 in PCRE) and also to nested atomic groups.
828
829 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
830 subject after a captured substring, to make it easier to tell which of a
831 number of identical substrings has been captured.
832
833 24. The way atomic groups are processed by pcre_exec() has been changed so that
834 if they are repeated, backtracking one repetition now resets captured
835 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
836 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
837 "aaa". Previously, it would have been "a". As part of this code
838 refactoring, the way recursive calls are handled has also been changed.
839
840 25. If an assertion condition captured any substrings, they were not passed
841 back unless some other capturing happened later. For example, if
842 (?(?=(a))a) was matched against "a", no capturing was returned.
843
844 26. When studying a pattern that contained subroutine calls or assertions,
845 the code for finding the minimum length of a possible match was handling
846 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
847 group 1 called group 2 while simultaneously a separate group 2 called group
848 1). A stack overflow occurred in this case. I have fixed this by limiting
849 the recursion depth to 10.
850
851 27. Updated RunTest.bat in the distribution to the version supplied by Tom
852 Fortmann. This supports explicit test numbers on the command line, and has
853 argument validation and error reporting.
854
855 28. An instance of \X with an unlimited repeat could fail if at any point the
856 first character it looked at was a mark character.
857
858 29. Some minor code refactoring concerning Unicode properties and scripts
859 should reduce the stack requirement of match() slightly.
860
861 30. Added the '=' option to pcretest to check the setting of unused capturing
862 slots at the end of the pattern, which are documented as being -1, but are
863 not included in the return count.
864
865 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
866 compiled something random. Now it gives a compile-time error (as does
867 Perl).
868
869 32. A *MARK encountered during the processing of a positive assertion is now
870 recorded and passed back (compatible with Perl).
871
872 33. If --only-matching or --colour was set on a pcregrep call whose pattern
873 had alternative anchored branches, the search for a second match in a line
874 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
875 matched the line "0102" twice. The same bug affected patterns that started
876 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
877 twice.
878
879 34. Previously, PCRE did not allow quantification of assertions. However, Perl
880 does, and because of capturing effects, quantifying parenthesized
881 assertions may at times be useful. Quantifiers are now allowed for
882 parenthesized assertions.
883
884 35. A minor code tidy in pcre_compile() when checking options for \R usage.
885
886 36. \g was being checked for fancy things in a character class, when it should
887 just be a literal "g".
888
889 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
890 appearance of a nested POSIX class supersedes an apparent external class.
891 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
892 unescaped square brackets may also appear as part of class names. For
893 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
894 more like Perl. (But see 8.20/1 above.)
895
896 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
897 was because it thought it was \N{name}, which is not supported).
898
899 39. Add minix to OS list not supporting the -S option in pcretest.
900
901 40. PCRE tries to detect cases of infinite recursion at compile time, but it
902 cannot analyze patterns in sufficient detail to catch mutual recursions
903 such as ((?1))((?2)). There is now a runtime test that gives an error if a
904 subgroup is called recursively as a subpattern for a second time at the
905 same position in the subject string. In previous releases this might have
906 been caught by the recursion limit, or it might have run out of stack.
907
908 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
909 happen only once. PCRE was, however incorrectly giving a compile time error
910 "recursive call could loop indefinitely" because it cannot analyze the
911 pattern in sufficient detail. The compile time test no longer happens when
912 PCRE is compiling a conditional subpattern, but actual runaway loops are
913 now caught at runtime (see 40 above).
914
915 42. It seems that Perl allows any characters other than a closing parenthesis
916 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
917 has been changed to be the same.
918
919 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
920 as not to get warnings when autogen.sh is called. Also changed
921 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
922
923 44. To help people who use pcregrep to scan files containing exceedingly long
924 lines, the following changes have been made:
925
926 (a) The default value of the buffer size parameter has been increased from
927 8K to 20K. (The actual buffer used is three times this size.)
928
929 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
930 PCRE is built.
931
932 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
933 to be set at run time.
934
935 (d) Numerical values in pcregrep options can be followed by K or M, for
936 example --buffer-size=50K.
937
938 (e) If a line being scanned overflows pcregrep's buffer, an error is now
939 given and the return code is set to 2.
940
941 45. Add a pointer to the latest mark to the callout data block.
942
943 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
944 partial match of an empty string instead of no match. This was specific to
945 the use of ".".
946
947 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
948 complete match instead of a partial match. This bug was dependent on both
949 the PCRE_UTF8 and PCRE_DOTALL options being set.
950
951 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
952 starting byte set, because \b was not being ignored.
953
954
955 Version 8.12 15-Jan-2011
956 ------------------------
957
958 1. Fixed some typos in the markup of the man pages, and wrote a script that
959 checks for such things as part of the documentation building process.
960
961 2. On a big-endian 64-bit system, pcregrep did not correctly process the
962 --match-limit and --recursion-limit options (added for 8.11). In
963 particular, this made one of the standard tests fail. (The integer value
964 went into the wrong half of a long int.)
965
966 3. If the --colour option was given to pcregrep with -v (invert match), it
967 did strange things, either producing crazy output, or crashing. It should,
968 of course, ignore a request for colour when reporting lines that do not
969 match.
970
971 4. Another pcregrep bug caused similar problems if --colour was specified with
972 -M (multiline) and the pattern match finished with a line ending.
973
974 5. In pcregrep, when a pattern that ended with a literal newline sequence was
975 matched in multiline mode, the following line was shown as part of the
976 match. This seems wrong, so I have changed it.
977
978 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
979 the check for further matches in the same line (so they could be coloured)
980 to overrun the end of the current line. If another match was found, it was
981 incorrectly shown (and then shown again when found in the next line).
982
983 7. If pcregrep was compiled under Windows, there was a reference to the
984 function pcregrep_exit() before it was defined. I am assuming this was
985 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
986 reported by a user. I've moved the definition above the reference.
987
988
989 Version 8.11 10-Dec-2010
990 ------------------------
991
992 1. (*THEN) was not working properly if there were untried alternatives prior
993 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
994 backtracked to try for "b" instead of moving to the next alternative branch
995 at the same level (in this case, to look for "c"). The Perl documentation
996 is clear that when (*THEN) is backtracked onto, it goes to the "next
997 alternative in the innermost enclosing group".
998
999 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1000 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
1001 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1002 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1003 (*THEN).
1004
1005 3. If \s appeared in a character class, it removed the VT character from
1006 the class, even if it had been included by some previous item, for example
1007 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1008 of \s, but is part of the POSIX "space" class.)
1009
1010 4. A partial match never returns an empty string (because you can always
1011 match an empty string at the end of the subject); however the checking for
1012 an empty string was starting at the "start of match" point. This has been
1013 changed to the "earliest inspected character" point, because the returned
1014 data for a partial match starts at this character. This means that, for
1015 example, /(?<=abc)def/ gives a partial match for the subject "abc"
1016 (previously it gave "no match").
1017
1018 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1019 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1020 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1021 has an implication that the given string is incomplete (because a partial
1022 match is preferred over a full match). For this reason, these items now
1023 give a partial match in this situation. [Aside: previously, the one case
1024 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1025 match rather than a full match, which was wrong by the old rules, but is
1026 now correct.]
1027
1028 6. There was a bug in the handling of #-introduced comments, recognized when
1029 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1030 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1031 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1032 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1033 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1034 places in pcre_compile().
1035
1036 7. Related to (6) above, when pcre_compile() was skipping #-introduced
1037 comments when looking ahead for named forward references to subpatterns,
1038 the only newline sequence it recognized was NL. It now handles newlines
1039 according to the set newline convention.
1040
1041 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1042 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1043 cater for a lack of strerror(). These oversights have been fixed.
1044
1045 9. Added --match-limit and --recursion-limit to pcregrep.
1046
1047 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1048
1049 11. When the -o option was used, pcregrep was setting a return code of 1, even
1050 when matches were found, and --line-buffered was not being honoured.
1051
1052 12. Added an optional parentheses number to the -o and --only-matching options
1053 of pcregrep.
1054
1055 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1056 can match an empty string. The code to do it in pcretest and pcredemo
1057 needed fixing:
1058
1059 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1060 only one byte after an empty string match just before CRLF (this case
1061 just got forgotten; "any" and "anycrlf" were OK).
1062
1063 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1064 mode when an empty string match preceded an ASCII character followed by
1065 a non-ASCII character. (The code for advancing by one character rather
1066 than one byte was nonsense.)
1067
1068 (c) The pcredemo.c sample program did not have any code at all to handle
1069 the cases when CRLF is a valid newline sequence.
1070
1071 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1072 as a starting offset was within the subject string. There is now a new
1073 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1074 negative or greater than the length of the string. In order to test this,
1075 pcretest is extended to allow the setting of negative starting offsets.
1076
1077 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1078 starting offset points to the beginning of a UTF-8 character was
1079 unnecessarily clumsy. I tidied it up.
1080
1081 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1082 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1083
1084 17. Nobody had reported that the --include_dir option, which was added in
1085 release 7.7 should have been called --include-dir (hyphen, not underscore)
1086 for compatibility with GNU grep. I have changed it to --include-dir, but
1087 left --include_dir as an undocumented synonym, and the same for
1088 --exclude-dir, though that is not available in GNU grep, at least as of
1089 release 2.5.4.
1090
1091 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1092 characters from a string of bytes) have been redefined so as not to use
1093 loops, in order to improve performance in some environments. At the same
1094 time, I abstracted some of the common code into auxiliary macros to save
1095 repetition (this should not affect the compiled code).
1096
1097 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1098 compile-time error is now given if \c is not followed by an ASCII
1099 character, that is, a byte less than 128. (In EBCDIC mode, the code is
1100 different, and any byte value is allowed.)
1101
1102 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1103 START_OPTIMIZE option, which is now allowed at compile time - but just
1104 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1105 to pcregrep and other applications that have no direct access to PCRE
1106 options. The new /Y option in pcretest sets this option when calling
1107 pcre_compile().
1108
1109 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1110 back references. Groups containing recursive back references were forced to
1111 be atomic by that change, but in the case of named groups, the amount of
1112 memory required was incorrectly computed, leading to "Failed: internal
1113 error: code overflow". This has been fixed.
1114
1115 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1116 pcretest.c, to avoid build problems in some Borland environments.
1117
1118
1119 Version 8.10 25-Jun-2010
1120 ------------------------
1121
1122 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1123 THEN.
1124
1125 2. (*ACCEPT) was not working when inside an atomic group.
1126
1127 3. Inside a character class, \B is treated as a literal by default, but
1128 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1129 causes the error). The code is unchanged, but I tidied the documentation.
1130
1131 4. Inside a character class, PCRE always treated \R and \X as literals,
1132 whereas Perl faults them if its -w option is set. I have changed PCRE so
1133 that it faults them when PCRE_EXTRA is set.
1134
1135 5. Added support for \N, which always matches any character other than
1136 newline. (It is the same as "." when PCRE_DOTALL is not set.)
1137
1138 6. When compiling pcregrep with newer versions of gcc which may have
1139 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1140 declared with attribute warn_unused_result" were given. Just casting the
1141 result to (void) does not stop the warnings; a more elaborate fudge is
1142 needed. I've used a macro to implement this.
1143
1144 7. Minor change to pcretest.c to avoid a compiler warning.
1145
1146 8. Added four artifical Unicode properties to help with an option to make
1147 \s etc use properties (see next item). The new properties are: Xan
1148 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1149
1150 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1151 use Unicode properties. (*UCP) at the start of a pattern can be used to set
1152 this option. Modified pcretest to add /W to test this facility. Added
1153 REG_UCP to make it available via the POSIX interface.
1154
1155 10. Added --line-buffered to pcregrep.
1156
1157 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1158 studied, and the match started with a letter with a code point greater than
1159 127 whose first byte was different to the first byte of the other case of
1160 the letter, the other case of this starting letter was not recognized
1161 (#976).
1162
1163 12. If a pattern that was studied started with a repeated Unicode property
1164 test, for example, \p{Nd}+, there was the theoretical possibility of
1165 setting up an incorrect bitmap of starting bytes, but fortunately it could
1166 not have actually happened in practice until change 8 above was made (it
1167 added property types that matched character-matching opcodes).
1168
1169 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1170 possible starting bytes for non-anchored patterns.
1171
1172 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1173 \R, and also a number of cases that involve Unicode properties, both
1174 explicit and implicit when PCRE_UCP is set.
1175
1176 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1177 input, it could crash or give wrong results if characters with values
1178 greater than 0xc0 were present in the subject string. (Detail: it assumed
1179 UTF-8 input when processing these items.)
1180
1181 16. Added a lot of (int) casts to avoid compiler warnings in systems where
1182 size_t is 64-bit (#991).
1183
1184 17. Added a check for running out of memory when PCRE is compiled with
1185 --disable-stack-for-recursion (#990).
1186
1187 18. If the last data line in a file for pcretest does not have a newline on
1188 the end, a newline was missing in the output.
1189
1190 19. The default pcre_chartables.c file recognizes only ASCII characters (values
1191 less than 128) in its various bitmaps. However, there is a facility for
1192 generating tables according to the current locale when PCRE is compiled. It
1193 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1194 characters, are recognized by isspace() and therefore were getting set in
1195 these tables, and indeed these tables seem to approximate to ISO 8859. This
1196 caused a problem in UTF-8 mode when pcre_study() was used to create a list
1197 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1198 which of course cannot start UTF-8 characters. I have changed the code so
1199 that only real ASCII characters (less than 128) and the correct starting
1200 bytes for UTF-8 encodings are set for characters greater than 127 when in
1201 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1202 altogether.)
1203
1204 20. Added the /T option to pcretest so as to be able to run tests with non-
1205 standard character tables, thus making it possible to include the tests
1206 used for 19 above in the standard set of tests.
1207
1208 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1209 reference to a subpattern the other side of a comment that contains an
1210 opening parenthesis caused either an internal compiling error, or a
1211 reference to the wrong subpattern.
1212
1213
1214 Version 8.02 19-Mar-2010
1215 ------------------------
1216
1217 1. The Unicode data tables have been updated to Unicode 5.2.0.
1218
1219 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
1220 configured.
1221
1222 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
1223 original author of that file, following a query about its status.
1224
1225 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
1226 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1227
1228 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1229 quantifier applied to a forward-referencing subroutine call, could compile
1230 incorrect code or give the error "internal error: previously-checked
1231 referenced subpattern not found".
1232
1233 6. Both MS Visual Studio and Symbian OS have problems with initializing
1234 variables to point to external functions. For these systems, therefore,
1235 pcre_malloc etc. are now initialized to local functions that call the
1236 relevant global functions.
1237
1238 7. There were two entries missing in the vectors called coptable and poptable
1239 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1240 I've fixed the data, and added a kludgy way of testing at compile time that
1241 the lengths are correct (equal to the number of opcodes).
1242
1243 8. Following on from 7, I added a similar kludge to check the length of the
1244 eint vector in pcreposix.c.
1245
1246 9. Error texts for pcre_compile() are held as one long string to avoid too
1247 much relocation at load time. To find a text, the string is searched,
1248 counting zeros. There was no check for running off the end of the string,
1249 which could happen if a new error number was added without updating the
1250 string.
1251
1252 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1253
1254 11. \K was not working if it appeared in an atomic group or in a group that
1255 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1256 \K is "not well defined" if used in an assertion. PCRE now accepts it if
1257 the assertion is positive, but not if it is negative.
1258
1259 12. Change 11 fortuitously reduced the size of the stack frame used in the
1260 "match()" function of pcre_exec.c by one pointer. Forthcoming
1261 implementation of support for (*MARK) will need an extra pointer on the
1262 stack; I have reserved it now, so that the stack frame size does not
1263 decrease.
1264
1265 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1266 item in branch that calls a recursion is a subroutine call - as in the
1267 second branch in the above example - was incorrectly given the compile-
1268 time error "recursive call could loop indefinitely" because pcre_compile()
1269 was not correctly checking the subroutine for matching a non-empty string.
1270
1271 14. The checks for overrunning compiling workspace could trigger after an
1272 overrun had occurred. This is a "should never occur" error, but it can be
1273 triggered by pathological patterns such as hundreds of nested parentheses.
1274 The checks now trigger 100 bytes before the end of the workspace.
1275
1276 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1277
1278
1279 Version 8.01 19-Jan-2010
1280 ------------------------
1281
1282 1. If a pattern contained a conditional subpattern with only one branch (in
1283 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1284 computed the wrong minimum data length (which is of course zero for such
1285 subpatterns). This could cause incorrect "no match" results.
1286
1287 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1288 the pattern is reset in the first branch, pcre_compile() failed with
1289 "internal error: code overflow at offset...". This happened only when
1290 the reset was to the original external option setting. (An optimization
1291 abstracts leading options settings into an external setting, which was the
1292 cause of this.)
1293
1294 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1295 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1296 assertion pattern did not match (meaning that the assertion was true), it
1297 was incorrectly treated as false if the SKIP had been reached during the
1298 matching. This also applied to assertions used as conditions.
1299
1300 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
1301 assertion subpattern, including such a pattern used as a condition,
1302 unpredictable results occurred, instead of the error return
1303 PCRE_ERROR_DFA_UITEM.
1304
1305 5. The C++ GlobalReplace function was not working like Perl for the special
1306 situation when an empty string is matched. It now does the fancy magic
1307 stuff that is necessary.
1308
1309 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1310 removed. (These were left over from very, very early versions of PCRE.)
1311
1312 7. Some cosmetic changes to the code to make life easier when compiling it
1313 as part of something else:
1314
1315 (a) Change DEBUG to PCRE_DEBUG.
1316
1317 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1318 called "current" as "current_branch", to prevent a collision with the
1319 Linux macro when compiled as a kernel module.
1320
1321 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1322 prevent a collision with the Linux macro when compiled as a kernel
1323 module.
1324
1325 8. In pcre_compile() there are some checks for integer overflows that used to
1326 cast potentially large values to (double). This has been changed to that
1327 when building, a check for int64_t is made, and if it is found, it is used
1328 instead, thus avoiding the use of floating point arithmetic. (There is no
1329 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1330 double.
1331
1332 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1333 2005 (difference between two addresses compared to an unsigned value).
1334
1335 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1336 custom one, because of the following reported problem in Windows:
1337
1338 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1339 under Win32.
1340 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1341 therefore missing the function definition.
1342 - The compiler thus generates a "C" signature for the test function.
1343 - The linker fails to find the "C" function.
1344 - PCRE fails to configure if asked to do so against libbz2.
1345
1346 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1347 messages were output:
1348
1349 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1350 rerunning libtoolize, to keep the correct libtool macros in-tree.
1351 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1352
1353 I have done both of these things.
1354
1355 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1356 most of the time, it *can* run out if it is given a pattern that contains a
1357 runaway infinite recursion. I updated the discussion in the pcrestack man
1358 page.
1359
1360 13. Now that we have gone to the x.xx style of version numbers, the minor
1361 version may start with zero. Using 08 or 09 is a bad idea because users
1362 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1363 interpreted as invalid octal numbers. I've updated the previous comment in
1364 configure.ac, and also added a check that gives an error if 08 or 09 are
1365 used.
1366
1367 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1368 causing partial matching to fail when the end of the subject matched \W
1369 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1370
1371 15. There were some discrepancies between the declarations in pcre_internal.h
1372 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1373 their definitions. The declarations used "const uschar *" and the
1374 definitions used USPTR. Even though USPTR is normally defined as "const
1375 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1376 reported that: "This difference in casting confuses some C++ compilers, for
1377 example, SunCC recognizes above declarations as different functions and
1378 generates broken code for hbpcre." I have changed the declarations to use
1379 USPTR.
1380
1381 16. GNU libtool is named differently on some systems. The autogen.sh script now
1382 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1383 (FreeBSD).
1384
1385 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1386 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1387 comment: "Figure out how to create a longlong from a string: strtoll and
1388 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1389 instance, but it only takes 2 args instead of 3!"
1390
1391 18. A subtle bug concerned with back references has been fixed by a change of
1392 specification, with a corresponding code fix. A pattern such as
1393 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1394 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1395 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1396 same bug. Such groups have to be quantified to be useful, or contained
1397 inside another quantified group. (If there's no repetition, the reference
1398 can never match.) The problem arises because, having left the group and
1399 moved on to the rest of the pattern, a later failure that backtracks into
1400 the group uses the captured value from the final iteration of the group
1401 rather than the correct earlier one. I have fixed this in PCRE by forcing
1402 any group that contains a reference to itself to be an atomic group; that
1403 is, there cannot be any backtracking into it once it has completed. This is
1404 similar to recursive and subroutine calls.
1405
1406
1407 Version 8.00 19-Oct-09
1408 ----------------------
1409
1410 1. The table for translating pcre_compile() error codes into POSIX error codes
1411 was out-of-date, and there was no check on the pcre_compile() error code
1412 being within the table. This could lead to an OK return being given in
1413 error.
1414
1415 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1416 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1417 in a Windows environment.
1418
1419 3. The pcregrep --count option prints the count for each file even when it is
1420 zero, as does GNU grep. However, pcregrep was also printing all files when
1421 --files-with-matches was added. Now, when both options are given, it prints
1422 counts only for those files that have at least one match. (GNU grep just
1423 prints the file name in this circumstance, but including the count seems
1424 more useful - otherwise, why use --count?) Also ensured that the
1425 combination -clh just lists non-zero counts, with no names.
1426
1427 4. The long form of the pcregrep -F option was incorrectly implemented as
1428 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1429 but it seems right to fix it, and I didn't think it was worth preserving
1430 the old behaviour.
1431
1432 5. The command line items --regex=pattern and --regexp=pattern were not
1433 recognized by pcregrep, which required --regex pattern or --regexp pattern
1434 (with a space rather than an '='). The man page documented the '=' forms,
1435 which are compatible with GNU grep; these now work.
1436
1437 6. No libpcreposix.pc file was created for pkg-config; there was just
1438 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1439
1440 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1441 when UCP support is not needed, by modifying the Python script that
1442 generates it from Unicode data files. This should not matter if the module
1443 is correctly used as a library, but I received one complaint about 50K of
1444 unwanted data. My guess is that the person linked everything into his
1445 program rather than using a library. Anyway, it does no harm.
1446
1447 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1448 was a minimum greater than 1 for a wide character in a possessive
1449 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1450 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1451 character. Chaos in the form of incorrect output or a compiling loop could
1452 result.
1453
1454 9. The restrictions on what a pattern can contain when partial matching is
1455 requested for pcre_exec() have been removed. All patterns can now be
1456 partially matched by this function. In addition, if there are at least two
1457 slots in the offset vector, the offset of the earliest inspected character
1458 for the match and the offset of the end of the subject are set in them when
1459 PCRE_ERROR_PARTIAL is returned.
1460
1461 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1462 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1463 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1464 and may be more useful for multi-segment matching.
1465
1466 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1467 used to be given if ever the end of the subject was reached; now it is
1468 given only if matching could not proceed because another character was
1469 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1470 string "Z", which now yields "no match" instead of "partial match". In the
1471 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1472 final character ended with (*FAIL).
1473
1474 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1475 if the pattern had a "must contain" character that was already found in the
1476 earlier partial match, unless partial matching was again requested. For
1477 example, with the pattern /dog.(body)?/, the "must contain" character is
1478 "g". If the first part-match was for the string "dog", restarting with
1479 "sbody" failed. This bug has been fixed.
1480
1481 13. The string returned by pcre_dfa_exec() after a partial match has been
1482 changed so that it starts at the first inspected character rather than the
1483 first character of the match. This makes a difference only if the pattern
1484 starts with a lookbehind assertion or \b or \B (\K is not supported by
1485 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1486 matching functions compatible, and I think it's the right thing to do.
1487
1488 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1489 so that the demonstration program is easily available in environments where
1490 PCRE has not been installed from source.
1491
1492 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1493 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1494 library.
1495
1496 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1497 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1498 is not the first non-POSIX option to be added. Clearly some people find
1499 these options useful.
1500
1501 17. If a caller to the POSIX matching function regexec() passes a non-zero
1502 value for nmatch with a NULL value for pmatch, the value of
1503 nmatch is forced to zero.
1504
1505 18. RunGrepTest did not have a test for the availability of the -u option of
1506 the diff command, as RunTest does. It now checks in the same way as
1507 RunTest, and also checks for the -b option.
1508
1509 19. If an odd number of negated classes containing just a single character
1510 interposed, within parentheses, between a forward reference to a named
1511 subpattern and the definition of the subpattern, compilation crashed with
1512 an internal error, complaining that it could not find the referenced
1513 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1514 [The bug was that it was starting one character too far in when skipping
1515 over the character class, thus treating the ] as data rather than
1516 terminating the class. This meant it could skip too much.]
1517
1518 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1519 /g option in pcretest when the pattern contains \K, which makes it possible
1520 to have an empty string match not at the start, even when the pattern is
1521 anchored. Updated pcretest and pcredemo to use this option.
1522
1523 21. If the maximum number of capturing subpatterns in a recursion was greater
1524 than the maximum at the outer level, the higher number was returned, but
1525 with unset values at the outer level. The correct (outer level) value is
1526 now given.
1527
1528 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1529 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1530 make it do so. The string so far is captured, making this feature
1531 compatible with Perl.
1532
1533 23. The tests have been re-organized, adding tests 11 and 12, to make it
1534 possible to check the Perl 5.10 features against Perl 5.10.
1535
1536 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1537 pattern matches a fixed length string. PCRE did not allow this; now it
1538 does. Neither allows recursion.
1539
1540 25. I finally figured out how to implement a request to provide the minimum
1541 length of subject string that was needed in order to match a given pattern.
1542 (It was back references and recursion that I had previously got hung up
1543 on.) This code has now been added to pcre_study(); it finds a lower bound
1544 to the length of subject needed. It is not necessarily the greatest lower
1545 bound, but using it to avoid searching strings that are too short does give
1546 some useful speed-ups. The value is available to calling programs via
1547 pcre_fullinfo().
1548
1549 26. While implementing 25, I discovered to my embarrassment that pcretest had
1550 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1551 study optimizations had never been tested with that matching function.
1552 Oops. What is worse, even when it was passed study data, there was a bug in
1553 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1554 were also very few tests of studied patterns with pcre_dfa_exec().
1555
1556 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1557 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1558 on the other side of the coin, they are no longer allowed to have different
1559 names, because these cannot be distinguished in PCRE, and this has caused
1560 confusion. (This is a difference from Perl.)
1561
1562 28. When duplicate subpattern names are present (necessarily with different
1563 numbers, as required by 27 above), and a test is made by name in a
1564 conditional pattern, either for a subpattern having been matched, or for
1565 recursion in such a pattern, all the associated numbered subpatterns are
1566 tested, and the overall condition is true if the condition is true for any
1567 one of them. This is the way Perl works, and is also more like the way
1568 testing by number works.
1569
1570
1571 Version 7.9 11-Apr-09
1572 ---------------------
1573
1574 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1575 (pcretest), all targets were linked against these libraries. This included
1576 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1577 libraries. This caused unwanted dependencies to be created. This problem
1578 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1579 pcretest is linked with readline.
1580
1581 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1582 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1583 moved outside it again, because FALSE and TRUE are already defined in AIX,
1584 but BOOL is not.
1585
1586 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1587 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1588
1589 4. The pcregrep documentation said spaces were inserted as well as colons (or
1590 hyphens) following file names and line numbers when outputting matching
1591 lines. This is not true; no spaces are inserted. I have also clarified the
1592 wording for the --colour (or --color) option.
1593
1594 5. In pcregrep, when --colour was used with -o, the list of matching strings
1595 was not coloured; this is different to GNU grep, so I have changed it to be
1596 the same.
1597
1598 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1599 each matching line was coloured. Now it goes on to look for further matches
1600 of any of the test patterns, which is the same behaviour as GNU grep.
1601
1602 7. A pattern that could match an empty string could cause pcregrep to loop; it
1603 doesn't make sense to accept an empty string match in pcregrep, so I have
1604 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1605 seems to be how GNU grep behaves.
1606
1607 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1608 start or after a newline", because the conditional assertion was not being
1609 correctly handled. The rule now is that both the assertion and what follows
1610 in the first alternative must satisfy the test.
1611
1612 9. If auto-callout was enabled in a pattern with a conditional group whose
1613 condition was an assertion, PCRE could crash during matching, both with
1614 pcre_exec() and pcre_dfa_exec().
1615
1616 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1617 used for matching.
1618
1619 11. Unicode property support in character classes was not working for
1620 characters (bytes) greater than 127 when not in UTF-8 mode.
1621
1622 12. Added the -M command line option to pcretest.
1623
1624 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1625
1626 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1627
1628 16. Added comments and documentation about mis-use of no_arg in the C++
1629 wrapper.
1630
1631 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1632 from Martin Jerabek that uses macro names for all relevant character and
1633 string constants.
1634
1635 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1636 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1637 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1638 these, but not everybody uses configure.
1639
1640 19. A conditional group that had only one branch was not being correctly
1641 recognized as an item that could match an empty string. This meant that an
1642 enclosing group might also not be so recognized, causing infinite looping
1643 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1644 with the subject "ab", where knowledge that the repeated group can match
1645 nothing is needed in order to break the loop.
1646
1647 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1648 exec(), but without supplying a callout function, matching went wrong.
1649
1650 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1651 leak if the size of the offset vector was greater than 30. When the vector
1652 is smaller, the saved offsets during recursion go onto a local stack
1653 vector, but for larger vectors malloc() is used. It was failing to free
1654 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1655 error, in fact).
1656
1657 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1658 heapframe that is used only when UTF-8 support is enabled. This caused no
1659 problem, but was untidy.
1660
1661 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1662 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1663 included within another project.
1664
1665 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1666 slightly modified by me:
1667
1668 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1669 not building pcregrep.
1670
1671 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1672 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1673
1674 25. Forward references, both numeric and by name, in patterns that made use of
1675 duplicate group numbers, could behave incorrectly or give incorrect errors,
1676 because when scanning forward to find the reference group, PCRE was not
1677 taking into account the duplicate group numbers. A pattern such as
1678 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1679
1680 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1681 the feature of a custom pointer more persuasive (as requested by a user).
1682
1683 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1684 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1685 pre-defined.
1686
1687 28. Added support for (*UTF8) at the start of a pattern.
1688
1689 29. Arrange for flags added by the "release type" setting in CMake to be shown
1690 in the configuration summary.
1691
1692
1693 Version 7.8 05-Sep-08
1694 ---------------------
1695
1696 1. Replaced UCP searching code with optimized version as implemented for Ad
1697 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1698 stage table and inline lookup instead of a function, giving speed ups of 2
1699 to 5 times on some simple patterns that I tested. Permission was given to
1700 distribute the MultiStage2.py script that generates the tables (it's not in
1701 the tarball, but is in the Subversion repository).
1702
1703 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1704 scripts.
1705
1706 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1707 a group with a zero qualifier. The result of the study could be incorrect,
1708 or the function might crash, depending on the pattern.
1709
1710 4. Caseless matching was not working for non-ASCII characters in back
1711 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1712 It now works when Unicode Property Support is available.
1713
1714 5. In pcretest, an escape such as \x{de} in the data was always generating
1715 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1716 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1717 truncation.
1718
1719 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1720
1721 7. Added two (int) casts to pcregrep when printing the difference of two
1722 pointers, in case they are 64-bit values.
1723
1724 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1725 test 2 if it fails.
1726
1727 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1728 and a #define of that name to empty if it is not externally set. This is to
1729 allow users of MSVC to set it if necessary.
1730
1731 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1732 the convenience functions in the pcre_get.c source file.
1733
1734 11. An option change at the start of a pattern that had top-level alternatives
1735 could cause overwriting and/or a crash. This command provoked a crash in
1736 some environments:
1737
1738 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1739
1740 This potential security problem was recorded as CVE-2008-2371.
1741
1742 12. For a pattern where the match had to start at the beginning or immediately
1743 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1744 pcre_dfa_exec() could read past the end of the passed subject if there was
1745 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1746 pcretest so that it places the subject at the end of its malloc-ed buffer.
1747
1748 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1749 exec() might read past the end of the data buffer in UTF-8 mode.
1750
1751 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1752 the data contained the byte 0x85 as part of a UTF-8 character within its
1753 first line. This applied both to normal and DFA matching.
1754
1755 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1756 /^[^d]*?$/8 failed to match "abc".
1757
1758 16. Added a missing copyright notice to pcrecpp_internal.h.
1759
1760 17. Make it more clear in the documentation that values returned from
1761 pcre_exec() in ovector are byte offsets, not character counts.
1762
1763 18. Tidied a few places to stop certain compilers from issuing warnings.
1764
1765 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1766 supplied by Stefan Weber. I made a further small update for 7.8 because
1767 there is a change of source arrangements: the pcre_searchfuncs.c module is
1768 replaced by pcre_ucd.c.
1769
1770
1771 Version 7.7 07-May-08
1772 ---------------------
1773
1774 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1775 a string to a long long, pretend we don't even have a long long." This is
1776 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1777
1778 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1779 pre-7.6 versions, which defined a global no_arg variable instead of putting
1780 it in the RE class. (See also #8 below.)
1781
1782 3. Remove a line of dead code, identified by coverity and reported by Nuno
1783 Lopes.
1784
1785 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1786
1787 (1) The include/exclude patterns were being applied to the whole pathnames
1788 of files, instead of just to the final components.
1789
1790 (2) If there was more than one level of directory, the subdirectories were
1791 skipped unless they satisfied the include/exclude conditions. This is
1792 inconsistent with GNU grep (and could even be seen as contrary to the
1793 pcregrep specification - which I improved to make it absolutely clear).
1794 The action now is always to scan all levels of directory, and just
1795 apply the include/exclude patterns to regular files.
1796
1797 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1798 --exclude_dir in the tests to avoid scanning .svn directories.
1799
1800 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1801 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1802 doesn't support NULs in patterns.
1803
1804 7. Added some missing "const"s to declarations of static tables in
1805 pcre_compile.c and pcre_dfa_exec.c.
1806
1807 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1808 caused by fix #2 above. (Subsequently also a second patch to fix the
1809 first patch. And a third patch - this was a messy problem.)
1810
1811 9. Applied Craig's patch to remove the use of push_back().
1812
1813 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1814 matching function regexec().
1815
1816 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1817 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1818 references. PCRE supports relative numbers with this syntax (I don't think
1819 Oniguruma does).
1820
1821 12. Previously, a group with a zero repeat such as (...){0} was completely
1822 omitted from the compiled regex. However, this means that if the group
1823 was called as a subroutine from elsewhere in the pattern, things went wrong
1824 (an internal error was given). Such groups are now left in the compiled
1825 pattern, with a new opcode that causes them to be skipped at execution
1826 time.
1827
1828 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1829 to the way PCRE behaves:
1830
1831 (a) A lone ] character is dis-allowed (Perl treats it as data).
1832
1833 (b) A back reference to an unmatched subpattern matches an empty string
1834 (Perl fails the current match path).
1835
1836 (c) A data ] in a character class must be notated as \] because if the
1837 first data character in a class is ], it defines an empty class. (In
1838 Perl it is not possible to have an empty class.) The empty class []
1839 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1840 The negative empty class [^] matches any one character, independently
1841 of the DOTALL setting.
1842
1843 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1844 non-existent subpattern following a character class starting with ']' and
1845 containing () gave an internal compiling error instead of "reference to
1846 non-existent subpattern". Fortunately, when the pattern did exist, the
1847 compiled code was correct. (When scanning forwards to check for the
1848 existencd of the subpattern, it was treating the data ']' as terminating
1849 the class, so got the count wrong. When actually compiling, the reference
1850 was subsequently set up correctly.)
1851
1852 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1853 it was being rejected as not supported by pcre_dfa_exec(), even though
1854 other assertions are supported. I have made pcre_dfa_exec() support
1855 (*FAIL).
1856
1857 16. The implementation of 13c above involved the invention of a new opcode,
1858 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1859 cannot be changed at match time, I realized I could make a small
1860 improvement to matching performance by compiling OP_ALLANY instead of
1861 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1862 on the OP_ANY path.
1863
1864 17. Compiling pcretest on Windows with readline support failed without the
1865 following two fixes: (1) Make the unistd.h include conditional on
1866 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1867
1868 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1869 ncurses library to be included for pcretest when ReadLine support is
1870 requested, but also to allow for it to be overridden. This patch came from
1871 Daniel Bergström.
1872
1873 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1874 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1875 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1876 spotting this.
1877
1878
1879 Version 7.6 28-Jan-08
1880 ---------------------
1881
1882 1. A character class containing a very large number of characters with
1883 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1884 overflow.
1885
1886 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1887 HAVE_LONG_LONG is not defined.
1888
1889 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1890 bring them up to date and include new features. This patch includes:
1891
1892 - Fixed PH's badly added libz and libbz2 support.
1893 - Fixed a problem with static linking.
1894 - Added pcredemo. [But later removed - see 7 below.]
1895 - Fixed dftables problem and added an option.
1896 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1897 HAVE_LONG_LONG.
1898 - Added readline support for pcretest.
1899 - Added an listing of the option settings after cmake has run.
1900
1901 4. A user submitted a patch to Makefile that makes it easy to create
1902 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1903 Makefile.am that cause it to include this special target, without
1904 affecting anything else. Note that the same mingw target plus all
1905 the other distribution libraries and programs are now supported
1906 when configuring with CMake (see 6 below) instead of with
1907 Configure/Make.
1908
1909 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1910 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1911 exported in the Windows port". It has not yet been confirmed that the patch
1912 solves the problem, but it does no harm.
1913
1914 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1915 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1916 with CMake, and also correct the comment about stack recursion.
1917
1918 7. Remove the automatic building of pcredemo from the ./configure system and
1919 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1920 of a program that users should build themselves after PCRE is installed, so
1921 building it automatically is not really right. What is more, it gave
1922 trouble in some build environments.
1923
1924 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1925
1926
1927 Version 7.5 10-Jan-08
1928 ---------------------
1929
1930 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1931 values in parens when parsing an RE using the C++ wrapper."
1932
1933 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1934 Characters greater than 255 were excluded from the class instead of being
1935 included.
1936
1937 3. The same bug as (2) above applied to negated POSIX classes such as
1938 [:^space:].
1939
1940 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1941 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1942 I have changed it.
1943
1944 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1945 first named subpattern) and a construct such as (?&a) would reference the
1946 first named subpattern whose name started with "a" (in other words, the
1947 length check was missing). Both these problems are fixed. "Subpattern name
1948 expected" is now given for (?&) (a zero-length name), and this patch also
1949 makes it give the same error for \k'' (previously it complained that that
1950 was a reference to a non-existent subpattern).
1951
1952 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1953 this is right because (?- can be followed by option settings as well as by
1954 digits. I have, however, made the messages clearer.
1955
1956 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1957 than the number used in the conditional) now cause a compile-time error.
1958 This is actually not compatible with Perl, which accepts such patterns, but
1959 treats the conditional as always being FALSE (as PCRE used to), but it
1960 seems to me that giving a diagnostic is better.
1961
1962 8. Change "alphameric" to the more common word "alphanumeric" in comments
1963 and messages.
1964
1965 9. Fix two occurrences of "backslash" in comments that should have been
1966 "backspace".
1967
1968 10. Remove two redundant lines of code that can never be obeyed (their function
1969 was moved elsewhere).
1970
1971 11. The program that makes PCRE's Unicode character property table had a bug
1972 which caused it to generate incorrect table entries for sequences of
1973 characters that have the same character type, but are in different scripts.
1974 It amalgamated them into a single range, with the script of the first of
1975 them. In other words, some characters were in the wrong script. There were
1976 thirteen such cases, affecting characters in the following ranges:
1977
1978 U+002b0 - U+002c1
1979 U+0060c - U+0060d
1980 U+0061e - U+00612
1981 U+0064b - U+0065e
1982 U+0074d - U+0076d
1983 U+01800 - U+01805
1984 U+01d00 - U+01d77
1985 U+01d9b - U+01dbf
1986 U+0200b - U+0200f
1987 U+030fc - U+030fe
1988 U+03260 - U+0327f
1989 U+0fb46 - U+0fbb1
1990 U+10450 - U+1049d
1991
1992 12. The -o option (show only the matching part of a line) for pcregrep was not
1993 compatible with GNU grep in that, if there was more than one match in a
1994 line, it showed only the first of them. It now behaves in the same way as
1995 GNU grep.
1996
1997 13. If the -o and -v options were combined for pcregrep, it printed a blank
1998 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1999 does the same. The return code can be used to tell if there were any
2000 non-matching lines.
2001
2002 14. Added --file-offsets and --line-offsets to pcregrep.
2003
2004 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2005 infinitely looping recursion. The bug was that positive lookaheads were not
2006 being skipped when checking for a possible empty match (negative lookaheads
2007 and both kinds of lookbehind were skipped).
2008
2009 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2010 inclusion of <windows.h> to before rather than after the definition of
2011 INVALID_FILE_ATTRIBUTES (patch from David Byron).
2012
2013 17. Specifying a possessive quantifier with a specific limit for a Unicode
2014 character property caused pcre_compile() to compile bad code, which led at
2015 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2016 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2017 caused the error; without that there was no problem.
2018
2019 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2020
2021 19. Added --enable-pcretest-libreadline.
2022
2023 20. In pcrecpp.cc, the variable 'count' was incremented twice in
2024 RE::GlobalReplace(). As a result, the number of replacements returned was
2025 double what it should be. I removed one of the increments, but Craig sent a
2026 later patch that removed the other one (the right fix) and added unit tests
2027 that check the return values (which was not done before).
2028
2029 21. Several CMake things:
2030
2031 (1) Arranged that, when cmake is used on Unix, the libraries end up with
2032 the names libpcre and libpcreposix, not just pcre and pcreposix.
2033
2034 (2) The above change means that pcretest and pcregrep are now correctly
2035 linked with the newly-built libraries, not previously installed ones.
2036
2037 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2038
2039 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2040 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2041 UTF-8 newline character). The key issue is that the pattern starts .*;
2042 this means that the match must be either at the beginning, or after a
2043 newline. The bug was in the code for advancing after a failed match and
2044 checking that the new position followed a newline. It was not taking
2045 account of UTF-8 characters correctly.
2046
2047 23. PCRE was behaving differently from Perl in the way it recognized POSIX
2048 character classes. PCRE was not treating the sequence [:...:] as a
2049 character class unless the ... were all letters. Perl, however, seems to
2050 allow any characters between [: and :], though of course it rejects as
2051 unknown any "names" that contain non-letters, because all the known class
2052 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2053 for example, whereas PCRE did not - it did not recognize a POSIX character
2054 class. This seemed a bit dangerous, so the code has been changed to be
2055 closer to Perl. The behaviour is not identical to Perl, because PCRE will
2056 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2057 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2058 Perl does, and where it didn't before.
2059
2060 24. Rewrite so as to remove the single use of %n from pcregrep because in some
2061 Windows environments %n is disabled by default.
2062
2063
2064 Version 7.4 21-Sep-07
2065 ---------------------
2066
2067 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
2068 means that a class such as [\s] counted as "explicit reference to CR or
2069 LF". That isn't really right - the whole point of the change was to try to
2070 help when there was an actual mention of one of the two characters. So now
2071 the change happens only if \r or \n (or a literal CR or LF) character is
2072 encountered.
2073
2074 2. The 32-bit options word was also used for 6 internal flags, but the numbers
2075 of both had grown to the point where there were only 3 bits left.
2076 Fortunately, there was spare space in the data structure, and so I have
2077 moved the internal flags into a new 16-bit field to free up more option
2078 bits.
2079
2080 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2081 but did not set the internal JCHANGED flag - either of these is enough to
2082 control the way the "get" function works - but the PCRE_INFO_JCHANGED
2083 facility is supposed to tell if (?J) was ever used, so now (?J) at the
2084 start sets both bits.
2085
2086 4. Added options (at build time, compile time, exec time) to change \R from
2087 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2088
2089 5. doc/pcresyntax.html was missing from the distribution.
2090
2091 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2092 compatibility, even though it is no longer used.
2093
2094 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2095 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2096 windows.h file is present (where different names are used). [This was
2097 reversed later after testing - see 16 below.]
2098
2099 8. Changed all #include <config.h> to #include "config.h". There were also
2100 some further <pcre.h> cases that I changed to "pcre.h".
2101
2102 9. When pcregrep was used with the --colour option, it missed the line ending
2103 sequence off the lines that it output.
2104
2105 10. It was pointed out to me that arrays of string pointers cause lots of
2106 relocations when a shared library is dynamically loaded. A technique of
2107 using a single long string with a table of offsets can drastically reduce
2108 these. I have refactored PCRE in four places to do this. The result is
2109 dramatic:
2110
2111 Originally: 290
2112 After changing UCP table: 187
2113 After changing error message table: 43
2114 After changing table of "verbs" 36
2115 After changing table of Posix names 22
2116
2117 Thanks to the folks working on Gregex for glib for this insight.
2118
2119 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2120 unicode-properties was also set.
2121
2122 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2123
2124 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2125 checked only for CRLF.
2126
2127 14. Added casts to pcretest.c to avoid compiler warnings.
2128
2129 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2130
2131 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2132 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2133 entirely. This removes changes made in 7 above.
2134
2135 17. The CMake files have been updated, and there is now more information about
2136 building with CMake in the NON-UNIX-USE document.
2137
2138
2139 Version 7.3 28-Aug-07
2140 ---------------------
2141
2142 1. In the rejigging of the build system that eventually resulted in 7.1, the
2143 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2144 brackets there is not right, since it causes compilers to look for an
2145 installed pcre.h, not the version that is in the source that is being
2146 compiled (which of course may be different). I have changed it back to:
2147
2148 #include "pcre.h"
2149
2150 I have a vague recollection that the change was concerned with compiling in
2151 different directories, but in the new build system, that is taken care of
2152 by the VPATH setting the Makefile.
2153
2154 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2155 when the subject happened to end in the byte 0x85 (e.g. if the last
2156 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2157 characters but of course it shouldn't be taken as a newline when it is part
2158 of another character. The bug was that, for an unlimited repeat of . in
2159 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2160 characters when looking for a newline.
2161
2162 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2163
2164 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2165 in debug output.
2166
2167 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2168 long printing in the pcrecpp unittest when running under MinGW.
2169
2170 6. ESC_K was left out of the EBCDIC table.
2171
2172 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2173 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2174 limit also applies to "virtual nesting" when a pattern is recursive, and in
2175 this case 1000 isn't so big. I have been able to remove this limit at the
2176 expense of backing off one optimization in certain circumstances. Normally,
2177 when pcre_exec() would call its internal match() function recursively and
2178 immediately return the result unconditionally, it uses a "tail recursion"
2179 feature to save stack. However, when a subpattern that can match an empty
2180 string has an unlimited repetition quantifier, it no longer makes this
2181 optimization. That gives it a stack frame in which to save the data for
2182 checking that an empty string has been matched. Previously this was taken
2183 from the 1000-entry workspace that had been reserved. So now there is no
2184 explicit limit, but more stack is used.
2185
2186 8. Applied Daniel's patches to solve problems with the import/export magic
2187 syntax that is required for Windows, and which was going wrong for the
2188 pcreposix and pcrecpp parts of the library. These were overlooked when this
2189 problem was solved for the main library.
2190
2191 9. There were some crude static tests to avoid integer overflow when computing
2192 the size of patterns that contain repeated groups with explicit upper
2193 limits. As the maximum quantifier is 65535, the maximum group length was
2194 set at 30,000 so that the product of these two numbers did not overflow a
2195 32-bit integer. However, it turns out that people want to use groups that
2196 are longer than 30,000 bytes (though not repeat them that many times).
2197 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2198 made it possible to implement the integer overflow checks in a much more
2199 dynamic way, which I have now done. The artificial limitation on group
2200 length has been removed - we now have only the limit on the total length of
2201 the compiled pattern, which depends on the LINK_SIZE setting.
2202
2203 10. Fixed a bug in the documentation for get/copy named substring when
2204 duplicate names are permitted. If none of the named substrings are set, the
2205 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2206 empty string.
2207
2208 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2209 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2210 because the ] is interpreted as the first data character and the
2211 terminating ] is not found. PCRE has been made compatible with Perl in this
2212 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2213 cause memory overwriting.
2214
2215 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2216 string has been matched (to stop an infinite loop). It was not recognizing
2217 a conditional subpattern that could match an empty string if that
2218 subpattern was within another subpattern. For example, it looped when
2219 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
2220 condition was not nested. This bug has been fixed.
2221
2222 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2223 past the start of the subject in the presence of bytes with the top bit
2224 set, for example "\x8aBCD".
2225
2226 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2227 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2228
2229 14. Optimized (?!) to (*FAIL).
2230
2231 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2232 This restricts code points to be within the range 0 to 0x10FFFF, excluding
2233 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2234 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2235 does: it's just the validity check that is more restrictive.
2236
2237 16. Inserted checks for integer overflows during escape sequence (backslash)
2238 processing, and also fixed erroneous offset values for syntax errors during
2239 backslash processing.
2240
2241 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2242 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2243
2244 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2245 caused an overrun.
2246
2247 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2248 something other than just ASCII characters) inside a group that had an
2249 unlimited repeat caused a loop at compile time (while checking to see
2250 whether the group could match an empty string).
2251
2252 20. Debugging a pattern containing \p or \P could cause a crash. For example,
2253 [\P{Any}] did so. (Error in the code for printing property names.)
2254
2255 21. An orphan \E inside a character class could cause a crash.
2256
2257 22. A repeated capturing bracket such as (A)? could cause a wild memory
2258 reference during compilation.
2259
2260 23. There are several functions in pcre_compile() that scan along a compiled
2261 expression for various reasons (e.g. to see if it's fixed length for look
2262 behind). There were bugs in these functions when a repeated \p or \P was
2263 present in the pattern. These operators have additional parameters compared
2264 with \d, etc, and these were not being taken into account when moving along
2265 the compiled data. Specifically:
2266
2267 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2268 length.
2269
2270 (b) An item such as \pL+ within a repeated group could cause crashes or
2271 loops.
2272
2273 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2274 "reference to non-existent subpattern" error.
2275
2276 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2277
2278 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2279 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2280
2281 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2282
2283 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2284 character were causing crashes (broken optimization).
2285
2286 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2287 \p or \P) caused a compile-time loop.
2288
2289 28. More problems have arisen in unanchored patterns when CRLF is a valid line
2290 break. For example, the unstudied pattern [\r\n]A does not match the string
2291 "\r\nA" because change 7.0/46 below moves the current point on by two
2292 characters after failing to match at the start. However, the pattern \nA
2293 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2294 the same is true. There doesn't seem any very clean way out of this, but
2295 what I have chosen to do makes the common cases work: PCRE now takes note
2296 of whether there can be an explicit match for \r or \n anywhere in the
2297 pattern, and if so, 7.0/46 no longer applies. As part of this change,
2298 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2299 pattern has explicit CR or LF references.
2300
2301 29. Added (*CR) etc for changing newline setting at start of pattern.
2302
2303
2304 Version 7.2 19-Jun-07
2305 ---------------------
2306
2307 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2308 which is apparently normally available under Windows.
2309
2310 2. Re-jig the pcregrep tests with different newline settings in an attempt
2311 to make them independent of the local environment's newline setting.
2312
2313 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2314
2315 4. Some of the "internals" tests were previously cut out when the link size
2316 was not 2, because the output contained actual offsets. The recent new
2317 "Z" feature of pcretest means that these can be cut out, making the tests
2318 usable with all link sizes.
2319
2320 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2321 stack recursion. This gives a massive performance boost under BSD, but just
2322 a small improvement under Linux. However, it saves one field in the frame
2323 in all cases.
2324
2325 6. Added more features from the forthcoming Perl 5.10:
2326
2327 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2328 recursion call. It refers to the nth most recently opened parentheses.
2329
2330 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2331 to be opened parentheses.
2332
2333 (c) Conditions that refer to capturing parentheses can be specified
2334 relatively, for example, (?(-2)... or (?(+3)...
2335
2336 (d) \K resets the start of the current match so that everything before
2337 is not part of it.
2338
2339 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2340
2341 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2342 reference syntax.
2343
2344 (g) (?| introduces a group in which the numbering of parentheses in each
2345 alternative starts with the same number.
2346
2347 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2348
2349 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2350 PCRE_INFO_JCHANGED.
2351
2352 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2353 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2354 for detecting groups that can match an empty string.
2355
2356 9. A pattern with a very large number of alternatives (more than several
2357 hundred) was running out of internal workspace during the pre-compile
2358 phase, where pcre_compile() figures out how much memory will be needed. A
2359 bit of new cunning has reduced the workspace needed for groups with
2360 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2361 workspace instead of running out of the 4096 that are available.
2362
2363 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2364
2365 11. Applied patch from Google to remove an optimization that didn't quite work.
2366 The report of the bug said:
2367
2368 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2369 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2370 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2371
2372 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2373 it matched the wrong number of bytes.
2374
2375
2376 Version 7.1 24-Apr-07
2377 ---------------------
2378
2379 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2380 that is more "standard", making use of automake and other Autotools. There
2381 is some re-arrangement of the files and adjustment of comments consequent
2382 on this.
2383
2384 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2385 for recursive directory scanning broke on some systems because the files
2386 are not scanned in any specific order and on different systems the order
2387 was different. A call to "sort" has been inserted into RunGrepTest for the
2388 approprate test as a short-term fix. In the longer term there may be an
2389 alternative.
2390
2391 3. I had an email from Eric Raymond about problems translating some of PCRE's
2392 man pages to HTML (despite the fact that I distribute HTML pages, some
2393 people do their own conversions for various reasons). The problems
2394 concerned the use of low-level troff macros .br and .in. I have therefore
2395 removed all such uses from the man pages (some were redundant, some could
2396 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2397 HTML has been updated to handle .nf/.fi and to complain if it encounters
2398 .br or .in.
2399
2400 4. Updated comments in configure.ac that get placed in config.h.in and also
2401 arranged for config.h to be included in the distribution, with the name
2402 config.h.generic, for the benefit of those who have to compile without
2403 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2404
2405 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2406 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2407 makevp.bat for the current PCRE, using the additional files
2408 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2409
2410 6. A Windows user reported a minor discrepancy with test 2, which turned out
2411 to be caused by a trailing space on an input line that had got lost in his
2412 copy. The trailing space was an accident, so I've just removed it.
2413
2414 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2415 that is needed.
2416
2417 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2418 as "const" (a) because they are and (b) because it helps the PHP
2419 maintainers who have recently made a script to detect big data structures
2420 in the php code that should be moved to the .rodata section. I remembered
2421 to update Builducptable as well, so it won't revert if ucptable.h is ever
2422 re-created.
2423
2424 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2425 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2426 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2427 support is not required. This saves 1.5-2K of code, which is important in
2428 some applications.
2429
2430 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2431 so as not to refer to the tables, even though these functions will never be
2432 called when UTF-8 support is disabled. Otherwise there are problems with a
2433 shared library.
2434
2435 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2436
2437 (a) It was defining its arguments as char * instead of void *.
2438
2439 (b) It was assuming that all moves were upwards in memory; this was true
2440 a long time ago when I wrote it, but is no longer the case.
2441
2442 The emulated memove() is provided for those environments that have neither
2443 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2444 is clearly not the case, as these two bugs were recently reported.
2445
2446 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2447 and Detrail to create the HTML documentation, the .txt form of the man
2448 pages, and it removes trailing spaces from listed files. It also creates
2449 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2450 case, it wraps all the #defines with #ifndefs. This script should be run
2451 before "make dist".
2452
2453 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2454 with Unicode property support.
2455
2456 (a) For a maximizing quantifier, if the two different cases of the
2457 character were of different lengths in their UTF-8 codings (there are
2458 some cases like this - I found 11), and the matching function had to
2459 back up over a mixture of the two cases, it incorrectly assumed they
2460 were both the same length.
2461
2462 (b) When PCRE was configured to use the heap rather than the stack for
2463 recursion during matching, it was not correctly preserving the data for
2464 the other case of a UTF-8 character when checking ahead for a match
2465 while processing a minimizing repeat. If the check also involved
2466 matching a wide character, but failed, corruption could cause an
2467 erroneous result when trying to check for a repeat of the original
2468 character.
2469
2470 13. Some tidying changes to the testing mechanism:
2471
2472 (a) The RunTest script now detects the internal link size and whether there
2473 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2474 values substituted by "configure". (The RunGrepTest script already did
2475 this for UTF-8.) The configure.ac script no longer substitutes the
2476 relevant variables.
2477
2478 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2479 with length and offset values. This means that the output is different
2480 for different internal link sizes. Test 2 is skipped for link sizes
2481 other than 2 because of this, bypassing the problem. Unfortunately,
2482 there was also a test in test 3 (the locale tests) that used /B and
2483 failed for link sizes other than 2. Rather than cut the whole test out,
2484 I have added a new /Z option to pcretest that replaces the length and
2485 offset values with spaces. This is now used to make test 3 independent
2486 of link size. (Test 2 will be tidied up later.)
2487
2488 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2489 segmentation fault instead of returning the appropriate error message.
2490
2491 15. In multiline mode when the newline sequence was set to "any", the pattern
2492 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2493 This doesn't seem right; it now treats the CRLF combination as the line
2494 ending, and so does not match in that case. It's only a pattern such as ^$
2495 that would hit this one: something like ^ABC$ would have failed after \r
2496 and then tried again after \r\n.
2497
2498 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2499 in an attempt to make files that differ only in their line terminators
2500 compare equal. This works on Linux.
2501
2502 17. Under certain error circumstances pcregrep might try to free random memory
2503 as it exited. This is now fixed, thanks to valgrind.
2504
2505 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2506 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2507 was because its rules for how to advance for /g after matching an empty
2508 string at the end of a line did not allow for this case. They now check for
2509 it specially.
2510
2511 20. pcretest is supposed to handle patterns and data of any length, by
2512 extending its buffers when necessary. It was getting this wrong when the
2513 buffer for a data line had to be extended.
2514
2515 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2516 CRLF as a newline sequence.
2517
2518 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2519 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2520 I have nevertheless tidied it up.
2521
2522 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2523
2524 24. Added a man page for pcre-config.
2525
2526
2527 Version 7.0 19-Dec-06
2528 ---------------------
2529
2530 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2531 moving to gcc 4.1.1.
2532
2533 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2534 sys/time.h, which is documented as needed for this function. It doesn't
2535 seem to matter on Linux, but it showed up on some releases of OS X.
2536
2537 3. It seems that there are systems where bytes whose values are greater than
2538 127 match isprint() in the "C" locale. The "C" locale should be the
2539 default when a C program starts up. In most systems, only ASCII printing
2540 characters match isprint(). This difference caused the output from pcretest
2541 to vary, making some of the tests fail. I have changed pcretest so that:
2542
2543 (a) When it is outputting text in the compiled version of a pattern, bytes
2544 other than 32-126 are always shown as hex escapes.
2545
2546 (b) When it is outputting text that is a matched part of a subject string,
2547 it does the same, unless a different locale has been set for the match
2548 (using the /L modifier). In this case, it uses isprint() to decide.
2549
2550 4. Fixed a major bug that caused incorrect computation of the amount of memory
2551 required for a compiled pattern when options that changed within the
2552 pattern affected the logic of the preliminary scan that determines the
2553 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2554 that the computed length was too small. The symptoms of this bug were
2555 either the PCRE error "internal error: code overflow" from pcre_compile(),
2556 or a glibc crash with a message such as "pcretest: free(): invalid next
2557 size (fast)". Examples of patterns that provoked this bug (shown in
2558 pcretest format) are:
2559
2560 /(?-x: )/x
2561 /(?x)(?-x: \s*#\s*)/
2562 /((?i)[\x{c0}])/8
2563 /(?i:[\x{c0}])/8
2564
2565 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2566 is now done differently.
2567
2568 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2569 wrapper classes; (b) implement a new function in the C++ scanner that is
2570 more efficient than the old way of doing things because it avoids levels of
2571 recursion in the regex matching; (c) add a paragraph to the documentation
2572 for the FullMatch() function.
2573
2574 6. The escape sequence \n was being treated as whatever was defined as
2575 "newline". Not only was this contrary to the documentation, which states
2576 that \n is character 10 (hex 0A), but it also went horribly wrong when
2577 "newline" was defined as CRLF. This has been fixed.
2578
2579 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2580 was being set to -1 for the "end of line" case (supposedly a value that no
2581 character can have). Though this value is never used (the check for end of
2582 line is "zero bytes in current character"), it caused compiler complaints.
2583 I've changed it to 0xffffffff.
2584
2585 8. In pcre_version.c, the version string was being built by a sequence of
2586 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2587 string (as it is for production releases) called a macro with an empty
2588 argument. The C standard says the result of this is undefined. The gcc
2589 compiler treats it as an empty string (which was what was wanted) but it is
2590 reported that Visual C gives an error. The source has been hacked around to
2591 avoid this problem.
2592
2593 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2594 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2595 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2596 of them did).
2597
2598 10. Originally, pcretest opened its input and output without "b"; then I was
2599 told that "b" was needed in some environments, so it was added for release
2600 5.0 to both the input and output. (It makes no difference on Unix-like
2601 systems.) Later I was told that it is wrong for the input on Windows. I've
2602 now abstracted the modes into two macros, to make it easier to fiddle with
2603 them, and removed "b" from the input mode under Windows.
2604
2605 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2606
2607 12. Added -help and --help to pcretest as an official way of being reminded
2608 of the options.
2609
2610 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2611 and pcrecpp.cc because they annoy compilers at high warning levels.
2612
2613 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2614
2615 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2616 scripts are not C programs :-) and which was not noticed because it works
2617 on Linux.
2618
2619 16. pcretest is supposed to handle any length of pattern and data line (as one
2620 line or as a continued sequence of lines) by extending its input buffer if
2621 necessary. This feature was broken for very long pattern lines, leading to
2622 a string of junk being passed to pcre_compile() if the pattern was longer
2623 than about 50K.
2624
2625 17. I have done a major re-factoring of the way pcre_compile() computes the
2626 amount of memory needed for a compiled pattern. Previously, there was code
2627 that made a preliminary scan of the pattern in order to do this. That was
2628 OK when PCRE was new, but as the facilities have expanded, it has become
2629 harder and harder to keep it in step with the real compile phase, and there
2630 have been a number of bugs (see for example, 4 above). I have now found a
2631 cunning way of running the real compile function in a "fake" mode that
2632 enables it to compute how much memory it would need, while actually only
2633 ever using a few hundred bytes of working memory and without too many
2634 tests of the mode. This should make future maintenance and development
2635 easier. A side effect of this work is that the limit of 200 on the nesting
2636 depth of parentheses has been removed (though this was never a serious
2637 limitation, I suspect). However, there is a downside: pcre_compile() now
2638 runs more slowly than before (30% or more, depending on the pattern). I
2639 hope this isn't a big issue. There is no effect on runtime performance.
2640
2641 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2642 newline (only possible for the last line of a file) and it was a
2643 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2644
2645 19. Added additional timing features to pcretest. (1) The -tm option now times
2646 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2647 separate command line item, by a number that specifies the number of
2648 repeats to use when timing. The default is 50000; this gives better
2649 precision, but takes uncomfortably long for very large patterns.
2650
2651 20. Extended pcre_study() to be more clever in cases where a branch of a
2652 subpattern has no definite first character. For example, (a*|b*)[cd] would
2653 previously give no result from pcre_study(). Now it recognizes that the
2654 first character must be a, b, c, or d.
2655
2656 21. There was an incorrect error "recursive call could loop indefinitely" if
2657 a subpattern (or the entire pattern) that was being tested for matching an
2658 empty string contained only one non-empty item after a nested subpattern.
2659 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2660 incorrectly, because the \d was being skipped in the check.
2661
2662 22. The pcretest program now has a new pattern option /B and a command line
2663 option -b, which is equivalent to adding /B to every pattern. This causes
2664 it to show the compiled bytecode, without the additional information that
2665 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2666 is the same as /B/I).
2667
2668 23. A new optimization is now able automatically to treat some sequences such
2669 as a*b as a*+b. More specifically, if something simple (such as a character
2670 or a simple class like \d) has an unlimited quantifier, and is followed by
2671 something that cannot possibly match the quantified thing, the quantifier
2672 is automatically "possessified".
2673
2674 24. A recursive reference to a subpattern whose number was greater than 39
2675 went wrong under certain circumstances in UTF-8 mode. This bug could also
2676 have affected the operation of pcre_study().
2677
2678 25. Realized that a little bit of performance could be had by replacing
2679 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2680
2681 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2682
2683 27. Possessive quantifiers such as a++ were previously implemented by turning
2684 them into atomic groups such as ($>a+). Now they have their own opcodes,
2685 which improves performance. This includes the automatically created ones
2686 from 23 above.
2687
2688 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2689 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2690 the first matched character to be a colon. This applied both to named and
2691 numbered groups.
2692
2693 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2694
2695 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2696 building PCRE on Windows easier, so I have included it in the distribution.
2697
2698 31. There is now a check in pcretest against a ridiculously large number being
2699 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2700 loop, the loop is abandoned.
2701
2702 32. Forward references to subpatterns in conditions such as (?(2)...) where
2703 subpattern 2 is defined later cause pcre_compile() to search forwards in
2704 the pattern for the relevant set of parentheses. This search went wrong
2705 when there were unescaped parentheses in a character class, parentheses
2706 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2707
2708 33. "Subroutine" calls and backreferences were previously restricted to
2709 referencing subpatterns earlier in the regex. This restriction has now
2710 been removed.
2711
2712 34. Added a number of extra features that are going to be in Perl 5.10. On the
2713 whole, these are just syntactic alternatives for features that PCRE had
2714 previously implemented using the Python syntax or my own invention. The
2715 other formats are all retained for compatibility.
2716
2717 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2718 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2719 also .NET compatible.
2720
2721 (b) A recursion or subroutine call to a named group can now be defined as
2722 (?&name) as well as (?P>name).
2723
2724 (c) A backreference to a named group can now be defined as \k<name> or
2725 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2726 5.10, are also .NET compatible.
2727
2728 (d) A conditional reference to a named group can now use the syntax
2729 (?(<name>) or (?('name') as well as (?(name).
2730
2731 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2732 groups (named and numbered) that are never evaluated inline, but can be
2733 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2734 is always false. There may be only one alternative in such a group.
2735
2736 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2737 as the simple (?(R). The condition is true only if the most recent
2738 recursion is that of the given number or name. It does not search out
2739 through the entire recursion stack.
2740
2741 (g) The escape \gN or \g{N} has been added, where N is a positive or
2742 negative number, specifying an absolute or relative reference.
2743
2744 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2745 some "unreachable code" warnings.
2746
2747 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2748 things, this adds five new scripts.
2749
2750 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2751 There were also incompatibilities regarding the handling of \Q..\E inside
2752 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2753 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2754
2755 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2756 matches an empty string, and forcibly breaks the loop. There were bugs in
2757 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2758 against aaaa the result was just "a" rather than "aaaa", for example. Two
2759 separate and independent bugs (that affected different cases) have been
2760 fixed.
2761
2762 39. Refactored the code to abolish the use of different opcodes for small
2763 capturing bracket numbers. This is a tidy that I avoided doing when I
2764 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2765 The new approach is not only tidier, it makes it possible to reduce the
2766 memory needed to fix the previous bug (38).
2767
2768 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2769 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2770 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2771 mode.
2772
2773 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2774 report.
2775
2776 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2777 copy construction and assignment in the C++ wrapper.
2778
2779 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2780 couple of bugs that could have given wrong results in the "--newline=crlf"
2781 case.
2782
2783 44. Added a number of casts and did some reorganization of signed/unsigned int
2784 variables following suggestions from Dair Grant. Also renamed the variable
2785 "this" as "item" because it is a C++ keyword.
2786
2787 45. Arranged for dftables to add
2788
2789 #include "pcre_internal.h"
2790
2791 to pcre_chartables.c because without it, gcc 4.x may remove the array
2792 definition from the final binary if PCRE is built into a static library and
2793 dead code stripping is activated.
2794
2795 46. For an unanchored pattern, if a match attempt fails at the start of a
2796 newline sequence, and the newline setting is CRLF or ANY, and the next two
2797 characters are CRLF, advance by two characters instead of one.
2798
2799
2800 Version 6.7 04-Jul-06
2801 ---------------------
2802
2803 1. In order to handle tests when input lines are enormously long, pcretest has
2804 been re-factored so that it automatically extends its buffers when
2805 necessary. The code is crude, but this _is_ just a test program. The
2806 default size has been increased from 32K to 50K.
2807
2808 2. The code in pcre_study() was using the value of the re argument before
2809 testing it for NULL. (Of course, in any sensible call of the function, it
2810 won't be NULL.)
2811
2812 3. The memmove() emulation function in pcre_internal.h, which is used on
2813 systems that lack both memmove() and bcopy() - that is, hardly ever -
2814 was missing a "static" storage class specifier.
2815
2816 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2817 containing an extended class (one that cannot be represented by a bitmap
2818 because it contains high-valued characters or Unicode property items, e.g.
2819 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2820 pattern, but PCRE should not loop if you do not (it no longer does).
2821 [Detail: two cases were found: (a) a repeated subpattern containing an
2822 extended class; (b) a recursive reference to a subpattern that followed a
2823 previous extended class. It wasn't skipping over the extended class
2824 correctly when UTF-8 mode was not set.]
2825
2826 5. A negated single-character class was not being recognized as fixed-length
2827 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2828 compile error "lookbehind assertion is not fixed length".
2829
2830 6. The RunPerlTest auxiliary script was showing an unexpected difference
2831 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2832 write a Perl script that can interpret lines of an input file either as
2833 byte characters or as UTF-8, which is what "perltest" was being required to
2834 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2835 can't do is switch easily at run time between having the "use utf8;" pragma
2836 or not. In the end, I fudged it by using the RunPerlTest script to insert
2837 "use utf8;" explicitly for the UTF-8 tests.
2838
2839 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2840 the end of the subject string, contrary to the documentation and to what
2841 Perl does. This was true of both matching functions. Now it matches only at
2842 the start of the subject and immediately after *internal* newlines.
2843
2844 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2845 a pointer to an int instead of a pointer to an unsigned long int. This
2846 caused problems on 64-bit systems.
2847
2848 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2849 instance of the 'standard' template library not being so standard".
2850
2851 10. There was no check on the number of named subpatterns nor the maximum
2852 length of a subpattern name. The product of these values is used to compute
2853 the size of the memory block for a compiled pattern. By supplying a very
2854 long subpattern name and a large number of named subpatterns, the size
2855 computation could be caused to overflow. This is now prevented by limiting
2856 the length of names to 32 characters, and the number of named subpatterns
2857 to 10,000.
2858
2859 11. Subpatterns that are repeated with specific counts have to be replicated in
2860 the compiled pattern. The size of memory for this was computed from the
2861 length of the subpattern and the repeat count. The latter is limited to
2862 65535, but there was no limit on the former, meaning that integer overflow
2863 could in principle occur. The compiled length of a repeated subpattern is
2864 now limited to 30,000 bytes in order to prevent this.
2865
2866 12. Added the optional facility to have named substrings with the same name.
2867
2868 13. Added the ability to use a named substring as a condition, using the
2869 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2870 are numbers (not recommended). Forward references are permitted.
2871
2872 14. Added forward references in named backreferences (if you see what I mean).
2873
2874 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2875 pattern could run off the end of the subject. For example, the pattern
2876 "(?s)(.{1,5})"8 did this with the subject "ab".
2877
2878 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2879 PCRE_CASELESS was set when matching characters that were quantified with ?
2880 or *.
2881
2882 17. A character class other than a single negated character that had a minimum
2883 but no maximum quantifier - for example [ab]{6,} - was not handled
2884 correctly by pce_dfa_exec(). It would match only one character.
2885
2886 18. A valid (though odd) pattern that looked like a POSIX character
2887 class but used an invalid character after [ (for example [[,abc,]]) caused
2888 pcre_compile() to give the error "Failed: internal error: code overflow" or
2889 in some cases to crash with a glibc free() error. This could even happen if
2890 the pattern terminated after [[ but there just happened to be a sequence of
2891 letters, a binary zero, and a closing ] in the memory that followed.
2892
2893 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2894 over the years. Originally (before any Unicode support), just the bottom 8
2895 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2896 output from "man perlunicode" includes this:
2897
2898 The regular expression compiler produces polymorphic opcodes. That
2899 is, the pattern adapts to the data and automatically switches to
2900 the Unicode character scheme when presented with Unicode data--or
2901 instead uses a traditional byte scheme when presented with byte
2902 data.
2903
2904 Sadly, a wide octal escape does not cause a switch, and in a string with
2905 no other multibyte characters, these octal escapes are treated as before.
2906 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2907 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2908 Unicode string.
2909
2910 I have not perpetrated such confusion in PCRE. Up till now, it took just
2911 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2912 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2913 translate to the appropriate multibyte character.
2914
2915 29. Applied some refactoring to reduce the number of warnings from Microsoft
2916 and Borland compilers. This has included removing the fudge introduced
2917 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2918 a warning about an unused variable.
2919
2920 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2921 characters since release 4.0, because Perl (from release 5.004) does not.
2922 [Or at least, is documented not to: some releases seem to be in conflict
2923 with the documentation.] However, when a pattern was studied with
2924 pcre_study() and all its branches started with \s, PCRE still included VT
2925 as a possible starting character. Of course, this did no harm; it just
2926 caused an unnecessary match attempt.
2927
2928 22. Removed a now-redundant internal flag bit that recorded the fact that case
2929 dependency changed within the pattern. This was once needed for "required
2930 byte" processing, but is no longer used. This recovers a now-scarce options
2931 bit. Also moved the least significant internal flag bit to the most-
2932 significant bit of the word, which was not previously used (hangover from
2933 the days when it was an int rather than a uint) to free up another bit for
2934 the future.
2935
2936 23. Added support for CRLF line endings as well as CR and LF. As well as the
2937 default being selectable at build time, it can now be changed at runtime
2938 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2939 specify that it is scanning data with non-default line endings.
2940
2941 24. Changed the definition of CXXLINK to make it agree with the definition of
2942 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2943
2944 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2945 recursions. This makes a big different to stack usage for some patterns.
2946
2947 26. If a subpattern containing a named recursion or subroutine reference such
2948 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2949 the space required for the compiled pattern went wrong and gave too small a
2950 value. Depending on the environment, this could lead to "Failed: internal
2951 error: code overflow at offset 49" or "glibc detected double free or
2952 corruption" errors.
2953
2954 27. Applied patches from Google (a) to support the new newline modes and (b) to
2955 advance over multibyte UTF-8 characters in GlobalReplace.
2956
2957 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2958 difference for some implementation of PCRE in some Windows version.
2959
2960 29. Added some extra testing facilities to pcretest:
2961
2962 \q<number> in a data line sets the "match limit" value
2963 \Q<number> in a data line sets the "match recursion limt" value
2964 -S <number> sets the stack size, where <number> is in megabytes
2965
2966 The -S option isn't available for Windows.
2967
2968
2969 Version 6.6 06-Feb-06
2970 ---------------------
2971
2972 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2973 in pcreposix.h. I have copied the definition from pcre.h.
2974
2975 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2976 because pcre.h is no longer a built file.
2977
2978 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2979 not normally included in the compiled code.
2980
2981
2982 Version 6.5 01-Feb-06
2983 ---------------------
2984
2985 1. When using the partial match feature with pcre_dfa_exec(), it was not
2986 anchoring the second and subsequent partial matches at the new starting
2987 point. This could lead to incorrect results. For example, with the pattern
2988 /1234/, partially matching against "123" and then "a4" gave a match.
2989
2990 2. Changes to pcregrep:
2991
2992 (a) All non-match returns from pcre_exec() were being treated as failures
2993 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2994 error message is output. Some extra information is given for the
2995 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2996 probably the only errors that are likely to be caused by users (by
2997 specifying a regex that has nested indefinite repeats, for instance).
2998 If there are more than 20 of these errors, pcregrep is abandoned.
2999
3000 (b) A binary zero was treated as data while matching, but terminated the
3001 output line if it was written out. This has been fixed: binary zeroes
3002 are now no different to any other data bytes.
3003
3004 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3005 used to set a locale for matching. The --locale=xxxx long option has
3006 been added (no short equivalent) to specify a locale explicitly on the
3007 pcregrep command, overriding the environment variables.
3008
3009 (d) When -B was used with -n, some line numbers in the output were one less
3010 than they should have been.
3011
3012 (e) Added the -o (--only-matching) option.
3013
3014 (f) If -A or -C was used with -c (count only), some lines of context were
3015 accidentally printed for the final match.
3016
3017 (g) Added the -H (--with-filename) option.
3018
3019 (h) The combination of options -rh failed to suppress file names for files
3020 that were found from directory arguments.
3021
3022 (i) Added the -D (--devices) and -d (--directories) options.
3023
3024 (j) Added the -F (--fixed-strings) option.
3025
3026 (k) Allow "-" to be used as a file name for -f as well as for a data file.
3027
3028 (l) Added the --colo(u)r option.
3029
3030 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3031 is not present by default.
3032
3033 3. A nasty bug was discovered in the handling of recursive patterns, that is,
3034 items such as (?R) or (?1), when the recursion could match a number of
3035 alternatives. If it matched one of the alternatives, but subsequently,
3036 outside the recursion, there was a failure, the code tried to back up into
3037 the recursion. However, because of the way PCRE is implemented, this is not
3038 possible, and the result was an incorrect result from the match.
3039
3040 In order to prevent this happening, the specification of recursion has
3041 been changed so that all such subpatterns are automatically treated as
3042 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3043
3044 4. I had overlooked the fact that, in some locales, there are characters for
3045 which isalpha() is true but neither isupper() nor islower() are true. In
3046 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3047 and ordfeminine) are like this. This affected the treatment of \w and \W
3048 when they appeared in character classes, but not when they appeared outside
3049 a character class. The bit map for "word" characters is now created
3050 separately from the results of isalnum() instead of just taking it from the
3051 upper, lower, and digit maps. (Plus the underscore character, of course.)
3052
3053 5. The above bug also affected the handling of POSIX character classes such as
3054 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3055 permanent tables. Instead, the bit maps for such a class were previously
3056 created as the appropriate unions of the upper, lower, and digit bitmaps.
3057 Now they are created by subtraction from the [[:word:]] class, which has
3058 its own bitmap.
3059
3060 6. The [[:blank:]] character class matches horizontal, but not vertical space.
3061 It is created by subtracting the vertical space characters (\x09, \x0a,
3062 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3063 subtraction was done in the overall bitmap for a character class, meaning
3064 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3065 be recognized. This bug has been fixed.
3066
3067 7. Patches from the folks at Google:
3068
3069 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3070 real life, but is still worth protecting against".
3071
3072 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3073 regular expressions".
3074
3075 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3076 have it.
3077
3078 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3079 "configure" and the latter not, in order to fix a problem somebody had
3080 with compiling the Arg class on HP-UX.
3081
3082 (e) Improve the error-handling of the C++ wrapper a little bit.
3083
3084 (f) New tests for checking recursion limiting.
3085
3086 8. The pcre_memmove() function, which is used only if the environment does not
3087 have a standard memmove() function (and is therefore rarely compiled),
3088 contained two bugs: (a) use of int instead of size_t, and (b) it was not
3089 returning a result (though PCRE never actually uses the result).
3090
3091 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3092 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3093 returned instead of calling malloc() with an overflowing number that would
3094 most likely cause subsequent chaos.
3095
3096 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3097
3098 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3099 with this option is matched, the nmatch and pmatch options of regexec() are
3100 ignored.
3101
3102 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3103 provided in case anyone wants to the the POSIX interface with UTF-8
3104 strings.
3105
3106 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3107 C++ linking (needed for some HP-UX environments).
3108
3109 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3110 (unused parameter) and in the pcre_printint() function (omitted "default"
3111 switch label when the default is to do nothing).
3112
3113 15. Added some code to make it possible, when PCRE is compiled as a C++
3114 library, to replace subject pointers for pcre_exec() with a smart pointer
3115 class, thus making it possible to process discontinuous strings.
3116
3117 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3118 much the same function. They were added by different people who were trying
3119 to make PCRE easy to compile on non-Unix systems. It has been suggested
3120 that PCRE_EXPORT be abolished now that there is more automatic apparatus
3121 for compiling on Windows systems. I have therefore replaced it with
3122 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3123 defaults to "extern" for C or "extern C" for C++, which works fine on
3124 Unix-like systems. It is now possible to override the value of PCRE_DATA_
3125 SCOPE with something explicit in config.h. In addition:
3126
3127 (a) pcreposix.h still had just "extern" instead of either of these macros;
3128 I have replaced it with PCRE_DATA_SCOPE.
3129
3130 (b) Functions such as _pcre_xclass(), which are internal to the library,
3131 but external in the C sense, all had PCRE_EXPORT in their definitions.
3132 This is apparently wrong for the Windows case, so I have removed it.
3133 (It makes no difference on Unix-like systems.)
3134
3135 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3136 of recursive calls to match(). This is different to MATCH_LIMIT because
3137 that limits the total number of calls to match(), not all of which increase
3138 the depth of recursion. Limiting the recursion depth limits the amount of
3139 stack (or heap if NO_RECURSE is set) that is used. The default can be set
3140 when PCRE is compiled, and changed at run time. A patch from Google adds
3141 this functionality to the C++ interface.
3142
3143 18. Changes to the handling of Unicode character properties:
3144
3145 (a) Updated the table to Unicode 4.1.0.
3146
3147 (b) Recognize characters that are not in the table as "Cn" (undefined).
3148
3149 (c) I revised the way the table is implemented to a much improved format
3150 which includes recognition of ranges. It now supports the ranges that
3151 are defined in UnicodeData.txt, and it also amalgamates other
3152 characters into ranges. This has reduced the number of entries in the
3153 table from around 16,000 to around 3,000, thus reducing its size
3154 considerably. I realized I did not need to use a tree structure after
3155 all - a binary chop search is just as efficient. Having reduced the
3156 number of entries, I extended their size from 6 bytes to 8 bytes to
3157 allow for more data.
3158
3159 (d) Added support for Unicode script names via properties such as \p{Han}.
3160
3161 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3162 matching that character.
3163
3164 20. When matching a repeated Unicode property with a minimum greater than zero,
3165 (for example \pL{2,}), PCRE could look past the end of the subject if it
3166 reached it while seeking the minimum number of characters. This could
3167 happen only if some of the characters were more than one byte long, because
3168 there is a check for at least the minimum number of bytes.
3169
3170 21. Refactored the implementation of \p and \P so as to be more general, to
3171 allow for more different types of property in future. This has changed the
3172 compiled form incompatibly. Anybody with saved compiled patterns that use
3173 \p or \P will have to recompile them.
3174
3175 22. Added "Any" and "L&" to the supported property types.
3176
3177 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3178 but give a compile time error if the value is greater than 0xff.
3179
3180 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3181 accidentally not being installed or uninstalled.
3182
3183 25. The pcre.h file was built from pcre.h.in, but the only changes that were
3184 made were to insert the current release number. This seemed silly, because
3185 it made things harder for people building PCRE on systems that don't run
3186 "configure". I have turned pcre.h into a distributed file, no longer built
3187 by "configure", with the version identification directly included. There is
3188 no longer a pcre.h.in file.
3189
3190 However, this change necessitated a change to the pcre-config script as
3191 well. It is built from pcre-config.in, and one of the substitutions was the
3192 release number. I have updated configure.ac so that ./configure now finds
3193 the release number by grepping pcre.h.
3194
3195 26. Added the ability to run the tests under valgrind.
3196
3197
3198 Version 6.4 05-Sep-05
3199 ---------------------
3200
3201 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3202 "--" to be printed when multiple files were scanned, even when none of the
3203 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3204 consider it to be a bug, and have restored the previous behaviour.
3205
3206 2. A couple of code tidies to get rid of compiler warnings.
3207
3208 3. The pcretest program used to cheat by referring to symbols in the library
3209 whose names begin with _pcre_. These are internal symbols that are not
3210 really supposed to be visible externally, and in some environments it is
3211 possible to suppress them. The cheating is now confined to including
3212 certain files from the library's source, which is a bit cleaner.
3213
3214 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3215 file's purpose clearer.
3216
3217 5. Reorganized pcre_ucp_findchar().
3218
3219
3220 Version 6.3 15-Aug-05
3221 ---------------------
3222
3223 1. The file libpcre.pc.in did not have general read permission in the tarball.
3224
3225 2. There were some problems when building without C++ support:
3226
3227 (a) If C++ support was not built, "make install" and "make test" still
3228 tried to test it.
3229
3230 (b) There were problems when the value of CXX was explicitly set. Some
3231 changes have been made to try to fix these, and ...
3232
3233 (c) --disable-cpp can now be used to explicitly disable C++ support.
3234
3235 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3236 backslash in a target when C++ was disabled. This confuses some
3237 versions of "make", apparently. Using an intermediate variable solves
3238 this. (Same for CPP_LOBJ.)
3239
3240 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3241 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3242 necessary on certain architectures.
3243
3244 4. Added a setting of -export-symbols-regex to the link command to remove
3245 those symbols that are exported in the C sense, but actually are local
3246 within the library, and not documented. Their names all begin with
3247 "_pcre_". This is not a perfect job, because (a) we have to except some
3248 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3249 available (and never for static libraries). I have made a note to try to
3250 find a way round (a) in the future.
3251
3252
3253 Version 6.2 01-Aug-05
3254 ---------------------
3255
3256 1. There was no test for integer overflow of quantifier values. A construction
3257 such as {1111111111111111} would give undefined results. What is worse, if
3258 a minimum quantifier for a parenthesized subpattern overflowed and became
3259 negative, the calculation of the memory size went wrong. This could have
3260 led to memory overwriting.
3261
3262 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3263
3264 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3265 operating environments where this matters.
3266
3267 4. Applied Giuseppe Maxia's patch to add additional features for controlling
3268 PCRE options from within the C++ wrapper.
3269
3270 5. Named capturing subpatterns were not being correctly counted when a pattern
3271 was compiled. This caused two problems: (a) If there were more than 100
3272 such subpatterns, the calculation of the memory needed for the whole
3273 compiled pattern went wrong, leading to an overflow error. (b) Numerical
3274 back references of the form \12, where the number was greater than 9, were
3275 not recognized as back references, even though there were sufficient
3276 previous subpatterns.
3277
3278 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3279 versions of gcc, e.g. 2.95.4.
3280
3281
3282 Version 6.1 21-Jun-05
3283 ---------------------
3284
3285 1. There was one reference to the variable "posix" in pcretest.c that was not
3286 surrounded by "#if !defined NOPOSIX".
3287
3288 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3289 the cross-check on the old pcre_info() function, for the benefit of the
3290 cut-down version of PCRE that is currently imported into Exim.
3291
3292 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3293 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3294 patterns that start (?i) but I don't think that matters. The use of (?i) is
3295 just an example; this all applies to the other options as well.
3296
3297 4. Since libtool seems to echo the compile commands it is issuing, the output
3298 from "make" can be reduced a bit by putting "@" in front of each libtool
3299 compile command.
3300
3301 5. Patch from the folks at Google for configure.in to be a bit more thorough
3302 in checking for a suitable C++ installation before trying to compile the
3303 C++ stuff. This should fix a reported problem when a compiler was present,
3304 but no suitable headers.
3305
3306 6. The man pages all had just "PCRE" as their title. I have changed them to
3307 be the relevant file name. I have also arranged that these names are
3308 retained in the file doc/pcre.txt, which is a concatenation in text format
3309 of all the man pages except the little individual ones for each function.
3310
3311 7. The NON-UNIX-USE file had not been updated for the different set of source
3312 files that come with release 6. I also added a few comments about the C++
3313 wrapper.
3314
3315
3316 Version 6.0 07-Jun-05
3317 ---------------------
3318
3319 1. Some minor internal re-organization to help with my DFA experiments.
3320
3321 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3322 didn't matter for the library itself when fully configured, but did matter
3323 when compiling without UCP support, or within Exim, where the ucp files are
3324 not imported.
3325
3326 3. Refactoring of the library code to split up the various functions into
3327 different source modules. The addition of the new DFA matching code (see
3328 below) to a single monolithic source would have made it really too
3329 unwieldy, quite apart from causing all the code to be include in a
3330 statically linked application, when only some functions are used. This is
3331 relevant even without the DFA addition now that patterns can be compiled in
3332 one application and matched in another.
3333
3334 The downside of splitting up is that there have to be some external
3335 functions and data tables that are used internally in different modules of
3336 the library but which are not part of the API. These have all had their
3337 names changed to start with "_pcre_" so that they are unlikely to clash
3338 with other external names.
3339
3340 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3341 a different (DFA) algorithm. Although it is slower than the original
3342 function, it does have some advantages for certain types of matching
3343 problem.
3344
3345 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3346 including restarting after a partial match.
3347
3348 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3349 defined when compiling for Windows was sent to me. I have put it into the
3350 code, though I have no means of testing or verifying it.
3351
3352 7. Added the pcre_refcount() auxiliary function.
3353
3354 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3355 match before or at the first newline in the subject string. In pcretest,
3356 the /f option on a pattern can be used to set this.
3357
3358 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3359 would behave wrongly. This has been present in PCRE since release 4.0.
3360
3361 10. A number of changes to the pcregrep command:
3362
3363 (a) Refactored how -x works; insert ^(...)$ instead of setting
3364 PCRE_ANCHORED and checking the length, in preparation for adding
3365 something similar for -w.
3366
3367 (b) Added the -w (match as a word) option.
3368
3369 (c) Refactored the way lines are read and buffered so as to have more
3370 than one at a time available.
3371
3372 (d) Implemented a pcregrep test script.
3373
3374 (e) Added the -M (multiline match) option. This allows patterns to match
3375 over several lines of the subject. The buffering ensures that at least
3376 8K, or the rest of the document (whichever is the shorter) is available
3377 for matching (and similarly the previous 8K for lookbehind assertions).
3378
3379 (f) Changed the --help output so that it now says
3380
3381 -w, --word-regex(p)
3382
3383 instead of two lines, one with "regex" and the other with "regexp"
3384 because that confused at least one person since the short forms are the
3385 same. (This required a bit of code, as the output is generated
3386 automatically from a table. It wasn't just a text change.)
3387
3388 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3389 option but starts with a hyphen. Could be a pattern or a path name
3390 starting with a hyphen, for instance.
3391
3392 (h) "-" can be given as a file name to represent stdin.
3393
3394 (i) When file names are being printed, "(standard input)" is used for
3395 the standard input, for compatibility with GNU grep. Previously
3396 "<stdin>" was used.
3397
3398 (j) The option --label=xxx can be used to supply a name to be used for
3399 stdin when file names are being printed. There is no short form.
3400
3401 (k) Re-factored the options decoding logic because we are going to add
3402 two more options that take data. Such options can now be given in four
3403 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3404
3405 (l) Added the -A, -B, and -C options for requesting that lines of context
3406 around matches be printed.
3407
3408 (m) Added the -L option to print the names of files that do not contain
3409 any matching lines, that is, the complement of -l.
3410
3411 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3412 continue to scan other files.
3413
3414 (o) The -s option was incorrectly implemented. For compatibility with other
3415 greps, it now suppresses the error message for a non-existent or non-
3416 accessible file (but not the return code). There is a new option called
3417 -q that suppresses the output of matching lines, which was what -s was
3418 previously doing.
3419
3420 (p) Added --include and --exclude options to specify files for inclusion
3421 and exclusion when recursing.
3422
3423 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3424 Hopefully, it now does.
3425
3426 12. Missing cast in pcre_study().
3427
3428 13. Added an "uninstall" target to the makefile.
3429
3430 14. Replaced "extern" in the function prototypes in Makefile.in with
3431 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3432 world, but is set differently for Windows.
3433
3434 15. Added a second compiling function called pcre_compile2(). The only
3435 difference is that it has an extra argument, which is a pointer to an
3436 integer error code. When there is a compile-time failure, this is set
3437 non-zero, in addition to the error test pointer being set to point to an
3438 error message. The new argument may be NULL if no error number is required
3439 (but then you may as well call pcre_compile(), which is now just a
3440 wrapper). This facility is provided because some applications need a
3441 numeric error indication, but it has also enabled me to tidy up the way
3442 compile-time errors are handled in the POSIX wrapper.
3443
3444 16. Added VPATH=.libs to the makefile; this should help when building with one
3445 prefix path and installing with another. (Or so I'm told by someone who
3446 knows more about this stuff than I do.)
3447
3448 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3449 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3450 match everything, including newlines. This is not POSIX-compatible, but
3451 somebody wanted the feature. From pcretest it can be activated by using
3452 both the P and the s flags.
3453
3454 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3455
3456 19. libpcre.pc was being incorrectly installed as executable.
3457
3458 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3459 it now also looks for '\r' so that it will work unmodified on Windows.
3460
3461 21. Added Google's contributed C++ wrapper to the distribution.
3462
3463 22. Added some untidy missing memory free() calls in pcretest, to keep
3464 Electric Fence happy when testing.
3465
3466
3467
3468 Version 5.0 13-Sep-04
3469 ---------------------
3470
3471 1. Internal change: literal characters are no longer packed up into items
3472 containing multiple characters in a single byte-string. Each character
3473 is now matched using a separate opcode. However, there may be more than one
3474 byte in the character in UTF-8 mode.
3475
3476 2. The pcre_callout_block structure has two new fields: pattern_position and
3477 next_item_length. These contain the offset in the pattern to the next match
3478 item, and its length, respectively.
3479
3480 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3481 insertion of callouts before each pattern item. Added the /C option to
3482 pcretest to make use of this.
3483
3484 4. On the advice of a Windows user, the lines
3485
3486 #if defined(_WIN32) || defined(WIN32)
3487 _setmode( _fileno( stdout ), 0x8000 );
3488 #endif /* defined(_WIN32) || defined(WIN32) */
3489
3490 have been added to the source of pcretest. This apparently does useful
3491 magic in relation to line terminators.
3492
3493 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3494 for the benefit of those environments where the "b" makes a difference.
3495
3496 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3497 to know about it. I have put a hack into configure.in that adds in code
3498 to set GCC=yes if CC=icc. This seems to end up at a point in the
3499 generated configure script that is early enough to affect the setting of
3500 compiler options, which is what is needed, but I have no means of testing
3501 whether it really works. (The user who reported this had patched the
3502 generated configure script, which of course I cannot do.)
3503
3504 LATER: After change 22 below (new libtool files), the configure script
3505 seems to know about icc (and also ecc). Therefore, I have commented out
3506 this hack in configure.in.
3507
3508 7. Added support for pkg-config (2 patches were sent in).
3509
3510 8. Negated POSIX character classes that used a combination of internal tables
3511 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3512 [[:^ascii]]. Typically, they would match almost any characters. The other
3513 POSIX classes were not broken in this way.
3514
3515 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3516 to find the match, as PCRE was deluded into thinking that the match had to
3517 start at the start point or following a newline. The same bug applied to
3518 patterns with negative forward assertions or any backward assertions
3519 preceding ".*" at the start, unless the pattern required a fixed first
3520 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3521
3522 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3523 starting at the last subject character, bytes beyond the end of the subject
3524 string were read.
3525
3526 11. Renamed the variable "class" as "classbits" to make life easier for C++
3527 users. (Previously there was a macro definition, but it apparently wasn't
3528 enough.)
3529
3530 12. Added the new field "tables" to the extra data so that tables can be passed
3531 in at exec time, or the internal tables can be re-selected. This allows
3532 a compiled regex to be saved and re-used at a later time by a different
3533 program that might have everything at different addresses.
3534
3535 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3536 -R library as well as a -L library.
3537
3538 14. The debugging options of pcretest (-d on the command line or D on a
3539 pattern) showed incorrect output for anything following an extended class
3540 that contained multibyte characters and which was followed by a quantifier.
3541
3542 15. Added optional support for general category Unicode character properties
3543 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3544 support. It adds about 90K to the size of the library. The meanings of the
3545 inbuilt class escapes such as \d and \s have NOT been changed.
3546
3547 16. Updated pcredemo.c to include calls to free() to release the memory for the
3548 compiled pattern.
3549
3550 17. The generated file chartables.c was being created in the source directory
3551 instead of in the building directory. This caused the build to fail if the
3552 source directory was different from the building directory, and was
3553 read-only.
3554
3555 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3556 file. No doubt somebody will tell me if they don't make sense... Also added
3557 Dan Mooney's comments about building on OpenVMS.
3558
3559 19. Added support for partial matching via the PCRE_PARTIAL option for
3560 pcre_exec() and the \P data escape in pcretest.
3561
3562 20. Extended pcretest with 3 new pattern features:
3563
3564 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3565 write the compiled pattern to the file whose name is "rest-of-line".
3566 This is a straight binary dump of the data, with the saved pointer to
3567 the character tables forced to be NULL. The study data, if any, is
3568 written too. After writing, pcretest reads a new pattern.
3569
3570 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3571 compiled pattern from the given file. There must not be any
3572 occurrences of "<" in the file name (pretty unlikely); if there are,
3573 pcretest will instead treat the initial "<" as a pattern delimiter.
3574 After reading in the pattern, pcretest goes on to read data lines as
3575 usual.
3576
3577 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3578 and 16-bit fields in a compiled pattern, to simulate a pattern that
3579 was compiled on a host of opposite endianness.
3580
3581 21. The pcre-exec() function can now cope with patterns that were compiled on
3582 hosts of opposite endianness, with this restriction:
3583
3584 As for any compiled expression that is saved and used later, the tables
3585 pointer field cannot be preserved; the extra_data field in the arguments
3586 to pcre_exec() should be used to pass in a tables address if a value
3587 other than the default internal tables were used at compile time.
3588
3589 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3590 now diagnosed as an error. Previously, most of the time, a negative number
3591 would have been treated as zero, but if in addition "ovector" was passed as
3592 NULL, a crash could occur.
3593
3594 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3595 new versions from the libtool 1.5 distribution (the last one is a copy of
3596 a file called libtool.m4). This seems to have fixed the need to patch
3597 "configure" to support Darwin 1.3 (which I used to do). However, I still
3598 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3599 workstation).
3600
3601 24. Changed the PCRE licence to be the more standard "BSD" licence.
3602
3603
3604 Version 4.5 01-Dec-03
3605 ---------------------
3606
3607 1. There has been some re-arrangement of the code for the match() function so
3608 that it can be compiled in a version that does not call itself recursively.
3609 Instead, it keeps those local variables that need separate instances for
3610 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3611 needs to "recurse". Keeping track of where control must go is done by means
3612 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3613 hide most of the details from the main code, and operates only if
3614 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3615 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3616 operating.
3617
3618 To make it easier for callers to provide specially tailored get/free
3619 functions for this usage, two new functions, pcre_stack_malloc, and
3620 pcre_stack_free, are used. They are always called in strict stacking order,
3621 and the size of block requested is always the same.
3622
3623 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3624 PCRE has been compiled to use the stack or the heap for recursion. The
3625 -C option of pcretest uses this to show which version is compiled.
3626
3627 A new data escape \S, is added to pcretest; it causes the amounts of store
3628 obtained and freed by both kinds of malloc/free at match time to be added
3629 to the output.
3630
3631 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3632 what's available on my current Linux desktop machine.
3633
3634 3. When matching a UTF-8 string, the test for a valid string at the start has
3635 been extended. If start_offset is not zero, PCRE now checks that it points
3636 to a byte that is the start of a UTF-8 character. If not, it returns
3637 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3638 this is necessary because there may be backward assertions in the pattern.
3639 When matching the same subject several times, it may save resources to use
3640 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3641
3642 4. The code for checking the validity of UTF-8 strings has been tightened so
3643 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3644 containing "overlong sequences".
3645
3646 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3647 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3648 should have been "&", but it just so happened that all the cases this let
3649 through by mistake were picked up later in the function.
3650
3651 6. I had used a variable called "isblank" - this is a C99 function, causing
3652 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3653
3654 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3655 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3656 so the version is shown; (c) stop "make test" echoing "./RunTest".
3657
3658 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3659
3660 9. The prototype for memmove() for systems that don't have it was using
3661 size_t, but the inclusion of the header that defines size_t was later. I've
3662 moved the #includes for the C headers earlier to avoid this.
3663
3664 10. Added some adjustments to the code to make it easier to compiler on certain
3665 special systems:
3666
3667 (a) Some "const" qualifiers were missing.
3668 (b) Added the macro EXPORT before all exported functions; by default this
3669 is defined to be empty.
3670 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3671 that it reads its output file name as an argument instead of writing
3672 to the standard output and assuming this can be redirected.
3673
3674 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3675 class containing characters with values greater than 255, PCRE compilation
3676 went into a loop.
3677
3678 12. A recursive reference to a subpattern that was within another subpattern
3679 that had a minimum quantifier of zero caused PCRE to crash. For example,
3680 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3681 recursion. If the recursively-called subpattern itself had a zero repeat,
3682 that was OK.
3683
3684 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3685 buffer into which it was copied (for escape processing) was still set at
3686 1024, so long lines caused crashes.
3687
3688 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3689 "internal error: code overflow...". This applied to any character class
3690 that was followed by a possessive quantifier.
3691
3692 15. Modified the Makefile to add libpcre.la as a prerequisite for
3693 libpcreposix.la because I was told this is needed for a parallel build to
3694 work.
3695
3696 16. If a pattern that contained .* following optional items at the start was
3697 studied, the wrong optimizing data was generated, leading to matching
3698 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3699 matching string must start with a or b or c. The correct conclusion for
3700 this pattern is that a match can start with any character.
3701
3702
3703 Version 4.4 13-Aug-03
3704 ---------------------
3705
3706 1. In UTF-8 mode, a character class containing characters with values between
3707 127 and 255 was not handled correctly if the compiled pattern was studied.
3708 In fixing this, I have also improved the studying algorithm for such
3709 classes (slightly).
3710
3711 2. Three internal functions had redundant arguments passed to them. Removal
3712 might give a very teeny performance improvement.
3713
3714 3. Documentation bug: the value of the capture_top field in a callout is *one
3715 more than* the number of the hightest numbered captured substring.
3716
3717 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3718 in incorrectly linking with a previously installed version. They now link
3719 explicitly with libpcre.la.
3720
3721 5. configure.in no longer needs to recognize Cygwin specially.
3722
3723 6. A problem in pcre.in for Windows platforms is fixed.
3724
3725 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3726 pcretest, it used to include the size of the study block as part of its
3727 output. Unfortunately, the structure contains a field that has a different
3728 size on different hardware architectures. This meant that the tests that
3729 showed this size failed. As the block is currently always of a fixed size,
3730 this information isn't actually particularly useful in pcretest output, so
3731 I have just removed it.
3732
3733 8. Three pre-processor statements accidentally did not start in column 1.
3734 Sadly, there are *still* compilers around that complain, even though
3735 standard C has not required this for well over a decade. Sigh.
3736
3737 9. In pcretest, the code for checking callouts passed small integers in the
3738 callout_data field, which is a void * field. However, some picky compilers
3739 complained about the casts involved for this on 64-bit systems. Now
3740 pcretest passes the address of the small integer instead, which should get
3741 rid of the warnings.
3742
3743 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3744 both compile and run time, and gives an error if an invalid UTF-8 sequence
3745 is found. There is a option for disabling this check in cases where the
3746 string is known to be correct and/or the maximum performance is wanted.
3747
3748 11. In response to a bug report, I changed one line in Makefile.in from
3749
3750 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3751 to
3752 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3753
3754 to look similar to other lines, but I have no way of telling whether this
3755 is the right thing to do, as I do not use Windows. No doubt I'll get told
3756 if it's wrong...
3757
3758
3759 Version 4.3 21-May-03
3760 ---------------------
3761
3762 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3763 Makefile.
3764
3765 2. Some refactoring to improve the quality of the code:
3766
3767 (i) The utf8_table... variables are now declared "const".
3768
3769 (ii) The code for \cx, which used the "case flipping" table to upper case
3770 lower case letters, now just substracts 32. This is ASCII-specific,
3771 but the whole concept of \cx is ASCII-specific, so it seems
3772 reasonable.
3773
3774 (iii) PCRE was using its character types table to recognize decimal and
3775 hexadecimal digits in the pattern. This is silly, because it handles
3776 only 0-9, a-f, and A-F, but the character types table is locale-
3777 specific, which means strange things might happen. A private
3778 table is now used for this - though it costs 256 bytes, a table is
3779 much faster than multiple explicit tests. Of course, the standard
3780 character types table is still used for matching digits in subject
3781 strings against \d.
3782
3783 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3784 ending in _t are). So I've renamed it as ESC_tee.
3785
3786 3. The first argument for regexec() in the POSIX wrapper should have been
3787 defined as "const".
3788
3789 4. Changed pcretest to use malloc() for its buffers so that they can be
3790 Electric Fenced for debugging.
3791
3792 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3793 to read one or more bytes before the start of the subject string. Often this
3794 had no effect on PCRE's behaviour, but in some circumstances it could
3795 provoke a segmentation fault.
3796
3797 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3798 to try to read one or more bytes before the start of the subject string.
3799
3800 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3801 UTF-8 support could misbehave in various ways if the subject string
3802 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3803 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3804 back over UTF-8 characters.)
3805
3806
3807 Version 4.2 14-Apr-03
3808 ---------------------
3809
3810 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3811
3812 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3813 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3814 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3815 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3816 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3817 and BUILD_EXEEXT
3818 Note: automatic setting of the BUILD variables is not yet working
3819 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3820 compile-time but not at link-time
3821 [LINK]: use for linking executables only
3822 make different versions for Windows and non-Windows
3823 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3824 libraries
3825 [LINK_FOR_BUILD]: new variable
3826 [OBJEXT]: use throughout
3827 [EXEEXT]: use throughout
3828 <winshared>: new target
3829 <wininstall>: new target
3830 <dftables.o>: use native compiler
3831 <dftables>: use native linker
3832 <install>: handle Windows platform correctly
3833 <clean>: ditto
3834 <check>: ditto
3835 copy DLL to top builddir before testing
3836
3837 As part of these changes, -no-undefined was removed again. This was reported
3838 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3839 in any case.
3840
3841 3. Some tidies to get rid of compiler warnings:
3842
3843 . In the match_data structure, match_limit was an unsigned long int, whereas
3844 match_call_count was an int. I've made them both unsigned long ints.
3845
3846 . In pcretest the fact that a const uschar * doesn't automatically cast to
3847 a void * provoked a warning.
3848
3849 . Turning on some more compiler warnings threw up some "shadow" variables
3850 and a few more missing casts.
3851
3852 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3853 option, a class that contained a single character with a value between 128
3854 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3855
3856 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3857 option, a class that contained several characters, but with at least one
3858 whose value was between 128 and 255 caused PCRE to crash.
3859
3860
3861 Version 4.1 12-Mar-03
3862 ---------------------
3863
3864 1. Compiling with gcc -pedantic found a couple of places where casts were
3865 needed, and a string in dftables.c that was longer than standard compilers are
3866 required to support.
3867
3868 2. Compiling with Sun's compiler found a few more places where the code could
3869 be tidied up in order to avoid warnings.
3870
3871 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3872 first of these names is deprecated in the latest Autoconf in favour of the name
3873 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3874 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3875 analogy I have changed it to CFLAGS_FOR_BUILD.
3876
3877 4. Added -no-undefined to the linking command in the Makefile, because this is
3878 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3879 linking step for the pcreposix library.
3880
3881 5. PCRE was failing to diagnose the case of two named groups with the same
3882 name.
3883
3884 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3885 literal character that is needed in the subject for a match, and scans along to
3886 ensure that it is present before embarking on the full matching process. This
3887 saves time in cases of nested unlimited repeats that are never going to match.
3888 Problem: the scan can take a lot of time if the subject is very long (e.g.
3889 megabytes), thus penalizing straightforward matches. It is now done only if the
3890 amount of subject to be scanned is less than 1000 bytes.
3891
3892 7. A lesser problem with the same optimization is that it was recording the
3893 first character of an anchored pattern as "needed", thus provoking a search
3894 right along the subject, even when the first match of the pattern was going to
3895 fail. The "needed" character is now not set for anchored patterns, unless it
3896 follows something in the pattern that is of non-fixed length. Thus, it still
3897 fulfils its original purpose of finding quick non-matches in cases of nested
3898 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3899
3900
3901 Version 4.0 17-Feb-03
3902 ---------------------
3903
3904 1. If a comment in an extended regex that started immediately after a meta-item
3905 extended to the end of string, PCRE compiled incorrect data. This could lead to
3906 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3907
3908 2. Moved to autoconf 2.53 and libtool 1.4.2.
3909
3910 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3911 the special perltest8 script is no longer needed - all the tests can be run
3912 from a single perltest script.
3913
3914 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3915 by \s. It has now been removed in PCRE. This means it isn't recognized as
3916 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3917 class [:space:] *does* include VT, thereby creating a mess.
3918
3919 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3920 space and tab.
3921
3922 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3923 its new features into the main test script, reducing the number of scripts.
3924
3925 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3926 were backward compatible, and made the (?i) apply to the whole pattern, as if
3927 /i were given. Now it behaves more logically, and applies the option setting
3928 only to what follows. PCRE has been changed to follow suit. However, if it
3929 finds options settings right at the start of the pattern, it extracts them into
3930 the global options, as before. Thus, they show up in the info data.
3931
3932 8. Added support for the \Q...\E escape sequence. Characters in between are
3933 treated as literals. This is slightly different from Perl in that $ and @ are
3934 also handled as literals inside the quotes. In Perl, they will cause variable
3935 interpolation. Note the following examples:
3936
3937 Pattern PCRE matches Perl matches
3938
3939 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3940 \Qabc\$xyz\E abc\$xyz abc\$xyz
3941 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3942
3943 For compatibility with Perl, \Q...\E sequences are recognized inside character
3944 classes as well as outside them.
3945
3946 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3947 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3948 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3949 signed/unsigned warnings.
3950
3951 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3952 option for pcretest, so I've replaced it by a simple function that does just
3953 that job.
3954
3955 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3956 "pcregrep -".
3957
3958 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3959 Java package. This provides some syntactic sugar for simple cases of what my
3960 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3961 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3962 item, you can use this simplified notation. Note that only makes sense with
3963 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3964 greediness, whatever the setting of the PCRE_UNGREEDY option.
3965
3966 13. A change of greediness default within a pattern was not taking effect at
3967 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3968 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3969 was abstracted outside.
3970
3971 14. PCRE now supports the \G assertion. It is true when the current matching
3972 position is at the start point of the match. This differs from \A when the
3973 starting offset is non-zero. Used with the /g option of pcretest (or similar
3974 code), it works in the same way as it does for Perl's /g option. If all
3975 alternatives of a regex begin with \G, the expression is anchored to the start
3976 match position, and the "anchored" flag is set in the compiled expression.
3977
3978 15. Some bugs concerning the handling of certain option changes within patterns
3979 have been fixed. These applied to options other than (?ims). For example,
3980 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3981 been the other way round. Some of this was related to change 7 above.
3982
3983 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3984 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3985 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3986 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3987
3988 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3989 mode. Unlike ".", it always matches newline, whatever the setting of
3990 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3991 assertions. Perl allows it, but it doesn't (in general) work because it can't
3992 calculate the length of the lookbehind. At least, that's the case for Perl
3993 5.8.0 - I've been told they are going to document that it doesn't work in
3994 future.
3995
3996 18. Added an error diagnosis for escapes that PCRE does not support: these are
3997 \L, \l, \N, \P, \p, \U, \u, and \X.
3998
3999 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
4000 reading past the end of the pattern in cases such as /[abcd/.
4001
4002 20. PCRE was getting more memory than necessary for patterns with classes that
4003 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
4004
4005 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
4006 compiling PCRE for use with Virtual Pascal.
4007
4008 22. Small fix to the Makefile to make it work properly if the build is done
4009 outside the source tree.
4010
4011 23. Added a new extension: a condition to go with recursion. If a conditional
4012 subpattern starts with (?(R) the "true" branch is used if recursion has
4013 happened, whereas the "false" branch is used only at the top level.
4014
4015 24. When there was a very long string of literal characters (over 255 bytes
4016 without UTF support, over 250 bytes with UTF support), the computation of how
4017 much memory was required could be incorrect, leading to segfaults or other
4018 strange effects.
4019
4020 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
4021 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
4022 there was a subsequent back reference to those brackets. This meant that, for
4023 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
4024 possible to check for precisely this case. All we can do is abandon the
4025 optimization if .* occurs inside capturing brackets when there are any back
4026 references whatsoever. (See below for a better fix that came later.)
4027
4028 26. The handling of the optimization for finding the first character of a
4029 non-anchored pattern, and for finding a character that is required later in the
4030 match were failing in some cases. This didn't break the matching; it just
4031 failed to optimize when it could. The way this is done has been re-implemented.
4032
4033 27. Fixed typo in error message for invalid (?R item (it said "(?p").
4034
4035 28. Added a new feature that provides some of the functionality that Perl
4036 provides with (?{...}). The facility is termed a "callout". The way it is done
4037 in PCRE is for the caller to provide an optional function, by setting
4038 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
4039 global variable. By default it is unset, which disables all calling out. To get
4040 the function called, the regex must include (?C) at appropriate points. This
4041 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
4042 This provides a means of identifying different callout points. When PCRE
4043 reaches such a point in the regex, if pcre_callout has been set, the external
4044 function is called. It is provided with data in a structure called
4045 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
4046 matching continues; if it returns a non-zero value, the match at the current
4047 point fails. However, backtracking will occur if possible. [This was changed
4048 later and other features added - see item 49 below.]
4049
4050 29. pcretest is upgraded to test the callout functionality. It provides a
4051 callout function that displays information. By default, it shows the start of
4052 the match and the current position in the text. There are some new data escapes
4053 to vary what happens:
4054
4055 \C+ in addition, show current contents of captured substrings
4056 \C- do not supply a callout function
4057 \C!n return 1 when callout number n is reached
4058 \C!n!m return 1 when callout number n is reached for the mth time
4059
4060 30. If pcregrep was called with the -l option and just a single file name, it
4061 output "<stdin>" if a match was found, instead of the file name.
4062
4063 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
4064 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
4065 pcre_exec(). This saves a malloc/free per call. The default value of
4066 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
4067 when configuring.
4068
4069 32. The default maximum size of a compiled pattern is 64K. There have been a
4070 few cases of people hitting this limit. The code now uses macros to handle the
4071 storing of links as offsets within the compiled pattern. It defaults to 2-byte
4072 links, but this can be changed to 3 or 4 bytes by --with-link-size when
4073 configuring. Tests 2 and 5 work only with 2-byte links because they output
4074 debugging information about compiled patterns.
4075
4076 33. Internal code re-arrangements:
4077
4078 (a) Moved the debugging function for printing out a compiled regex into
4079 its own source file (printint.c) and used #include to pull it into
4080 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
4081 separate copies.
4082
4083 (b) Defined the list of op-code names for debugging as a macro in
4084 internal.h so that it is next to the definition of the opcodes.
4085
4086 (c) Defined a table of op-code lengths for simpler skipping along compiled
4087 code. This is again a macro in internal.h so that it is next to the
4088 definition of the opcodes.
4089
4090 34. Added support for recursive calls to individual subpatterns, along the
4091 lines of Robin Houston's patch (but implemented somewhat differently).
4092
4093 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
4094 allow it to read and process whole directories in Win32. This code was
4095 contributed by Lionel Fourquaux; it has not been tested by me.
4096
4097 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
4098 used to name a group. Names consist of alphanumerics and underscores, and must
4099 be unique. Back references use the syntax (?P=name) and recursive calls use
4100 (?P>name) which is a PCRE extension to the Python extension. Groups still have
4101 numbers. The function pcre_fullinfo() can be used after compilation to extract
4102 a name/number map. There are three relevant calls:
4103
4104 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
4105 PCRE_INFO_NAMECOUNT yields the number of entries
4106 PCRE_INFO_NAMETABLE yields a pointer to the map.
4107
4108 The map is a vector of fixed-size entries. The size of each entry depends on
4109 the length of the longest name used. The first two bytes of each entry are the
4110 group number, most significant byte first. There follows the corresponding
4111 name, zero terminated. The names are in alphabetical order.
4112
4113 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
4114 case instead of 255. Making it the same both with and without UTF-8 support
4115 means that the same test output works with both.
4116
4117 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
4118 calling malloc() with a zero argument.
4119
4120 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
4121 optimization. I've improved things by keeping a bitmap of backreferences with
4122 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
4123 fact referenced, the optimization can be applied. It is unlikely that a
4124 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
4125 the match to follow \n) will appear inside brackets with a number greater than
4126 31, but if it does, any back reference > 31 suppresses the optimization.
4127
4128 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
4129 of disabling numbered capturing parentheses. Any opening parenthesis that is
4130 not followed by ? behaves as if it were followed by ?: but named parentheses
4131 can still be used for capturing (and they will acquire numbers in the usual
4132 way).
4133
4134 41. Redesigned the return codes from the match() function into yes/no/error so
4135 that errors can be passed back from deep inside the nested calls. A malloc
4136 failure while inside a recursive subpattern call now causes the
4137 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
4138
4139 42. It is now possible to set a limit on the number of times the match()
4140 function is called in a call to pcre_exec(). This facility makes it possible to
4141 limit the amount of recursion and backtracking, though not in a directly
4142 obvious way, because the match() function is used in a number of different
4143 circumstances. The count starts from zero for each position in the subject
4144 string (for non-anchored patterns). The default limit is, for compatibility, a
4145 large number, namely 10 000 000. You can change this in two ways:
4146
4147 (a) When configuring PCRE before making, you can use --with-match-limit=n
4148 to set a default value for the compiled library.
4149
4150 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
4151 a different value is set. See 45 below.
4152
4153 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
4154
4155 43. Added a new function pcre_config(int, void *) to enable run-time extraction
4156 of things that can be changed at compile time. The first argument specifies
4157 what is wanted and the second points to where the information is to be placed.
4158 The current list of available information is:
4159
4160 PCRE_CONFIG_UTF8
4161
4162 The output is an integer that is set to one if UTF-8 support is available;
4163 otherwise it is set to zero.
4164
4165 PCRE_CONFIG_NEWLINE
4166
4167 The output is an integer that it set to the value of the code that is used for
4168 newline. It is either LF (10) or CR (13).
4169
4170 PCRE_CONFIG_LINK_SIZE
4171
4172 The output is an integer that contains the number of bytes used for internal
4173 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
4174
4175 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
4176
4177 The output is an integer that contains the threshold above which the POSIX
4178 interface uses malloc() for output vectors. See item 31 above.
4179
4180 PCRE_CONFIG_MATCH_LIMIT
4181
4182 The output is an unsigned integer that contains the default limit of the number
4183 of match() calls in a pcre_exec() execution. See 42 above.
4184
4185 44. pcretest has been upgraded by the addition of the -C option. This causes it
4186 to extract all the available output from the new pcre_config() function, and to
4187 output it. The program then exits immediately.
4188
4189 45. A need has arisen to pass over additional data with calls to pcre_exec() in
4190 order to support additional features. One way would have been to define
4191 pcre_exec2() (for example) with extra arguments, but this would not have been
4192 extensible, and would also have required all calls to the original function to
4193 be mapped to the new one. Instead, I have chosen to extend the mechanism that
4194 is used for passing in "extra" data from pcre_study().
4195
4196 The pcre_extra structure is now exposed and defined in pcre.h. It currently
4197 contains the following fields:
4198
4199 flags a bitmap indicating which of the following fields are set
4200 study_data opaque data from pcre_study()
4201 match_limit a way of specifying a limit on match() calls for a specific
4202 call to pcre_exec()
4203 callout_data data for callouts (see 49 below)
4204
4205 The flag bits are also defined in pcre.h, and are
4206
4207 PCRE_EXTRA_STUDY_DATA
4208 PCRE_EXTRA_MATCH_LIMIT
4209 PCRE_EXTRA_CALLOUT_DATA
4210
4211 The pcre_study() function now returns one of these new pcre_extra blocks, with
4212 the actual study data pointed to by the study_data field, and the
4213 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
4214 before. That is, this change is entirely upwards-compatible and requires no
4215 change to existing code.
4216
4217 If you want to pass in additional data to pcre_exec(), you can either place it
4218 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
4219 block.
4220
4221 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
4222 data string contains the escape sequence \M, pcretest calls pcre_exec() several
4223 times with different match limits, until it finds the minimum value needed for
4224 pcre_exec() to complete. The value is then output. This can be instructive; for
4225 most simple matches the number is quite small, but for pathological cases it
4226 gets very large very quickly.
4227
4228 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
4229 returns the size of the data block pointed to by the study_data field in a
4230 pcre_extra block, that is, the value that was passed as the argument to
4231 pcre_malloc() when PCRE was getting memory in which to place the information
4232 created by pcre_study(). The fourth argument should point to a size_t variable.
4233 pcretest has been extended so that this information is shown after a successful
4234 pcre_study() call when information about the compiled regex is being displayed.
4235
4236 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
4237 because what follows is always an absolute path. (Later: it turns out that this
4238 is more than cosmetic for MinGW, because it doesn't like empty path
4239 components.)
4240
4241 49. Some changes have been made to the callout feature (see 28 above):
4242
4243 (i) A callout function now has three choices for what it returns:
4244
4245 0 => success, carry on matching
4246 > 0 => failure at this point, but backtrack if possible
4247 < 0 => serious error, return this value from pcre_exec()
4248
4249 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
4250 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
4251 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
4252 use by callout functions. It will never be used by PCRE itself.
4253
4254 (ii) The pcre_extra structure (see 45 above) has a void * field called
4255 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
4256 pcre_callout_block structure has a field of the same name. The contents of
4257 the field passed in the pcre_extra structure are passed to the callout
4258 function in the corresponding field in the callout block. This makes it
4259 easier to use the same callout-containing regex from multiple threads. For
4260 testing, the pcretest program has a new data escape
4261
4262 \C*n pass the number n (may be negative) as callout_data
4263
4264 If the callout function in pcretest receives a non-zero value as
4265 callout_data, it returns that value.
4266
4267 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
4268 there were some redundant $(CFLAGS) in commands that are now specified as
4269 $(LINK), which already includes $(CFLAGS).
4270
4271 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
4272 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
4273 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
4274 one-byte characters throughout. Note that case-insensitive matching applies
4275 only to characters whose values are less than 256. PCRE doesn't support the
4276 notion of cases for higher-valued characters.
4277
4278 (i) A character class whose characters are all within 0-255 is handled as
4279 a bit map, and the map is inverted for negative classes. Previously, a
4280 character > 255 always failed to match such a class; however it should
4281 match if the class was a negative one (e.g. [^ab]). This has been fixed.
4282
4283 (ii) A negated character class with a single character < 255 is coded as
4284 "not this character" (OP_NOT). This wasn't working properly when the test
4285 character was multibyte, either singly or repeated.
4286
4287 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
4288 mode, for example: \x{100}{2,3}.
4289
4290 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
4291 singly or repeated) now correctly test multibyte characters. However,
4292 PCRE doesn't recognize any characters with values greater than 255 as
4293 digits, spaces, or word characters. Such characters always match \D, \S,
4294 and \W, and never match \d, \s, or \w.
4295
4296 (v) Classes may now contain characters and character ranges with values
4297 greater than 255. For example: [ab\x{100}-\x{400}].
4298
4299 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
4300 PCRE in UTF-8 mode.
4301
4302 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
4303 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
4304 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
4305 value.)
4306
4307 53. The single man page has become too large. I have therefore split it up into
4308 a number of separate man pages. These also give rise to individual HTML pages;
4309 these are now put in a separate directory, and there is an index.html page that
4310 lists them all. Some hyperlinking between the pages has been installed.
4311
4312 54. Added convenience functions for handling named capturing parentheses.
4313
4314 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4315 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4316 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4317 are faulted.
4318
4319 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4320 calling configure. These values are used when compiling the dftables.c program
4321 which is run to generate the source of the default character tables. They
4322 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4323 you will need to set these values.
4324
4325 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4326
4327
4328 Version 3.9 02-Jan-02
4329 ---------------------
4330
4331 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4332
4333 2. If --disable-static was given, the building process failed when trying to
4334 build pcretest and pcregrep. (For some reason it was using libtool to compile
4335 them, which is not right, as they aren't part of the library.)
4336
4337
4338 Version 3.8 18-Dec-01
4339 ---------------------
4340
4341 1. The experimental UTF-8 code was completely screwed up. It was packing the
4342 bytes in the wrong order. How dumb can you get?
4343
4344
4345 Version 3.7 29-Oct-01
4346 ---------------------
4347
4348 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4349 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4350 this didn't happen under Solaris 8, where I normally test things.
4351
4352 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4353 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4354 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4355 this fix an hour or so after the initial 3.7 release.)
4356
4357
4358 Version 3.6 23-Oct-01
4359 ---------------------
4360
4361 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4362 offsets passed as NULL with zero offset count.
4363
4364 2. The config.guess and config.sub files had not been updated when I moved to
4365 the latest autoconf.
4366
4367
4368 Version 3.5 15-Aug-01
4369 ---------------------
4370
4371 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4372 had been forgotten.
4373
4374 2. By using declared but undefined structures, we can avoid using "void"
4375 definitions in pcre.h while keeping the internal definitions of the structures
4376 private.
4377
4378 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4379 user point of view, this means that both static and shared libraries are built
4380 by default, but this can be individually controlled. More of the work of
4381 handling this static/shared cases is now inside libtool instead of PCRE's make
4382 file.
4383
4384 4. The pcretest utility is now installed along with pcregrep because it is
4385 useful for users (to test regexs) and by doing this, it automatically gets
4386 relinked by libtool. The documentation has been turned into a man page, so
4387 there are now .1, .txt, and .html versions in /doc.
4388
4389 5. Upgrades to pcregrep:
4390 (i) Added long-form option names like gnu grep.
4391 (ii) Added --help to list all options with an explanatory phrase.
4392 (iii) Added -r, --recursive to recurse into sub-directories.
4393 (iv) Added -f, --file to read patterns from a file.
4394
4395 6. pcre_exec() was referring to its "code" argument before testing that
4396 argument for NULL (and giving an error if it was NULL).
4397
4398 7. Upgraded Makefile.in to allow for compiling in a different directory from
4399 the source directory.
4400
4401 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4402 options bits, the pointer it was passed was to an int instead of to an unsigned
4403 long int. This mattered only on 64-bit systems.
4404
4405 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4406 generated) instead of pcre.in, which it its source. Also made the same change
4407 in several of the .c files.
4408
4409 10. A new release of gcc defines printf() as a macro, which broke pcretest
4410 because it had an ifdef in the middle of a string argument for printf(). Fixed
4411 by using separate calls to printf().
4412
4413 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4414 script, to force use of CR or LF instead of \n in the source. On non-Unix
4415 systems, the value can be set in config.h.
4416
4417 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4418 absolute limit. Changed the text of the error message to make this clear, and
4419 likewise updated the man page.
4420
4421 13. The limit of 99 on the number of capturing subpatterns has been removed.
4422 The new limit is 65535, which I hope will not be a "real" limit.
4423
4424
4425 Version 3.4 22-Aug-00
4426 ---------------------
4427
4428 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4429
4430 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4431
4432
4433 Version 3.3 01-Aug-00
4434 ---------------------
4435
4436 1. If an octal character was given, but the value was greater than \377, it
4437 was not getting masked to the least significant bits, as documented. This could
4438 lead to crashes in some systems.
4439
4440 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4441 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4442
4443 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4444 These just pass their arguments on to (pcre_free)(), but they are provided
4445 because some uses of PCRE bind it to non-C systems that can call its functions,
4446 but cannot call free() or pcre_free() directly.
4447
4448 4. Add "make test" as a synonym for "make check". Corrected some comments in
4449 the Makefile.
4450
4451 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4452 Makefile.
4453
4454 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4455 command called pgrep for grepping around the active processes.
4456
4457 7. Added the beginnings of support for UTF-8 character strings.
4458
4459 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4460 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4461 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4462 out for the ar command.)
4463
4464
4465 Version 3.2 12-May-00
4466 ---------------------
4467
4468 This is purely a bug fixing release.
4469
4470 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4471 of ZA. This was just one example of several cases that could provoke this bug,
4472 which was introduced by change 9 of version 2.00. The code for breaking
4473 infinite loops after an iteration that matches an empty string was't working
4474 correctly.
4475
4476 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4477 when matched against abbab (for example). After matching an empty string, it
4478 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4479 caused it to match further down the string than it should.
4480
4481 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4482 was there because it doesn't seem to be needed, and it causes trouble on some
4483 systems, as it is not a Standard C header. It has been removed.
4484
4485 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4486 were reported on the Macintosh. The changes were from
4487
4488 while ((c = *(++ptr)) != 0 && c != '\n');
4489 to
4490 while ((c = *(++ptr)) != 0 && c != '\n') ;
4491
4492 Totally extraordinary, but if that's what it takes...
4493
4494 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4495 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4496 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4497 assumes the way PCRE uses memmove() (always moving upwards).
4498
4499 6. PCRE is being used in one environment where strchr() is not available. There
4500 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4501 faster code anyway.
4502
4503
4504 Version 3.1 09-Feb-00
4505 ---------------------
4506
4507 The only change in this release is the fixing of some bugs in Makefile.in for
4508 the "install" target:
4509
4510 (1) It was failing to install pcreposix.h.
4511
4512 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4513
4514
4515 Version 3.0 01-Feb-00
4516 ---------------------
4517
4518 1. Add support for the /+ modifier to perltest (to output $` like it does in
4519 pcretest).
4520
4521 2. Add support for the /g modifier to perltest.
4522
4523 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4524 matches null strings.
4525
4526 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4527 pattern. Perl treats empty patterns specially - it reuses the most recent
4528 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4529 effect.
4530
4531 5. The POSIX interface was broken in that it was just handing over the POSIX
4532 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4533 required a bigger vector, with some working space on the end. This means that
4534 the POSIX wrapper now has to get and free some memory, and copy the results.
4535
4536 6. Added some simple autoconf support, placing the test data and the
4537 documentation in separate directories, re-organizing some of the
4538 information files, and making it build pcre-config (a GNU standard). Also added
4539 libtool support for building PCRE as a shared library, which is now the
4540 default.
4541
4542 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4543 09 are not valid octal constants. Single digits will be used for minor values
4544 less than 10.
4545
4546 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4547 existing programs that set these in the POSIX interface can use PCRE without
4548 modification.
4549
4550 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4551 return all that pcre_info() returns, plus additional data. The pcre_info()
4552 function is retained for compatibility, but is considered to be obsolete.
4553
4554 10. Added experimental recursion feature (?R) to handle one common case that
4555 Perl 5.6 will be able to do with (?p{...}).
4556
4557 11. Added support for POSIX character classes like [:alpha:], which Perl is
4558 adopting.
4559
4560
4561 Version 2.08 31-Aug-99
4562 ----------------------
4563
4564 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4565 trying to match at the startoffset position, but instead was moving forward to
4566 the next newline as if a previous match had failed.
4567
4568 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4569 and could get into a loop if a null string was matched other than at the start
4570 of the subject.
4571
4572 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4573 be distinguished at compile time, and for completeness also added PCRE_DATE.
4574
4575 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4576 in GnuWin32 environments.
4577
4578
4579 Version 2.07 29-Jul-99
4580 ----------------------
4581
4582 1. The documentation is now supplied in plain text form and HTML as well as in
4583 the form of man page sources.
4584
4585 2. C++ compilers don't like assigning (void *) values to other pointer types.
4586 In particular this affects malloc(). Although there is no problem in Standard
4587 C, I've put in casts to keep C++ compilers happy.
4588
4589 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4590 should be (const char *).
4591
4592 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4593 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4594 However, I haven't made this a standard facility. The documentation doesn't
4595 mention it, and the Makefile doesn't support it.
4596
4597 5. The Makefile now contains an "install" target, with editable destinations at
4598 the top of the file. The pcretest program is not installed.
4599
4600 6. pgrep -V now gives the PCRE version number and date.
4601
4602 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4603 causing the entire string to be ignored, instead of just the last character.
4604
4605 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4606 non-matching string, it can take a very, very long time, even for strings of
4607 quite modest length, because of the nested recursion. PCRE now does better in
4608 some of these cases. It does this by remembering the last required literal
4609 character in the pattern, and pre-searching the subject to ensure it is present
4610 before running the real match. In other words, it applies a heuristic to detect
4611 some types of certain failure quickly, and in the above example, if presented
4612 with a string that has no trailing " it gives "no match" very quickly.
4613
4614 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4615 other alternatives are tried instead.
4616
4617
4618 Version 2.06 09-Jun-99
4619 ----------------------
4620
4621 1. Change pcretest's output for amount of store used to show just the code
4622 space, because the remainder (the data block) varies in size between 32-bit and
4623 64-bit systems.
4624
4625 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4626 start matching at. This allows lookbehinds to work when searching for multiple
4627 occurrences in a string.
4628
4629 3. Added additional options to pcretest for testing multiple occurrences:
4630
4631 /+ outputs the rest of the string that follows a match
4632 /g loops for multiple occurrences, using the new startoffset argument
4633 /G loops for multiple occurrences by passing an incremented pointer
4634
4635 4. PCRE wasn't doing the "first character" optimization for patterns starting
4636 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4637 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4638 the letter 'x'. On long subject strings, this gives a significant speed-up.
4639
4640
4641 Version 2.05 21-Apr-99
4642 ----------------------
4643
4644 1. Changed the type of magic_number from int to long int so that it works
4645 properly on 16-bit systems.
4646
4647 2. Fixed a bug which caused patterns starting with .* not to work correctly
4648 when the subject string contained newline characters. PCRE was assuming
4649 anchoring for such patterns in all cases, which is not correct because .* will
4650 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4651 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4652 must be retried after every newline in the subject.
4653
4654
4655 Version 2.04 18-Feb-99
4656 ----------------------
4657
4658 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4659 computation of the store needed to hold the pattern was incorrect (too large).
4660 If such patterns were nested a few deep, this could multiply and become a real
4661 problem.
4662
4663 2. Added /M option to pcretest to show the memory requirement of a specific
4664 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4665
4666 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4667 compiled in such a way that the backtracking after subsequent failure was
4668 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4669 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4670
4671
4672 Version 2.03 02-Feb-99
4673 ----------------------
4674
4675 1. Fixed typo and small mistake in man page.
4676
4677 2. Added 4th condition (GPL supersedes if conflict) and created separate
4678 LICENCE file containing the conditions.
4679
4680 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4681 Perl, that is the internal \ allows the delimiter to be included in the
4682 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4683 the final delimiter, add \ to the end of the pattern (to test the error).
4684
4685 4. Added the convenience functions for extracting substrings after a successful
4686 match. Updated pcretest to make it able to test these functions.
4687
4688
4689 Version 2.02 14-Jan-99
4690 ----------------------
4691
4692 1. Initialized the working variables associated with each extraction so that
4693 their saving and restoring doesn't refer to uninitialized store.
4694
4695 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4696 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4697 fix the problem.
4698
4699 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4700 calls, and wasn't printing the correct value for compiling calls. Increased the
4701 default value of LOOPREPEAT, and the number of significant figures in the
4702 times.
4703
4704 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4705
4706 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4707 a building problem on Windows NT with a FAT file system.
4708
4709
4710 Version 2.01 21-Oct-98
4711 ----------------------
4712
4713 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4714 to character tables built by pcre_maketables() in the current locale. If NULL
4715 is passed, the default tables are used.
4716
4717
4718 Version 2.00 24-Sep-98
4719 ----------------------
4720
4721 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4722 it any more.
4723
4724 2. Allow quantification of (?>) groups, and make it work correctly.
4725
4726 3. The first character computation wasn't working for (?>) groups.
4727
4728 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4729 end of the subject) and add 5.005's \z, which really does match only at the
4730 very end of the subject.
4731
4732 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4733
4734 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4735 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4736 localized options. All options to pcre_study() were also removed.
4737
4738 7. Add other new features from 5.005:
4739
4740 $(?<= positive lookbehind
4741 $(?<! negative lookbehind
4742 (?imsx-imsx) added the unsetting capability
4743 such a setting is global if at outer level; local otherwise
4744 (?imsx-imsx:) non-capturing groups with option setting
4745 (?(cond)re|re) conditional pattern matching
4746
4747 A backreference to itself in a repeated group matches the previous
4748 captured string.
4749
4750 8. General tidying up of studying (both automatic and via "study")
4751 consequential on the addition of new assertions.
4752
4753 9. As in 5.005, unlimited repeated groups that could match an empty substring
4754 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4755 runtime if any iteration does actually match an empty substring.
4756
4757 10. Include the RunTest script in the distribution.
4758
4759 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4760 discrepancies, some of which were old and were also with respect to 5.004. They
4761 have now been fixed.
4762
4763
4764 Version 1.09 28-Apr-98
4765 ----------------------
4766
4767 1. A negated single character class followed by a quantifier with a minimum
4768 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4769 program crashes, or just wrong answers. This did not apply to negated classes
4770 containing more than one character, or to minima other than one.
4771
4772
4773 Version 1.08 27-Mar-98
4774 ----------------------
4775
4776 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4777
4778 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4779 latter must appear before anything that relies on it in the pattern.
4780
4781
4782 Version 1.07 16-Feb-98
4783 ----------------------
4784
4785 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4786 repeat of a potentially empty string).
4787
4788
4789 Version 1.06 23-Jan-98
4790 ----------------------
4791
4792 1. Added Markus Oberhumer's little patches for C++.
4793
4794 2. Literal strings longer than 255 characters were broken.
4795
4796
4797 Version 1.05 23-Dec-97
4798 ----------------------
4799
4800 1. Negated character classes containing more than one character were failing if
4801 PCRE_CASELESS was set at run time.
4802
4803
4804 Version 1.04 19-Dec-97
4805 ----------------------
4806
4807 1. Corrected the man page, where some "const" qualifiers had been omitted.
4808
4809 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4810 input syntax.
4811
4812 3. Fixed memory leak which occurred when a regex with back references was
4813 matched with an offsets vector that wasn't big enough. The temporary memory
4814 that is used in this case wasn't being freed if the match failed.
4815
4816 4. Tidied pcretest to ensure it frees memory that it gets.
4817
4818 5. Temporary memory was being obtained in the case where the passed offsets
4819 vector was exactly big enough.
4820
4821 6. Corrected definition of offsetof() from change 5 below.
4822
4823 7. I had screwed up change 6 below and broken the rules for the use of
4824 setjmp(). Now fixed.
4825
4826
4827 Version 1.03 18-Dec-97
4828 ----------------------
4829
4830 1. A erroneous regex with a missing opening parenthesis was correctly
4831 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4832 on some systems.
4833
4834 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4835 it was reported that one broken compiler failed on the former because "code" is
4836 also an independent variable.
4837
4838 3. The erroneous regex a[]b caused an array overrun reference.
4839
4840 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4841 fail on data ending with that character. (It was going on too far, and checking
4842 the next character, typically a binary zero.) This was specific to the
4843 optimized code for single-character negative classes.
4844
4845 5. Added a contributed patch from the TIN world which does the following:
4846
4847 + Add an undef for memmove, in case the the system defines a macro for it.
4848
4849 + Add a definition of offsetof(), in case there isn't one. (I don't know
4850 the reason behind this - offsetof() is part of the ANSI standard - but
4851 it does no harm).
4852
4853 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4854 most of the places where whitespace preceded '#'. I have given up and
4855 allowed the remaining 2 cases to be at the margin.
4856
4857 + Rename some variables in pcre to eliminate shadowing. This seems very
4858 pedantic, but does no harm, of course.
4859
4860 6. Moved the call to setjmp() into its own function, to get rid of warnings
4861 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4862
4863 7. Constructs such as \d{8,} were compiling into the equivalent of
4864 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4865 outcome, but in this particular case used more store than had been allocated,
4866 which caused the bug to be discovered because it threw up an internal error.
4867
4868 8. The debugging code in both pcre and pcretest for outputting the compiled
4869 form of a regex was going wrong in the case of back references followed by
4870 curly-bracketed repeats.
4871
4872
4873 Version 1.02 12-Dec-97
4874 ----------------------
4875
4876 1. Typos in pcre.3 and comments in the source fixed.
4877
4878 2. Applied a contributed patch to get rid of places where it used to remove
4879 'const' from variables, and fixed some signed/unsigned and uninitialized
4880 variable warnings.
4881
4882 3. Added the "runtest" target to Makefile.
4883
4884 4. Set default compiler flag to -O2 rather than just -O.
4885
4886
4887 Version 1.01 19-Nov-97
4888 ----------------------
4889
4890 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4891 like /([ab]*)*/, that is, for classes with more than one character in them.
4892
4893 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4894 as /((?>a*))*/ (a PCRE_EXTRA facility).
4895
4896
4897 Version 1.00 18-Nov-97
4898 ----------------------
4899
4900 1. Added compile-time macros to support systems such as SunOS4 which don't have
4901 memmove() or strerror() but have other things that can be used instead.
4902
4903 2. Arranged that "make clean" removes the executables.
4904
4905
4906 Version 0.99 27-Oct-97
4907 ----------------------
4908
4909 1. Fixed bug in code for optimizing classes with only one character. It was
4910 initializing a 32-byte map regardless, which could cause it to run off the end
4911 of the memory it had got.
4912
4913 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4914
4915
4916 Version 0.98 22-Oct-97
4917 ----------------------
4918
4919 1. Fixed bug in code for handling temporary memory usage when there are more
4920 back references than supplied space in the ovector. This could cause segfaults.
4921
4922
4923 Version 0.97 21-Oct-97
4924 ----------------------
4925
4926 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4927
4928 2. Optimized negated single characters not to use a bit map.
4929
4930 3. Brought error texts together as macro definitions; clarified some of them;
4931 fixed one that was wrong - it said "range out of order" when it meant "invalid
4932 escape sequence".
4933
4934 4. Changed some char * arguments to const char *.
4935
4936 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4937
4938 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4939 pcretest.
4940
4941
4942 Version 0.96 16-Oct-97
4943 ----------------------
4944
4945 1. Added a simple "pgrep" utility to the distribution.
4946
4947 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4948 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4949 where "ddd" means "one or more decimal digits".
4950
4951 3. Fixed serious bug. If a pattern had a back reference, but the call to
4952 pcre_exec() didn't supply a large enough ovector to record the related
4953 identifying subpattern, the match always failed. PCRE now remembers the number
4954 of the largest back reference, and gets some temporary memory in which to save
4955 the offsets during matching if necessary, in order to ensure that
4956 backreferences always work.
4957
4958 4. Increased the compatibility with Perl in a number of ways:
4959
4960 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4961 to request this handling. The option can be set at compile or exec time.
4962
4963 (b) $ matches before a terminating newline by default; an option
4964 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4965 mode). The option can be set at compile or exec time.
4966
4967 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4968 the same as Perl's. If the decimal number it represents is less than 10
4969 or there aren't that many previous left capturing parentheses, an octal
4970 escape is read. Inside a character class, it's always an octal escape,
4971 even if it is a single digit.
4972
4973 (d) An escaped but undefined alphabetic character is taken as a literal,
4974 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4975 escapes.
4976
4977 (e) {0} is now permitted. (The previous item is removed from the compiled
4978 pattern).
4979
4980 5. Changed all the names of code files so that the basic parts are no longer
4981 than 10 characters, and abolished the teeny "globals.c" file.
4982
4983 6. Changed the handling of character classes; they are now done with a 32-byte
4984 bit map always.
4985
4986 7. Added the -d and /D options to pcretest to make it possible to look at the
4987 internals of compilation without having to recompile pcre.
4988
4989
4990 Version 0.95 23-Sep-97
4991 ----------------------
4992
4993 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4994 \x20 at the start of a run of normal characters. These were being treated as
4995 real characters, instead of the source characters being re-checked.
4996
4997
4998 Version 0.94 18-Sep-97
4999 ----------------------
5000
5001 1. The functions are now thread-safe, with the caveat that the global variables
5002 containing pointers to malloc() and free() or alternative functions are the
5003 same for all threads.
5004
5005 2. Get pcre_study() to generate a bitmap of initial characters for non-
5006 anchored patterns when this is possible, and use it if passed to pcre_exec().
5007
5008
5009 Version 0.93 15-Sep-97
5010 ----------------------
5011
5012 1. /(b)|(:+)/ was computing an incorrect first character.
5013
5014 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
5015 but not actually doing anything yet.
5016
5017 3. Treat "-" characters in classes that cannot be part of ranges as literals,
5018 as Perl does (e.g. [-az] or [az-]).
5019
5020 4. Set the anchored flag if a branch starts with .* or .*? because that tests
5021 all possible positions.
5022
5023 5. Split up into different modules to avoid including unneeded functions in a
5024 compiled binary. However, compile and exec are still in one module. The "study"
5025 function is split off.
5026
5027 6. The character tables are now in a separate module whose source is generated
5028 by an auxiliary program - but can then be edited by hand if required. There are
5029 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
5030 toupper() in the code.
5031
5032 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
5033 make them global. Abolish the function for setting them, as the caller can now
5034 set them directly.
5035
5036
5037 Version 0.92 11-Sep-97
5038 ----------------------
5039
5040 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
5041 (e.g. /a{1,3}/) was broken (I mis-optimized it).
5042
5043 2. Caseless matching was not working in character classes if the characters in
5044 the pattern were in upper case.
5045
5046 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
5047
5048 4. Make PCRE_ANCHORED public and accept as a compile option.
5049
5050 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
5051 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
5052 pass them.
5053
5054 6. Give an error if bad option bits passed at compile or run time.
5055
5056 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
5057 pcretest to cause it to pass that flag.
5058
5059 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
5060 options, and the first character, if set.
5061
5062 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
5063
5064
5065 Version 0.91 10-Sep-97
5066 ----------------------
5067
5068 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
5069 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
5070
5071 2. PCRE was looping on encountering an indefinitely repeated back reference to
5072 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
5073 Perl does - treats the match as successful.
5074
5075 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5