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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1265 - (show annotations)
Sun Mar 3 10:42:46 2013 UTC (6 years, 9 months ago) by ph10
File size: 243996 byte(s)
Fix missing callout in alternative of conditional group when auto callout is 
set.
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
94 Version 8.32 30-November-2012
95 -----------------------------
96
97 1. Improved JIT compiler optimizations for first character search and single
98 character iterators.
99
100 2. Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
101 Patch by Daniel Richard G.
102
103 3. Single character iterator optimizations in the JIT compiler.
104
105 4. Improved JIT compiler optimizations for character ranges.
106
107 5. Rename the "leave" variable names to "quit" to improve WinCE compatibility.
108 Reported by Giuseppe D'Angelo.
109
110 6. The PCRE_STARTLINE bit, indicating that a match can occur only at the start
111 of a line, was being set incorrectly in cases where .* appeared inside
112 atomic brackets at the start of a pattern, or where there was a subsequent
113 *PRUNE or *SKIP.
114
115 7. Improved instruction cache flush for POWER/PowerPC.
116 Patch by Daniel Richard G.
117
118 8. Fixed a number of issues in pcregrep, making it more compatible with GNU
119 grep:
120
121 (a) There is now no limit to the number of patterns to be matched.
122
123 (b) An error is given if a pattern is too long.
124
125 (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
126 are now supported.
127
128 (d) --exclude-from and --include-from (multiple use) have been added.
129
130 (e) Exclusions and inclusions now apply to all files and directories, not
131 just to those obtained from scanning a directory recursively.
132
133 (f) Multiple uses of -f and --file-list are now supported.
134
135 (g) In a Windows environment, the default for -d has been changed from
136 "read" (the GNU grep default) to "skip", because otherwise the presence
137 of a directory in the file list provokes an error.
138
139 (h) The documentation has been revised and clarified in places.
140
141 9. Improve the matching speed of capturing brackets.
142
143 10. Changed the meaning of \X so that it now matches a Unicode extended
144 grapheme cluster.
145
146 11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
147 out POSIX threads when JIT support is configured.
148
149 12. Added support for PCRE_STUDY_EXTRA_NEEDED.
150
151 13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
152 structure could go wrong in environments where size_t is not the same size
153 as int.
154
155 14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
156 set.
157
158 15. The EBCDIC support had decayed; later updates to the code had included
159 explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
160 general tidy up of EBCDIC-related issues, and the documentation was also
161 not quite right. There is now a test that can be run on ASCII systems to
162 check some of the EBCDIC-related things (but is it not a full test).
163
164 16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
165 in a small tidy to the code.
166
167 17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
168
169 18. If the --only-matching (-o) option in pcregrep is specified multiple
170 times, each one causes appropriate output. For example, -o1 -o2 outputs the
171 substrings matched by the 1st and 2nd capturing parentheses. A separating
172 string can be specified by --om-separator (default empty).
173
174 19. Improving the first n character searches.
175
176 20. Turn case lists for horizontal and vertical white space into macros so that
177 they are defined only once.
178
179 21. This set of changes together give more compatible Unicode case-folding
180 behaviour for characters that have more than one other case when UCP
181 support is available.
182
183 (a) The Unicode property table now has offsets into a new table of sets of
184 three or more characters that are case-equivalent. The MultiStage2.py
185 script that generates these tables (the pcre_ucd.c file) now scans
186 CaseFolding.txt instead of UnicodeData.txt for character case
187 information.
188
189 (b) The code for adding characters or ranges of characters to a character
190 class has been abstracted into a generalized function that also handles
191 case-independence. In UTF-mode with UCP support, this uses the new data
192 to handle characters with more than one other case.
193
194 (c) A bug that is fixed as a result of (b) is that codepoints less than 256
195 whose other case is greater than 256 are now correctly matched
196 caselessly. Previously, the high codepoint matched the low one, but not
197 vice versa.
198
199 (d) The processing of \h, \H, \v, and \ in character classes now makes use
200 of the new class addition function, using character lists defined as
201 macros alongside the case definitions of 20 above.
202
203 (e) Caseless back references now work with characters that have more than
204 one other case.
205
206 (f) General caseless matching of characters with more than one other case
207 is supported.
208
209 22. Unicode character properties were updated from Unicode 6.2.0
210
211 23. Improved CMake support under Windows. Patch by Daniel Richard G.
212
213 24. Add support for 32-bit character strings, and UTF-32
214
215 25. Major JIT compiler update (code refactoring and bugfixing).
216 Experimental Sparc 32 support is added.
217
218 26. Applied a modified version of Daniel Richard G's patch to create
219 pcre.h.generic and config.h.generic by "make" instead of in the
220 PrepareRelease script.
221
222 27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
223 pcre_compile.c when checking for a zero character.
224
225 28. Introducing a native interface for JIT. Through this interface, the compiled
226 machine code can be directly executed. The purpose of this interface is to
227 provide fast pattern matching, so several sanity checks are not performed.
228 However, feature tests are still performed. The new interface provides
229 1.4x speedup compared to the old one.
230
231 29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
232 the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
233 was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
234
235 30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
236 greater than 0x7fffffff (which cannot be represented in UTF-8, even under
237 the "old" RFC 2279). Instead, it ended up passing a negative length to
238 pcre_exec().
239
240 31. Add support for GCC's visibility feature to hide internal functions.
241
242 32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
243 "unknown -C option" after outputting 0 or 1.
244
245 33. There is now support for generating a code coverage report for the test
246 suite in environments where gcc is the compiler and lcov is installed. This
247 is mainly for the benefit of the developers.
248
249 34. If PCRE is built with --enable-valgrind, certain memory regions are marked
250 unaddressable using valgrind annotations, allowing valgrind to detect
251 invalid memory accesses. This is mainly for the benefit of the developers.
252
253 25. (*UTF) can now be used to start a pattern in any of the three libraries.
254
255 26. Give configure error if --enable-cpp but no C++ compiler found.
256
257
258 Version 8.31 06-July-2012
259 -------------------------
260
261 1. Fixing a wrong JIT test case and some compiler warnings.
262
263 2. Removed a bashism from the RunTest script.
264
265 3. Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
266 to unsigned type, result still unsigned" that was given by an MS compiler
267 on encountering the code "-sizeof(xxx)".
268
269 4. Partial matching support is added to the JIT compiler.
270
271 5. Fixed several bugs concerned with partial matching of items that consist
272 of more than one character:
273
274 (a) /^(..)\1/ did not partially match "aba" because checking references was
275 done on an "all or nothing" basis. This also applied to repeated
276 references.
277
278 (b) \R did not give a hard partial match if \r was found at the end of the
279 subject.
280
281 (c) \X did not give a hard partial match after matching one or more
282 characters at the end of the subject.
283
284 (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
285 a partial match for the string "\r".
286
287 (e) When newline was set to CRLF, the metacharacter "." did not recognize
288 a partial match for a CR character at the end of the subject string.
289
290 6. If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
291 running pcretest, the text "(JIT)" added to the output whenever JIT is
292 actually used to run the match.
293
294 7. Individual JIT compile options can be set in pcretest by following -s+[+]
295 or /S+[+] with a digit between 1 and 7.
296
297 8. OP_NOT now supports any UTF character not just single-byte ones.
298
299 9. (*MARK) control verb is now supported by the JIT compiler.
300
301 10. The command "./RunTest list" lists the available tests without actually
302 running any of them. (Because I keep forgetting what they all are.)
303
304 11. Add PCRE_INFO_MAXLOOKBEHIND.
305
306 12. Applied a (slightly modified) user-supplied patch that improves performance
307 when the heap is used for recursion (compiled with --disable-stack-for-
308 recursion). Instead of malloc and free for each heap frame each time a
309 logical recursion happens, frames are retained on a chain and re-used where
310 possible. This sometimes gives as much as 30% improvement.
311
312 13. As documented, (*COMMIT) is now confined to within a recursive subpattern
313 call.
314
315 14. As documented, (*COMMIT) is now confined to within a positive assertion.
316
317 15. It is now possible to link pcretest with libedit as an alternative to
318 libreadline.
319
320 16. (*COMMIT) control verb is now supported by the JIT compiler.
321
322 17. The Unicode data tables have been updated to Unicode 6.1.0.
323
324 18. Added --file-list option to pcregrep.
325
326 19. Added binary file support to pcregrep, including the -a, --binary-files,
327 -I, and --text options.
328
329 20. The madvise function is renamed for posix_madvise for QNX compatibility
330 reasons. Fixed by Giuseppe D'Angelo.
331
332 21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
333
334 22. Changed the option for creating symbolic links for 16-bit man pages from
335 -s to -sf so that re-installing does not cause issues.
336
337 23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
338
339 24. Fixed a very old bug in pcretest that caused errors with restarted DFA
340 matches in certain environments (the workspace was not being correctly
341 retained). Also added to pcre_dfa_exec() a simple plausibility check on
342 some of the workspace data at the beginning of a restart.
343
344 25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
345 was not doing so when it should - probably a typo introduced by SVN 528
346 (change 8.10/14).
347
348 26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
349 \w+ when the character tables indicated that \x{c4} was a word character.
350 There were several related cases, all because the tests for doing a table
351 lookup were testing for characters less than 127 instead of 255.
352
353 27. If a pattern contains capturing parentheses that are not used in a match,
354 their slots in the ovector are set to -1. For those that are higher than
355 any matched groups, this happens at the end of processing. In the case when
356 there were back references that the ovector was too small to contain
357 (causing temporary malloc'd memory to be used during matching), and the
358 highest capturing number was not used, memory off the end of the ovector
359 was incorrectly being set to -1. (It was using the size of the temporary
360 memory instead of the true size.)
361
362 28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
363 ovector size, it uses memory at the end of the block that it has got.
364
365 29. Check for an overlong MARK name and give an error at compile time. The
366 limit is 255 for the 8-bit library and 65535 for the 16-bit library.
367
368 30. JIT compiler update.
369
370 31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
371 Rill for the patch.
372
373 32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
374
375 33. Variable renamings in the PCRE-JIT compiler. No functionality change.
376
377 34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
378 SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
379 was enabled.
380
381 35. Improve JIT code generation for greedy plus quantifier.
382
383 36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
384 1 to "aa" instead of to an empty string. The bug affected repeated groups
385 that could potentially match an empty string.
386
387 37. Optimizing single character iterators in JIT.
388
389 38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
390 the same checks as \x{...} characters in non-JavaScript mode. Specifically,
391 codepoints that are too big for the mode are faulted, and in a UTF mode,
392 disallowed codepoints are also faulted.
393
394 39. If PCRE was compiled with UTF support, in three places in the DFA
395 matcher there was code that should only have been obeyed in UTF mode, but
396 was being obeyed unconditionally. In 8-bit mode this could cause incorrect
397 processing when bytes with values greater than 127 were present. In 16-bit
398 mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
399 both cases the values are those that cannot be the first data item in a UTF
400 character. The three items that might have provoked this were recursions,
401 possessively repeated groups, and atomic groups.
402
403 40. Ensure that libpcre is explicitly listed in the link commands for pcretest
404 and pcregrep, because some OS require shared objects to be explicitly
405 passed to ld, causing the link step to fail if they are not.
406
407 41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
408 mode, patterns that started with \h* or \R* might be incorrectly matched.
409
410
411 Version 8.30 04-February-2012
412 -----------------------------
413
414 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
415 name is defined in ctype.h.
416
417 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
418 only in quite long subpatterns.
419
420 3. Removed the function pcre_info(), which has been obsolete and deprecated
421 since it was replaced by pcre_fullinfo() in February 2000.
422
423 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
424 match a (*MARK), and the match failed at the start of the subject, a
425 reference to memory before the start of the subject could occur. This bug
426 was introduced by fix 17 of release 8.21.
427
428 5. A reference to an unset group with zero minimum repetition was giving
429 totally wrong answers (in non-JavaScript-compatibility mode). For example,
430 /(another)?(\1?)test/ matched against "hello world test". This bug was
431 introduced in release 8.13.
432
433 6. Add support for 16-bit character strings (a large amount of work involving
434 many changes and refactorings).
435
436 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
437 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
438 from a file.
439
440 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
441 rounding is not applied in this particular case).
442
443 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
444 if they appear, or are escaped, in patterns.
445
446 10. Get rid of a number of -Wunused-but-set-variable warnings.
447
448 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
449 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
450 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
451 also returns the mark "x". This bug applied to capturing parentheses,
452 non-capturing parentheses, and atomic parentheses. It also applied to some
453 assertions.
454
455 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
456 information out of configure.ac instead of relying on pcre.h.generic, which
457 is not stored in the repository.
458
459 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
460 -lreadline.
461
462 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
463
464 15. Applied Graycode's patch to put the top-level frame on the stack rather
465 than the heap when not using the stack for recursion. This gives a
466 performance improvement in many cases when recursion is not deep.
467
468 16. Experimental code added to "pcretest -C" to output the stack frame size.
469
470
471 Version 8.21 12-Dec-2011
472 ------------------------
473
474 1. Updating the JIT compiler.
475
476 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
477 are added as well.
478
479 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
480 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
481 calling _pcre_jit_exec. Some extra comments are added.
482
483 4. (*MARK) settings inside atomic groups that do not contain any capturing
484 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
485 was introduced by change 18 for 8.20.
486
487 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
488 ECMA-262 standard.
489
490 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
491 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
492 This bug was probably introduced by change 9 of 8.13.
493
494 7. While fixing 6 above, I noticed that a number of other items were being
495 incorrectly rejected as "not fixed length". This arose partly because newer
496 opcodes had not been added to the fixed-length checking code. I have (a)
497 corrected the bug and added tests for these items, and (b) arranged for an
498 error to occur if an unknown opcode is encountered while checking for fixed
499 length instead of just assuming "not fixed length". The items that were
500 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
501 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
502 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
503
504 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
505 being incorrectly compiled and would have given unpredicatble results.
506
507 9. A possessively repeated subpattern with minimum repeat count greater than
508 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
509 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
510 the first (A) could occur when it should not.
511
512 10. Add a cast and remove a redundant test from the code.
513
514 11. JIT should use pcre_malloc/pcre_free for allocation.
515
516 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
517 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
518 is anything other than /usr, -L is still shown).
519
520 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
521
522 14. Perl does not support \N without a following name in a [] class; PCRE now
523 also gives an error.
524
525 15. If a forward reference was repeated with an upper limit of around 2000,
526 it caused the error "internal error: overran compiling workspace". The
527 maximum number of forward references (including repeats) was limited by the
528 internal workspace, and dependent on the LINK_SIZE. The code has been
529 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
530 the default depends on LINK_SIZE. There is a new upper limit (for safety)
531 of around 200,000 forward references. While doing this, I also speeded up
532 the filling in of repeated forward references.
533
534 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
535 incorrectly expecting the subject to contain another "a" after the start.
536
537 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
538 in the match, the SKIP should be ignored. This was not happening; instead
539 the SKIP was being treated as NOMATCH. For patterns such as
540 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
541 tested.
542
543 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
544 now much more compatible with Perl, in particular in cases where the result
545 is a non-match for a non-anchored pattern. For example, if
546 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
547 "m", where previously it did not return a name. A side effect of this
548 change is that for partial matches, the last encountered mark name is
549 returned, as for non matches. A number of tests that were previously not
550 Perl-compatible have been moved into the Perl-compatible test files. The
551 refactoring has had the pleasing side effect of removing one argument from
552 the match() function, thus reducing its stack requirements.
553
554 19. If the /S+ option was used in pcretest to study a pattern using JIT,
555 subsequent uses of /S (without +) incorrectly behaved like /S+.
556
557 21. Retrieve executable code size support for the JIT compiler and fixing
558 some warnings.
559
560 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
561 not work when the shorter character appeared right at the end of the
562 subject string.
563
564 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
565 systems.
566
567 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
568 output it when the /M option is used in pcretest.
569
570 25. The CheckMan script was not being included in the distribution. Also, added
571 an explicit "perl" to run Perl scripts from the PrepareRelease script
572 because this is reportedly needed in Windows.
573
574 26. If study data was being save in a file and studying had not found a set of
575 "starts with" bytes for the pattern, the data written to the file (though
576 never used) was taken from uninitialized memory and so caused valgrind to
577 complain.
578
579 27. Updated RunTest.bat as provided by Sheri Pierce.
580
581 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
582
583 29. Computation of memory usage for the table of capturing group names was
584 giving an unnecessarily large value.
585
586
587 Version 8.20 21-Oct-2011
588 ------------------------
589
590 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
591 a POSIX class. After further experiments with Perl, which convinced me that
592 Perl has bugs and confusions, a closing square bracket is no longer allowed
593 in a POSIX name. This bug also affected patterns with classes that started
594 with full stops.
595
596 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
597 captured substring, but while checking the failing first alternative,
598 substring 1 is temporarily captured. If the output vector supplied to
599 pcre_exec() was not big enough for this capture, the yield of the function
600 was still zero ("insufficient space for captured substrings"). This cannot
601 be totally fixed without adding another stack variable, which seems a lot
602 of expense for a edge case. However, I have improved the situation in cases
603 such as /(a)(b)x|abc/ matched against "abc", where the return code
604 indicates that fewer than the maximum number of slots in the ovector have
605 been set.
606
607 3. Related to (2) above: when there are more back references in a pattern than
608 slots in the output vector, pcre_exec() uses temporary memory during
609 matching, and copies in the captures as far as possible afterwards. It was
610 using the entire output vector, but this conflicts with the specification
611 that only 2/3 is used for passing back captured substrings. Now it uses
612 only the first 2/3, for compatibility. This is, of course, another edge
613 case.
614
615 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
616 main code base, and can be used by building with --enable-jit. When this is
617 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
618 runtime --no-jit option is given.
619
620 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
621 ovector, the return from the function was zero, implying that there were
622 other matches that did not fit. The correct "exactly full" value is now
623 returned.
624
625 6. If a subpattern that was called recursively or as a subroutine contained
626 (*PRUNE) or any other control that caused it to give a non-standard return,
627 invalid errors such as "Error -26 (nested recursion at the same subject
628 position)" or even infinite loops could occur.
629
630 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
631 computing the minimum length on reaching *ACCEPT, and so ended up with the
632 wrong value of 1 rather than 0. Further investigation indicates that
633 computing a minimum subject length in the presence of *ACCEPT is difficult
634 (think back references, subroutine calls), and so I have changed the code
635 so that no minimum is registered for a pattern that contains *ACCEPT.
636
637 8. If (*THEN) was present in the first (true) branch of a conditional group,
638 it was not handled as intended. [But see 16 below.]
639
640 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
641 Sheri Pierce.
642
643 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
644 the first byte in a match must be "a".
645
646 11. Change 17 for 8.13 increased the recursion depth for patterns like
647 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
648 pattern contains any instances of (*THEN). If it does not, the old
649 optimizations are restored. It would be nice to do this on a per-group
650 basis, but at the moment that is not feasible.
651
652 12. In some environments, the output of pcretest -C is CRLF terminated. This
653 broke RunTest's code that checks for the link size. A single white space
654 character after the value is now allowed for.
655
656 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
657 For "fr", it uses the Windows-specific input and output files.
658
659 14. If (*THEN) appeared in a group that was called recursively or as a
660 subroutine, it did not work as intended. [But see next item.]
661
662 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
663 pattern fragments (but not containing any | characters). If A and B are
664 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
665 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
666 D. In other words, Perl considers parentheses that do not contain any |
667 characters to be part of a surrounding alternative, whereas PCRE was
668 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
669 differently. PCRE now behaves in the same way as Perl, except in the case
670 of subroutine/recursion calls such as (?1) which have in any case always
671 been different (but PCRE had them first :-).
672
673 16. Related to 15 above: Perl does not treat the | in a conditional group as
674 creating alternatives. Such a group is treated in the same way as an
675 ordinary group without any | characters when processing (*THEN). PCRE has
676 been changed to match Perl's behaviour.
677
678 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
679 RunGrepTest script failed.
680
681 18. Change 22 for version 13 caused atomic groups to use more stack. This is
682 inevitable for groups that contain captures, but it can lead to a lot of
683 stack use in large patterns. The old behaviour has been restored for atomic
684 groups that do not contain any capturing parentheses.
685
686 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
687 suppress the check for a minimum subject length at run time. (If it was
688 given to pcre_exec() or pcre_dfa_exec() it did work.)
689
690 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
691 fail to work when decoding hex characters in data strings in EBCDIC
692 environments.
693
694 21. It appears that in at least one Mac OS environment, the isxdigit() function
695 is implemented as a macro that evaluates to its argument more than once,
696 contravening the C 90 Standard (I haven't checked a later standard). There
697 was an instance in pcretest which caused it to go wrong when processing
698 \x{...} escapes in subject strings. The has been rewritten to avoid using
699 things like p++ in the argument of isxdigit().
700
701
702 Version 8.13 16-Aug-2011
703 ------------------------
704
705 1. The Unicode data tables have been updated to Unicode 6.0.0.
706
707 2. Two minor typos in pcre_internal.h have been fixed.
708
709 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
710 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
711 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
712
713 4. There were a number of related bugs in the code for matching backrefences
714 caselessly in UTF-8 mode when codes for the characters concerned were
715 different numbers of bytes. For example, U+023A and U+2C65 are an upper
716 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
717 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
718 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
719 2-byte code at the end of the subject (it thought there wasn't enough data
720 left).
721
722 5. Comprehensive information about what went wrong is now returned by
723 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
724 as the output vector has at least 2 elements. The offset of the start of
725 the failing character and a reason code are placed in the vector.
726
727 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
728 now returned is for the first byte of the failing character, instead of the
729 last byte inspected. This is an incompatible change, but I hope it is small
730 enough not to be a problem. It makes the returned offset consistent with
731 pcre_exec() and pcre_dfa_exec().
732
733 7. pcretest now gives a text phrase as well as the error number when
734 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
735 failure, the offset and reason code are output.
736
737 8. When \R was used with a maximizing quantifier it failed to skip backwards
738 over a \r\n pair if the subsequent match failed. Instead, it just skipped
739 back over a single character (\n). This seems wrong (because it treated the
740 two characters as a single entity when going forwards), conflicts with the
741 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
742 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
743 has been changed.
744
745 9. Some internal refactoring has changed the processing so that the handling
746 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
747 time (the PCRE_DOTALL option was changed this way some time ago: version
748 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
749 which was always a bit of a fudge. It also means that there is one less
750 argument for the match() function, which reduces its stack requirements
751 slightly. This change also fixes an incompatibility with Perl: the pattern
752 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
753
754 10. More internal refactoring has drastically reduced the number of recursive
755 calls to match() for possessively repeated groups such as (abc)++ when
756 using pcre_exec().
757
758 11. While implementing 10, a number of bugs in the handling of groups were
759 discovered and fixed:
760
761 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
762 (a|)*(?1) gave a compile-time internal error.
763 ((a|)+)+ did not notice that the outer group could match an empty string.
764 (^a|^)+ was not marked as anchored.
765 (.*a|.*)+ was not marked as matching at start or after a newline.
766
767 12. Yet more internal refactoring has removed another argument from the match()
768 function. Special calls to this function are now indicated by setting a
769 value in a variable in the "match data" data block.
770
771 13. Be more explicit in pcre_study() instead of relying on "default" for
772 opcodes that mean there is no starting character; this means that when new
773 ones are added and accidentally left out of pcre_study(), testing should
774 pick them up.
775
776 14. The -s option of pcretest has been documented for ages as being an old
777 synonym of -m (show memory usage). I have changed it to mean "force study
778 for every regex", that is, assume /S for every regex. This is similar to -i
779 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
780 using it. It makes it easier to run collections of tests with and without
781 study enabled, and thereby test pcre_study() more easily. All the standard
782 tests are now run with and without -s (but some patterns can be marked as
783 "never study" - see 20 below).
784
785 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
786 restoration of the capturing data to the outer values was not happening
787 correctly.
788
789 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
790 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
791 pattern had matched an empty string, and so incorrectly returned a no
792 match.
793
794 17. There was optimizing code for the last branch of non-capturing parentheses,
795 and also for the obeyed branch of a conditional subexpression, which used
796 tail recursion to cut down on stack usage. Unfortunately, now that there is
797 the possibility of (*THEN) occurring in these branches, tail recursion is
798 no longer possible because the return has to be checked for (*THEN). These
799 two optimizations have therefore been removed. [But see 8.20/11 above.]
800
801 18. If a pattern containing \R was studied, it was assumed that \R always
802 matched two bytes, thus causing the minimum subject length to be
803 incorrectly computed because \R can also match just one byte.
804
805 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
806 was incorrectly computed.
807
808 20. If /S is present twice on a test pattern in pcretest input, it now
809 *disables* studying, thereby overriding the use of -s on the command line
810 (see 14 above). This is necessary for one or two tests to keep the output
811 identical in both cases.
812
813 21. When (*ACCEPT) was used in an assertion that matched an empty string and
814 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
815
816 22. When an atomic group that contained a capturing parenthesis was
817 successfully matched, but the branch in which it appeared failed, the
818 capturing was not being forgotten if a higher numbered group was later
819 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
820 group 1 to "a", when in fact it should be unset. This applied to multi-
821 branched capturing and non-capturing groups, repeated or not, and also to
822 positive assertions (capturing in negative assertions does not happen
823 in PCRE) and also to nested atomic groups.
824
825 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
826 subject after a captured substring, to make it easier to tell which of a
827 number of identical substrings has been captured.
828
829 24. The way atomic groups are processed by pcre_exec() has been changed so that
830 if they are repeated, backtracking one repetition now resets captured
831 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
832 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
833 "aaa". Previously, it would have been "a". As part of this code
834 refactoring, the way recursive calls are handled has also been changed.
835
836 25. If an assertion condition captured any substrings, they were not passed
837 back unless some other capturing happened later. For example, if
838 (?(?=(a))a) was matched against "a", no capturing was returned.
839
840 26. When studying a pattern that contained subroutine calls or assertions,
841 the code for finding the minimum length of a possible match was handling
842 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
843 group 1 called group 2 while simultaneously a separate group 2 called group
844 1). A stack overflow occurred in this case. I have fixed this by limiting
845 the recursion depth to 10.
846
847 27. Updated RunTest.bat in the distribution to the version supplied by Tom
848 Fortmann. This supports explicit test numbers on the command line, and has
849 argument validation and error reporting.
850
851 28. An instance of \X with an unlimited repeat could fail if at any point the
852 first character it looked at was a mark character.
853
854 29. Some minor code refactoring concerning Unicode properties and scripts
855 should reduce the stack requirement of match() slightly.
856
857 30. Added the '=' option to pcretest to check the setting of unused capturing
858 slots at the end of the pattern, which are documented as being -1, but are
859 not included in the return count.
860
861 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
862 compiled something random. Now it gives a compile-time error (as does
863 Perl).
864
865 32. A *MARK encountered during the processing of a positive assertion is now
866 recorded and passed back (compatible with Perl).
867
868 33. If --only-matching or --colour was set on a pcregrep call whose pattern
869 had alternative anchored branches, the search for a second match in a line
870 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
871 matched the line "0102" twice. The same bug affected patterns that started
872 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
873 twice.
874
875 34. Previously, PCRE did not allow quantification of assertions. However, Perl
876 does, and because of capturing effects, quantifying parenthesized
877 assertions may at times be useful. Quantifiers are now allowed for
878 parenthesized assertions.
879
880 35. A minor code tidy in pcre_compile() when checking options for \R usage.
881
882 36. \g was being checked for fancy things in a character class, when it should
883 just be a literal "g".
884
885 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
886 appearance of a nested POSIX class supersedes an apparent external class.
887 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
888 unescaped square brackets may also appear as part of class names. For
889 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
890 more like Perl. (But see 8.20/1 above.)
891
892 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
893 was because it thought it was \N{name}, which is not supported).
894
895 39. Add minix to OS list not supporting the -S option in pcretest.
896
897 40. PCRE tries to detect cases of infinite recursion at compile time, but it
898 cannot analyze patterns in sufficient detail to catch mutual recursions
899 such as ((?1))((?2)). There is now a runtime test that gives an error if a
900 subgroup is called recursively as a subpattern for a second time at the
901 same position in the subject string. In previous releases this might have
902 been caught by the recursion limit, or it might have run out of stack.
903
904 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
905 happen only once. PCRE was, however incorrectly giving a compile time error
906 "recursive call could loop indefinitely" because it cannot analyze the
907 pattern in sufficient detail. The compile time test no longer happens when
908 PCRE is compiling a conditional subpattern, but actual runaway loops are
909 now caught at runtime (see 40 above).
910
911 42. It seems that Perl allows any characters other than a closing parenthesis
912 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
913 has been changed to be the same.
914
915 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
916 as not to get warnings when autogen.sh is called. Also changed
917 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
918
919 44. To help people who use pcregrep to scan files containing exceedingly long
920 lines, the following changes have been made:
921
922 (a) The default value of the buffer size parameter has been increased from
923 8K to 20K. (The actual buffer used is three times this size.)
924
925 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
926 PCRE is built.
927
928 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
929 to be set at run time.
930
931 (d) Numerical values in pcregrep options can be followed by K or M, for
932 example --buffer-size=50K.
933
934 (e) If a line being scanned overflows pcregrep's buffer, an error is now
935 given and the return code is set to 2.
936
937 45. Add a pointer to the latest mark to the callout data block.
938
939 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
940 partial match of an empty string instead of no match. This was specific to
941 the use of ".".
942
943 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
944 complete match instead of a partial match. This bug was dependent on both
945 the PCRE_UTF8 and PCRE_DOTALL options being set.
946
947 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
948 starting byte set, because \b was not being ignored.
949
950
951 Version 8.12 15-Jan-2011
952 ------------------------
953
954 1. Fixed some typos in the markup of the man pages, and wrote a script that
955 checks for such things as part of the documentation building process.
956
957 2. On a big-endian 64-bit system, pcregrep did not correctly process the
958 --match-limit and --recursion-limit options (added for 8.11). In
959 particular, this made one of the standard tests fail. (The integer value
960 went into the wrong half of a long int.)
961
962 3. If the --colour option was given to pcregrep with -v (invert match), it
963 did strange things, either producing crazy output, or crashing. It should,
964 of course, ignore a request for colour when reporting lines that do not
965 match.
966
967 4. Another pcregrep bug caused similar problems if --colour was specified with
968 -M (multiline) and the pattern match finished with a line ending.
969
970 5. In pcregrep, when a pattern that ended with a literal newline sequence was
971 matched in multiline mode, the following line was shown as part of the
972 match. This seems wrong, so I have changed it.
973
974 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
975 the check for further matches in the same line (so they could be coloured)
976 to overrun the end of the current line. If another match was found, it was
977 incorrectly shown (and then shown again when found in the next line).
978
979 7. If pcregrep was compiled under Windows, there was a reference to the
980 function pcregrep_exit() before it was defined. I am assuming this was
981 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
982 reported by a user. I've moved the definition above the reference.
983
984
985 Version 8.11 10-Dec-2010
986 ------------------------
987
988 1. (*THEN) was not working properly if there were untried alternatives prior
989 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
990 backtracked to try for "b" instead of moving to the next alternative branch
991 at the same level (in this case, to look for "c"). The Perl documentation
992 is clear that when (*THEN) is backtracked onto, it goes to the "next
993 alternative in the innermost enclosing group".
994
995 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
996 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
997 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
998 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
999 (*THEN).
1000
1001 3. If \s appeared in a character class, it removed the VT character from
1002 the class, even if it had been included by some previous item, for example
1003 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1004 of \s, but is part of the POSIX "space" class.)
1005
1006 4. A partial match never returns an empty string (because you can always
1007 match an empty string at the end of the subject); however the checking for
1008 an empty string was starting at the "start of match" point. This has been
1009 changed to the "earliest inspected character" point, because the returned
1010 data for a partial match starts at this character. This means that, for
1011 example, /(?<=abc)def/ gives a partial match for the subject "abc"
1012 (previously it gave "no match").
1013
1014 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1015 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1016 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1017 has an implication that the given string is incomplete (because a partial
1018 match is preferred over a full match). For this reason, these items now
1019 give a partial match in this situation. [Aside: previously, the one case
1020 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1021 match rather than a full match, which was wrong by the old rules, but is
1022 now correct.]
1023
1024 6. There was a bug in the handling of #-introduced comments, recognized when
1025 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1026 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1027 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1028 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1029 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1030 places in pcre_compile().
1031
1032 7. Related to (6) above, when pcre_compile() was skipping #-introduced
1033 comments when looking ahead for named forward references to subpatterns,
1034 the only newline sequence it recognized was NL. It now handles newlines
1035 according to the set newline convention.
1036
1037 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1038 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1039 cater for a lack of strerror(). These oversights have been fixed.
1040
1041 9. Added --match-limit and --recursion-limit to pcregrep.
1042
1043 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1044
1045 11. When the -o option was used, pcregrep was setting a return code of 1, even
1046 when matches were found, and --line-buffered was not being honoured.
1047
1048 12. Added an optional parentheses number to the -o and --only-matching options
1049 of pcregrep.
1050
1051 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1052 can match an empty string. The code to do it in pcretest and pcredemo
1053 needed fixing:
1054
1055 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1056 only one byte after an empty string match just before CRLF (this case
1057 just got forgotten; "any" and "anycrlf" were OK).
1058
1059 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1060 mode when an empty string match preceded an ASCII character followed by
1061 a non-ASCII character. (The code for advancing by one character rather
1062 than one byte was nonsense.)
1063
1064 (c) The pcredemo.c sample program did not have any code at all to handle
1065 the cases when CRLF is a valid newline sequence.
1066
1067 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1068 as a starting offset was within the subject string. There is now a new
1069 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1070 negative or greater than the length of the string. In order to test this,
1071 pcretest is extended to allow the setting of negative starting offsets.
1072
1073 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1074 starting offset points to the beginning of a UTF-8 character was
1075 unnecessarily clumsy. I tidied it up.
1076
1077 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1078 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1079
1080 17. Nobody had reported that the --include_dir option, which was added in
1081 release 7.7 should have been called --include-dir (hyphen, not underscore)
1082 for compatibility with GNU grep. I have changed it to --include-dir, but
1083 left --include_dir as an undocumented synonym, and the same for
1084 --exclude-dir, though that is not available in GNU grep, at least as of
1085 release 2.5.4.
1086
1087 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1088 characters from a string of bytes) have been redefined so as not to use
1089 loops, in order to improve performance in some environments. At the same
1090 time, I abstracted some of the common code into auxiliary macros to save
1091 repetition (this should not affect the compiled code).
1092
1093 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1094 compile-time error is now given if \c is not followed by an ASCII
1095 character, that is, a byte less than 128. (In EBCDIC mode, the code is
1096 different, and any byte value is allowed.)
1097
1098 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1099 START_OPTIMIZE option, which is now allowed at compile time - but just
1100 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1101 to pcregrep and other applications that have no direct access to PCRE
1102 options. The new /Y option in pcretest sets this option when calling
1103 pcre_compile().
1104
1105 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1106 back references. Groups containing recursive back references were forced to
1107 be atomic by that change, but in the case of named groups, the amount of
1108 memory required was incorrectly computed, leading to "Failed: internal
1109 error: code overflow". This has been fixed.
1110
1111 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1112 pcretest.c, to avoid build problems in some Borland environments.
1113
1114
1115 Version 8.10 25-Jun-2010
1116 ------------------------
1117
1118 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1119 THEN.
1120
1121 2. (*ACCEPT) was not working when inside an atomic group.
1122
1123 3. Inside a character class, \B is treated as a literal by default, but
1124 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1125 causes the error). The code is unchanged, but I tidied the documentation.
1126
1127 4. Inside a character class, PCRE always treated \R and \X as literals,
1128 whereas Perl faults them if its -w option is set. I have changed PCRE so
1129 that it faults them when PCRE_EXTRA is set.
1130
1131 5. Added support for \N, which always matches any character other than
1132 newline. (It is the same as "." when PCRE_DOTALL is not set.)
1133
1134 6. When compiling pcregrep with newer versions of gcc which may have
1135 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1136 declared with attribute warn_unused_result" were given. Just casting the
1137 result to (void) does not stop the warnings; a more elaborate fudge is
1138 needed. I've used a macro to implement this.
1139
1140 7. Minor change to pcretest.c to avoid a compiler warning.
1141
1142 8. Added four artifical Unicode properties to help with an option to make
1143 \s etc use properties (see next item). The new properties are: Xan
1144 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1145
1146 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1147 use Unicode properties. (*UCP) at the start of a pattern can be used to set
1148 this option. Modified pcretest to add /W to test this facility. Added
1149 REG_UCP to make it available via the POSIX interface.
1150
1151 10. Added --line-buffered to pcregrep.
1152
1153 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1154 studied, and the match started with a letter with a code point greater than
1155 127 whose first byte was different to the first byte of the other case of
1156 the letter, the other case of this starting letter was not recognized
1157 (#976).
1158
1159 12. If a pattern that was studied started with a repeated Unicode property
1160 test, for example, \p{Nd}+, there was the theoretical possibility of
1161 setting up an incorrect bitmap of starting bytes, but fortunately it could
1162 not have actually happened in practice until change 8 above was made (it
1163 added property types that matched character-matching opcodes).
1164
1165 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1166 possible starting bytes for non-anchored patterns.
1167
1168 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1169 \R, and also a number of cases that involve Unicode properties, both
1170 explicit and implicit when PCRE_UCP is set.
1171
1172 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1173 input, it could crash or give wrong results if characters with values
1174 greater than 0xc0 were present in the subject string. (Detail: it assumed
1175 UTF-8 input when processing these items.)
1176
1177 16. Added a lot of (int) casts to avoid compiler warnings in systems where
1178 size_t is 64-bit (#991).
1179
1180 17. Added a check for running out of memory when PCRE is compiled with
1181 --disable-stack-for-recursion (#990).
1182
1183 18. If the last data line in a file for pcretest does not have a newline on
1184 the end, a newline was missing in the output.
1185
1186 19. The default pcre_chartables.c file recognizes only ASCII characters (values
1187 less than 128) in its various bitmaps. However, there is a facility for
1188 generating tables according to the current locale when PCRE is compiled. It
1189 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1190 characters, are recognized by isspace() and therefore were getting set in
1191 these tables, and indeed these tables seem to approximate to ISO 8859. This
1192 caused a problem in UTF-8 mode when pcre_study() was used to create a list
1193 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1194 which of course cannot start UTF-8 characters. I have changed the code so
1195 that only real ASCII characters (less than 128) and the correct starting
1196 bytes for UTF-8 encodings are set for characters greater than 127 when in
1197 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1198 altogether.)
1199
1200 20. Added the /T option to pcretest so as to be able to run tests with non-
1201 standard character tables, thus making it possible to include the tests
1202 used for 19 above in the standard set of tests.
1203
1204 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1205 reference to a subpattern the other side of a comment that contains an
1206 opening parenthesis caused either an internal compiling error, or a
1207 reference to the wrong subpattern.
1208
1209
1210 Version 8.02 19-Mar-2010
1211 ------------------------
1212
1213 1. The Unicode data tables have been updated to Unicode 5.2.0.
1214
1215 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
1216 configured.
1217
1218 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
1219 original author of that file, following a query about its status.
1220
1221 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
1222 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1223
1224 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1225 quantifier applied to a forward-referencing subroutine call, could compile
1226 incorrect code or give the error "internal error: previously-checked
1227 referenced subpattern not found".
1228
1229 6. Both MS Visual Studio and Symbian OS have problems with initializing
1230 variables to point to external functions. For these systems, therefore,
1231 pcre_malloc etc. are now initialized to local functions that call the
1232 relevant global functions.
1233
1234 7. There were two entries missing in the vectors called coptable and poptable
1235 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1236 I've fixed the data, and added a kludgy way of testing at compile time that
1237 the lengths are correct (equal to the number of opcodes).
1238
1239 8. Following on from 7, I added a similar kludge to check the length of the
1240 eint vector in pcreposix.c.
1241
1242 9. Error texts for pcre_compile() are held as one long string to avoid too
1243 much relocation at load time. To find a text, the string is searched,
1244 counting zeros. There was no check for running off the end of the string,
1245 which could happen if a new error number was added without updating the
1246 string.
1247
1248 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1249
1250 11. \K was not working if it appeared in an atomic group or in a group that
1251 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1252 \K is "not well defined" if used in an assertion. PCRE now accepts it if
1253 the assertion is positive, but not if it is negative.
1254
1255 12. Change 11 fortuitously reduced the size of the stack frame used in the
1256 "match()" function of pcre_exec.c by one pointer. Forthcoming
1257 implementation of support for (*MARK) will need an extra pointer on the
1258 stack; I have reserved it now, so that the stack frame size does not
1259 decrease.
1260
1261 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1262 item in branch that calls a recursion is a subroutine call - as in the
1263 second branch in the above example - was incorrectly given the compile-
1264 time error "recursive call could loop indefinitely" because pcre_compile()
1265 was not correctly checking the subroutine for matching a non-empty string.
1266
1267 14. The checks for overrunning compiling workspace could trigger after an
1268 overrun had occurred. This is a "should never occur" error, but it can be
1269 triggered by pathological patterns such as hundreds of nested parentheses.
1270 The checks now trigger 100 bytes before the end of the workspace.
1271
1272 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1273
1274
1275 Version 8.01 19-Jan-2010
1276 ------------------------
1277
1278 1. If a pattern contained a conditional subpattern with only one branch (in
1279 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1280 computed the wrong minimum data length (which is of course zero for such
1281 subpatterns). This could cause incorrect "no match" results.
1282
1283 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1284 the pattern is reset in the first branch, pcre_compile() failed with
1285 "internal error: code overflow at offset...". This happened only when
1286 the reset was to the original external option setting. (An optimization
1287 abstracts leading options settings into an external setting, which was the
1288 cause of this.)
1289
1290 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1291 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1292 assertion pattern did not match (meaning that the assertion was true), it
1293 was incorrectly treated as false if the SKIP had been reached during the
1294 matching. This also applied to assertions used as conditions.
1295
1296 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
1297 assertion subpattern, including such a pattern used as a condition,
1298 unpredictable results occurred, instead of the error return
1299 PCRE_ERROR_DFA_UITEM.
1300
1301 5. The C++ GlobalReplace function was not working like Perl for the special
1302 situation when an empty string is matched. It now does the fancy magic
1303 stuff that is necessary.
1304
1305 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1306 removed. (These were left over from very, very early versions of PCRE.)
1307
1308 7. Some cosmetic changes to the code to make life easier when compiling it
1309 as part of something else:
1310
1311 (a) Change DEBUG to PCRE_DEBUG.
1312
1313 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1314 called "current" as "current_branch", to prevent a collision with the
1315 Linux macro when compiled as a kernel module.
1316
1317 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1318 prevent a collision with the Linux macro when compiled as a kernel
1319 module.
1320
1321 8. In pcre_compile() there are some checks for integer overflows that used to
1322 cast potentially large values to (double). This has been changed to that
1323 when building, a check for int64_t is made, and if it is found, it is used
1324 instead, thus avoiding the use of floating point arithmetic. (There is no
1325 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1326 double.
1327
1328 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1329 2005 (difference between two addresses compared to an unsigned value).
1330
1331 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1332 custom one, because of the following reported problem in Windows:
1333
1334 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1335 under Win32.
1336 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1337 therefore missing the function definition.
1338 - The compiler thus generates a "C" signature for the test function.
1339 - The linker fails to find the "C" function.
1340 - PCRE fails to configure if asked to do so against libbz2.
1341
1342 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1343 messages were output:
1344
1345 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1346 rerunning libtoolize, to keep the correct libtool macros in-tree.
1347 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1348
1349 I have done both of these things.
1350
1351 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1352 most of the time, it *can* run out if it is given a pattern that contains a
1353 runaway infinite recursion. I updated the discussion in the pcrestack man
1354 page.
1355
1356 13. Now that we have gone to the x.xx style of version numbers, the minor
1357 version may start with zero. Using 08 or 09 is a bad idea because users
1358 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1359 interpreted as invalid octal numbers. I've updated the previous comment in
1360 configure.ac, and also added a check that gives an error if 08 or 09 are
1361 used.
1362
1363 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1364 causing partial matching to fail when the end of the subject matched \W
1365 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1366
1367 15. There were some discrepancies between the declarations in pcre_internal.h
1368 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1369 their definitions. The declarations used "const uschar *" and the
1370 definitions used USPTR. Even though USPTR is normally defined as "const
1371 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1372 reported that: "This difference in casting confuses some C++ compilers, for
1373 example, SunCC recognizes above declarations as different functions and
1374 generates broken code for hbpcre." I have changed the declarations to use
1375 USPTR.
1376
1377 16. GNU libtool is named differently on some systems. The autogen.sh script now
1378 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1379 (FreeBSD).
1380
1381 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1382 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1383 comment: "Figure out how to create a longlong from a string: strtoll and
1384 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1385 instance, but it only takes 2 args instead of 3!"
1386
1387 18. A subtle bug concerned with back references has been fixed by a change of
1388 specification, with a corresponding code fix. A pattern such as
1389 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1390 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1391 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1392 same bug. Such groups have to be quantified to be useful, or contained
1393 inside another quantified group. (If there's no repetition, the reference
1394 can never match.) The problem arises because, having left the group and
1395 moved on to the rest of the pattern, a later failure that backtracks into
1396 the group uses the captured value from the final iteration of the group
1397 rather than the correct earlier one. I have fixed this in PCRE by forcing
1398 any group that contains a reference to itself to be an atomic group; that
1399 is, there cannot be any backtracking into it once it has completed. This is
1400 similar to recursive and subroutine calls.
1401
1402
1403 Version 8.00 19-Oct-09
1404 ----------------------
1405
1406 1. The table for translating pcre_compile() error codes into POSIX error codes
1407 was out-of-date, and there was no check on the pcre_compile() error code
1408 being within the table. This could lead to an OK return being given in
1409 error.
1410
1411 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1412 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1413 in a Windows environment.
1414
1415 3. The pcregrep --count option prints the count for each file even when it is
1416 zero, as does GNU grep. However, pcregrep was also printing all files when
1417 --files-with-matches was added. Now, when both options are given, it prints
1418 counts only for those files that have at least one match. (GNU grep just
1419 prints the file name in this circumstance, but including the count seems
1420 more useful - otherwise, why use --count?) Also ensured that the
1421 combination -clh just lists non-zero counts, with no names.
1422
1423 4. The long form of the pcregrep -F option was incorrectly implemented as
1424 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1425 but it seems right to fix it, and I didn't think it was worth preserving
1426 the old behaviour.
1427
1428 5. The command line items --regex=pattern and --regexp=pattern were not
1429 recognized by pcregrep, which required --regex pattern or --regexp pattern
1430 (with a space rather than an '='). The man page documented the '=' forms,
1431 which are compatible with GNU grep; these now work.
1432
1433 6. No libpcreposix.pc file was created for pkg-config; there was just
1434 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1435
1436 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1437 when UCP support is not needed, by modifying the Python script that
1438 generates it from Unicode data files. This should not matter if the module
1439 is correctly used as a library, but I received one complaint about 50K of
1440 unwanted data. My guess is that the person linked everything into his
1441 program rather than using a library. Anyway, it does no harm.
1442
1443 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1444 was a minimum greater than 1 for a wide character in a possessive
1445 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1446 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1447 character. Chaos in the form of incorrect output or a compiling loop could
1448 result.
1449
1450 9. The restrictions on what a pattern can contain when partial matching is
1451 requested for pcre_exec() have been removed. All patterns can now be
1452 partially matched by this function. In addition, if there are at least two
1453 slots in the offset vector, the offset of the earliest inspected character
1454 for the match and the offset of the end of the subject are set in them when
1455 PCRE_ERROR_PARTIAL is returned.
1456
1457 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1458 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1459 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1460 and may be more useful for multi-segment matching.
1461
1462 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1463 used to be given if ever the end of the subject was reached; now it is
1464 given only if matching could not proceed because another character was
1465 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1466 string "Z", which now yields "no match" instead of "partial match". In the
1467 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1468 final character ended with (*FAIL).
1469
1470 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1471 if the pattern had a "must contain" character that was already found in the
1472 earlier partial match, unless partial matching was again requested. For
1473 example, with the pattern /dog.(body)?/, the "must contain" character is
1474 "g". If the first part-match was for the string "dog", restarting with
1475 "sbody" failed. This bug has been fixed.
1476
1477 13. The string returned by pcre_dfa_exec() after a partial match has been
1478 changed so that it starts at the first inspected character rather than the
1479 first character of the match. This makes a difference only if the pattern
1480 starts with a lookbehind assertion or \b or \B (\K is not supported by
1481 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1482 matching functions compatible, and I think it's the right thing to do.
1483
1484 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1485 so that the demonstration program is easily available in environments where
1486 PCRE has not been installed from source.
1487
1488 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1489 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1490 library.
1491
1492 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1493 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1494 is not the first non-POSIX option to be added. Clearly some people find
1495 these options useful.
1496
1497 17. If a caller to the POSIX matching function regexec() passes a non-zero
1498 value for nmatch with a NULL value for pmatch, the value of
1499 nmatch is forced to zero.
1500
1501 18. RunGrepTest did not have a test for the availability of the -u option of
1502 the diff command, as RunTest does. It now checks in the same way as
1503 RunTest, and also checks for the -b option.
1504
1505 19. If an odd number of negated classes containing just a single character
1506 interposed, within parentheses, between a forward reference to a named
1507 subpattern and the definition of the subpattern, compilation crashed with
1508 an internal error, complaining that it could not find the referenced
1509 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1510 [The bug was that it was starting one character too far in when skipping
1511 over the character class, thus treating the ] as data rather than
1512 terminating the class. This meant it could skip too much.]
1513
1514 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1515 /g option in pcretest when the pattern contains \K, which makes it possible
1516 to have an empty string match not at the start, even when the pattern is
1517 anchored. Updated pcretest and pcredemo to use this option.
1518
1519 21. If the maximum number of capturing subpatterns in a recursion was greater
1520 than the maximum at the outer level, the higher number was returned, but
1521 with unset values at the outer level. The correct (outer level) value is
1522 now given.
1523
1524 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1525 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1526 make it do so. The string so far is captured, making this feature
1527 compatible with Perl.
1528
1529 23. The tests have been re-organized, adding tests 11 and 12, to make it
1530 possible to check the Perl 5.10 features against Perl 5.10.
1531
1532 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1533 pattern matches a fixed length string. PCRE did not allow this; now it
1534 does. Neither allows recursion.
1535
1536 25. I finally figured out how to implement a request to provide the minimum
1537 length of subject string that was needed in order to match a given pattern.
1538 (It was back references and recursion that I had previously got hung up
1539 on.) This code has now been added to pcre_study(); it finds a lower bound
1540 to the length of subject needed. It is not necessarily the greatest lower
1541 bound, but using it to avoid searching strings that are too short does give
1542 some useful speed-ups. The value is available to calling programs via
1543 pcre_fullinfo().
1544
1545 26. While implementing 25, I discovered to my embarrassment that pcretest had
1546 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1547 study optimizations had never been tested with that matching function.
1548 Oops. What is worse, even when it was passed study data, there was a bug in
1549 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1550 were also very few tests of studied patterns with pcre_dfa_exec().
1551
1552 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1553 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1554 on the other side of the coin, they are no longer allowed to have different
1555 names, because these cannot be distinguished in PCRE, and this has caused
1556 confusion. (This is a difference from Perl.)
1557
1558 28. When duplicate subpattern names are present (necessarily with different
1559 numbers, as required by 27 above), and a test is made by name in a
1560 conditional pattern, either for a subpattern having been matched, or for
1561 recursion in such a pattern, all the associated numbered subpatterns are
1562 tested, and the overall condition is true if the condition is true for any
1563 one of them. This is the way Perl works, and is also more like the way
1564 testing by number works.
1565
1566
1567 Version 7.9 11-Apr-09
1568 ---------------------
1569
1570 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1571 (pcretest), all targets were linked against these libraries. This included
1572 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1573 libraries. This caused unwanted dependencies to be created. This problem
1574 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1575 pcretest is linked with readline.
1576
1577 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1578 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1579 moved outside it again, because FALSE and TRUE are already defined in AIX,
1580 but BOOL is not.
1581
1582 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1583 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1584
1585 4. The pcregrep documentation said spaces were inserted as well as colons (or
1586 hyphens) following file names and line numbers when outputting matching
1587 lines. This is not true; no spaces are inserted. I have also clarified the
1588 wording for the --colour (or --color) option.
1589
1590 5. In pcregrep, when --colour was used with -o, the list of matching strings
1591 was not coloured; this is different to GNU grep, so I have changed it to be
1592 the same.
1593
1594 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1595 each matching line was coloured. Now it goes on to look for further matches
1596 of any of the test patterns, which is the same behaviour as GNU grep.
1597
1598 7. A pattern that could match an empty string could cause pcregrep to loop; it
1599 doesn't make sense to accept an empty string match in pcregrep, so I have
1600 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1601 seems to be how GNU grep behaves.
1602
1603 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1604 start or after a newline", because the conditional assertion was not being
1605 correctly handled. The rule now is that both the assertion and what follows
1606 in the first alternative must satisfy the test.
1607
1608 9. If auto-callout was enabled in a pattern with a conditional group whose
1609 condition was an assertion, PCRE could crash during matching, both with
1610 pcre_exec() and pcre_dfa_exec().
1611
1612 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1613 used for matching.
1614
1615 11. Unicode property support in character classes was not working for
1616 characters (bytes) greater than 127 when not in UTF-8 mode.
1617
1618 12. Added the -M command line option to pcretest.
1619
1620 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1621
1622 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1623
1624 16. Added comments and documentation about mis-use of no_arg in the C++
1625 wrapper.
1626
1627 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1628 from Martin Jerabek that uses macro names for all relevant character and
1629 string constants.
1630
1631 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1632 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1633 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1634 these, but not everybody uses configure.
1635
1636 19. A conditional group that had only one branch was not being correctly
1637 recognized as an item that could match an empty string. This meant that an
1638 enclosing group might also not be so recognized, causing infinite looping
1639 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1640 with the subject "ab", where knowledge that the repeated group can match
1641 nothing is needed in order to break the loop.
1642
1643 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1644 exec(), but without supplying a callout function, matching went wrong.
1645
1646 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1647 leak if the size of the offset vector was greater than 30. When the vector
1648 is smaller, the saved offsets during recursion go onto a local stack
1649 vector, but for larger vectors malloc() is used. It was failing to free
1650 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1651 error, in fact).
1652
1653 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1654 heapframe that is used only when UTF-8 support is enabled. This caused no
1655 problem, but was untidy.
1656
1657 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1658 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1659 included within another project.
1660
1661 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1662 slightly modified by me:
1663
1664 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1665 not building pcregrep.
1666
1667 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1668 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1669
1670 25. Forward references, both numeric and by name, in patterns that made use of
1671 duplicate group numbers, could behave incorrectly or give incorrect errors,
1672 because when scanning forward to find the reference group, PCRE was not
1673 taking into account the duplicate group numbers. A pattern such as
1674 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1675
1676 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1677 the feature of a custom pointer more persuasive (as requested by a user).
1678
1679 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1680 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1681 pre-defined.
1682
1683 28. Added support for (*UTF8) at the start of a pattern.
1684
1685 29. Arrange for flags added by the "release type" setting in CMake to be shown
1686 in the configuration summary.
1687
1688
1689 Version 7.8 05-Sep-08
1690 ---------------------
1691
1692 1. Replaced UCP searching code with optimized version as implemented for Ad
1693 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1694 stage table and inline lookup instead of a function, giving speed ups of 2
1695 to 5 times on some simple patterns that I tested. Permission was given to
1696 distribute the MultiStage2.py script that generates the tables (it's not in
1697 the tarball, but is in the Subversion repository).
1698
1699 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1700 scripts.
1701
1702 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1703 a group with a zero qualifier. The result of the study could be incorrect,
1704 or the function might crash, depending on the pattern.
1705
1706 4. Caseless matching was not working for non-ASCII characters in back
1707 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1708 It now works when Unicode Property Support is available.
1709
1710 5. In pcretest, an escape such as \x{de} in the data was always generating
1711 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1712 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1713 truncation.
1714
1715 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1716
1717 7. Added two (int) casts to pcregrep when printing the difference of two
1718 pointers, in case they are 64-bit values.
1719
1720 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1721 test 2 if it fails.
1722
1723 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1724 and a #define of that name to empty if it is not externally set. This is to
1725 allow users of MSVC to set it if necessary.
1726
1727 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1728 the convenience functions in the pcre_get.c source file.
1729
1730 11. An option change at the start of a pattern that had top-level alternatives
1731 could cause overwriting and/or a crash. This command provoked a crash in
1732 some environments:
1733
1734 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1735
1736 This potential security problem was recorded as CVE-2008-2371.
1737
1738 12. For a pattern where the match had to start at the beginning or immediately
1739 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1740 pcre_dfa_exec() could read past the end of the passed subject if there was
1741 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1742 pcretest so that it places the subject at the end of its malloc-ed buffer.
1743
1744 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1745 exec() might read past the end of the data buffer in UTF-8 mode.
1746
1747 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1748 the data contained the byte 0x85 as part of a UTF-8 character within its
1749 first line. This applied both to normal and DFA matching.
1750
1751 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1752 /^[^d]*?$/8 failed to match "abc".
1753
1754 16. Added a missing copyright notice to pcrecpp_internal.h.
1755
1756 17. Make it more clear in the documentation that values returned from
1757 pcre_exec() in ovector are byte offsets, not character counts.
1758
1759 18. Tidied a few places to stop certain compilers from issuing warnings.
1760
1761 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1762 supplied by Stefan Weber. I made a further small update for 7.8 because
1763 there is a change of source arrangements: the pcre_searchfuncs.c module is
1764 replaced by pcre_ucd.c.
1765
1766
1767 Version 7.7 07-May-08
1768 ---------------------
1769
1770 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1771 a string to a long long, pretend we don't even have a long long." This is
1772 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1773
1774 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1775 pre-7.6 versions, which defined a global no_arg variable instead of putting
1776 it in the RE class. (See also #8 below.)
1777
1778 3. Remove a line of dead code, identified by coverity and reported by Nuno
1779 Lopes.
1780
1781 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1782
1783 (1) The include/exclude patterns were being applied to the whole pathnames
1784 of files, instead of just to the final components.
1785
1786 (2) If there was more than one level of directory, the subdirectories were
1787 skipped unless they satisfied the include/exclude conditions. This is
1788 inconsistent with GNU grep (and could even be seen as contrary to the
1789 pcregrep specification - which I improved to make it absolutely clear).
1790 The action now is always to scan all levels of directory, and just
1791 apply the include/exclude patterns to regular files.
1792
1793 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1794 --exclude_dir in the tests to avoid scanning .svn directories.
1795
1796 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1797 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1798 doesn't support NULs in patterns.
1799
1800 7. Added some missing "const"s to declarations of static tables in
1801 pcre_compile.c and pcre_dfa_exec.c.
1802
1803 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1804 caused by fix #2 above. (Subsequently also a second patch to fix the
1805 first patch. And a third patch - this was a messy problem.)
1806
1807 9. Applied Craig's patch to remove the use of push_back().
1808
1809 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1810 matching function regexec().
1811
1812 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1813 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1814 references. PCRE supports relative numbers with this syntax (I don't think
1815 Oniguruma does).
1816
1817 12. Previously, a group with a zero repeat such as (...){0} was completely
1818 omitted from the compiled regex. However, this means that if the group
1819 was called as a subroutine from elsewhere in the pattern, things went wrong
1820 (an internal error was given). Such groups are now left in the compiled
1821 pattern, with a new opcode that causes them to be skipped at execution
1822 time.
1823
1824 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1825 to the way PCRE behaves:
1826
1827 (a) A lone ] character is dis-allowed (Perl treats it as data).
1828
1829 (b) A back reference to an unmatched subpattern matches an empty string
1830 (Perl fails the current match path).
1831
1832 (c) A data ] in a character class must be notated as \] because if the
1833 first data character in a class is ], it defines an empty class. (In
1834 Perl it is not possible to have an empty class.) The empty class []
1835 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1836 The negative empty class [^] matches any one character, independently
1837 of the DOTALL setting.
1838
1839 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1840 non-existent subpattern following a character class starting with ']' and
1841 containing () gave an internal compiling error instead of "reference to
1842 non-existent subpattern". Fortunately, when the pattern did exist, the
1843 compiled code was correct. (When scanning forwards to check for the
1844 existencd of the subpattern, it was treating the data ']' as terminating
1845 the class, so got the count wrong. When actually compiling, the reference
1846 was subsequently set up correctly.)
1847
1848 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1849 it was being rejected as not supported by pcre_dfa_exec(), even though
1850 other assertions are supported. I have made pcre_dfa_exec() support
1851 (*FAIL).
1852
1853 16. The implementation of 13c above involved the invention of a new opcode,
1854 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1855 cannot be changed at match time, I realized I could make a small
1856 improvement to matching performance by compiling OP_ALLANY instead of
1857 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1858 on the OP_ANY path.
1859
1860 17. Compiling pcretest on Windows with readline support failed without the
1861 following two fixes: (1) Make the unistd.h include conditional on
1862 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1863
1864 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1865 ncurses library to be included for pcretest when ReadLine support is
1866 requested, but also to allow for it to be overridden. This patch came from
1867 Daniel Bergström.
1868
1869 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1870 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1871 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1872 spotting this.
1873
1874
1875 Version 7.6 28-Jan-08
1876 ---------------------
1877
1878 1. A character class containing a very large number of characters with
1879 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1880 overflow.
1881
1882 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1883 HAVE_LONG_LONG is not defined.
1884
1885 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1886 bring them up to date and include new features. This patch includes:
1887
1888 - Fixed PH's badly added libz and libbz2 support.
1889 - Fixed a problem with static linking.
1890 - Added pcredemo. [But later removed - see 7 below.]
1891 - Fixed dftables problem and added an option.
1892 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1893 HAVE_LONG_LONG.
1894 - Added readline support for pcretest.
1895 - Added an listing of the option settings after cmake has run.
1896
1897 4. A user submitted a patch to Makefile that makes it easy to create
1898 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1899 Makefile.am that cause it to include this special target, without
1900 affecting anything else. Note that the same mingw target plus all
1901 the other distribution libraries and programs are now supported
1902 when configuring with CMake (see 6 below) instead of with
1903 Configure/Make.
1904
1905 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1906 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1907 exported in the Windows port". It has not yet been confirmed that the patch
1908 solves the problem, but it does no harm.
1909
1910 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1911 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1912 with CMake, and also correct the comment about stack recursion.
1913
1914 7. Remove the automatic building of pcredemo from the ./configure system and
1915 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1916 of a program that users should build themselves after PCRE is installed, so
1917 building it automatically is not really right. What is more, it gave
1918 trouble in some build environments.
1919
1920 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1921
1922
1923 Version 7.5 10-Jan-08
1924 ---------------------
1925
1926 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1927 values in parens when parsing an RE using the C++ wrapper."
1928
1929 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1930 Characters greater than 255 were excluded from the class instead of being
1931 included.
1932
1933 3. The same bug as (2) above applied to negated POSIX classes such as
1934 [:^space:].
1935
1936 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1937 defined or documented. It seems to have been a typo for PCRE_STATIC, so
1938 I have changed it.
1939
1940 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
1941 first named subpattern) and a construct such as (?&a) would reference the
1942 first named subpattern whose name started with "a" (in other words, the
1943 length check was missing). Both these problems are fixed. "Subpattern name
1944 expected" is now given for (?&) (a zero-length name), and this patch also
1945 makes it give the same error for \k'' (previously it complained that that
1946 was a reference to a non-existent subpattern).
1947
1948 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
1949 this is right because (?- can be followed by option settings as well as by
1950 digits. I have, however, made the messages clearer.
1951
1952 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1953 than the number used in the conditional) now cause a compile-time error.
1954 This is actually not compatible with Perl, which accepts such patterns, but
1955 treats the conditional as always being FALSE (as PCRE used to), but it
1956 seems to me that giving a diagnostic is better.
1957
1958 8. Change "alphameric" to the more common word "alphanumeric" in comments
1959 and messages.
1960
1961 9. Fix two occurrences of "backslash" in comments that should have been
1962 "backspace".
1963
1964 10. Remove two redundant lines of code that can never be obeyed (their function
1965 was moved elsewhere).
1966
1967 11. The program that makes PCRE's Unicode character property table had a bug
1968 which caused it to generate incorrect table entries for sequences of
1969 characters that have the same character type, but are in different scripts.
1970 It amalgamated them into a single range, with the script of the first of
1971 them. In other words, some characters were in the wrong script. There were
1972 thirteen such cases, affecting characters in the following ranges:
1973
1974 U+002b0 - U+002c1
1975 U+0060c - U+0060d
1976 U+0061e - U+00612
1977 U+0064b - U+0065e
1978 U+0074d - U+0076d
1979 U+01800 - U+01805
1980 U+01d00 - U+01d77
1981 U+01d9b - U+01dbf
1982 U+0200b - U+0200f
1983 U+030fc - U+030fe
1984 U+03260 - U+0327f
1985 U+0fb46 - U+0fbb1
1986 U+10450 - U+1049d
1987
1988 12. The -o option (show only the matching part of a line) for pcregrep was not
1989 compatible with GNU grep in that, if there was more than one match in a
1990 line, it showed only the first of them. It now behaves in the same way as
1991 GNU grep.
1992
1993 13. If the -o and -v options were combined for pcregrep, it printed a blank
1994 line for every non-matching line. GNU grep prints nothing, and pcregrep now
1995 does the same. The return code can be used to tell if there were any
1996 non-matching lines.
1997
1998 14. Added --file-offsets and --line-offsets to pcregrep.
1999
2000 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2001 infinitely looping recursion. The bug was that positive lookaheads were not
2002 being skipped when checking for a possible empty match (negative lookaheads
2003 and both kinds of lookbehind were skipped).
2004
2005 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2006 inclusion of <windows.h> to before rather than after the definition of
2007 INVALID_FILE_ATTRIBUTES (patch from David Byron).
2008
2009 17. Specifying a possessive quantifier with a specific limit for a Unicode
2010 character property caused pcre_compile() to compile bad code, which led at
2011 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2012 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2013 caused the error; without that there was no problem.
2014
2015 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2016
2017 19. Added --enable-pcretest-libreadline.
2018
2019 20. In pcrecpp.cc, the variable 'count' was incremented twice in
2020 RE::GlobalReplace(). As a result, the number of replacements returned was
2021 double what it should be. I removed one of the increments, but Craig sent a
2022 later patch that removed the other one (the right fix) and added unit tests
2023 that check the return values (which was not done before).
2024
2025 21. Several CMake things:
2026
2027 (1) Arranged that, when cmake is used on Unix, the libraries end up with
2028 the names libpcre and libpcreposix, not just pcre and pcreposix.
2029
2030 (2) The above change means that pcretest and pcregrep are now correctly
2031 linked with the newly-built libraries, not previously installed ones.
2032
2033 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2034
2035 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2036 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2037 UTF-8 newline character). The key issue is that the pattern starts .*;
2038 this means that the match must be either at the beginning, or after a
2039 newline. The bug was in the code for advancing after a failed match and
2040 checking that the new position followed a newline. It was not taking
2041 account of UTF-8 characters correctly.
2042
2043 23. PCRE was behaving differently from Perl in the way it recognized POSIX
2044 character classes. PCRE was not treating the sequence [:...:] as a
2045 character class unless the ... were all letters. Perl, however, seems to
2046 allow any characters between [: and :], though of course it rejects as
2047 unknown any "names" that contain non-letters, because all the known class
2048 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2049 for example, whereas PCRE did not - it did not recognize a POSIX character
2050 class. This seemed a bit dangerous, so the code has been changed to be
2051 closer to Perl. The behaviour is not identical to Perl, because PCRE will
2052 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2053 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2054 Perl does, and where it didn't before.
2055
2056 24. Rewrite so as to remove the single use of %n from pcregrep because in some
2057 Windows environments %n is disabled by default.
2058
2059
2060 Version 7.4 21-Sep-07
2061 ---------------------
2062
2063 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
2064 means that a class such as [\s] counted as "explicit reference to CR or
2065 LF". That isn't really right - the whole point of the change was to try to
2066 help when there was an actual mention of one of the two characters. So now
2067 the change happens only if \r or \n (or a literal CR or LF) character is
2068 encountered.
2069
2070 2. The 32-bit options word was also used for 6 internal flags, but the numbers
2071 of both had grown to the point where there were only 3 bits left.
2072 Fortunately, there was spare space in the data structure, and so I have
2073 moved the internal flags into a new 16-bit field to free up more option
2074 bits.
2075
2076 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2077 but did not set the internal JCHANGED flag - either of these is enough to
2078 control the way the "get" function works - but the PCRE_INFO_JCHANGED
2079 facility is supposed to tell if (?J) was ever used, so now (?J) at the
2080 start sets both bits.
2081
2082 4. Added options (at build time, compile time, exec time) to change \R from
2083 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2084
2085 5. doc/pcresyntax.html was missing from the distribution.
2086
2087 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2088 compatibility, even though it is no longer used.
2089
2090 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2091 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2092 windows.h file is present (where different names are used). [This was
2093 reversed later after testing - see 16 below.]
2094
2095 8. Changed all #include <config.h> to #include "config.h". There were also
2096 some further <pcre.h> cases that I changed to "pcre.h".
2097
2098 9. When pcregrep was used with the --colour option, it missed the line ending
2099 sequence off the lines that it output.
2100
2101 10. It was pointed out to me that arrays of string pointers cause lots of
2102 relocations when a shared library is dynamically loaded. A technique of
2103 using a single long string with a table of offsets can drastically reduce
2104 these. I have refactored PCRE in four places to do this. The result is
2105 dramatic:
2106
2107 Originally: 290
2108 After changing UCP table: 187
2109 After changing error message table: 43
2110 After changing table of "verbs" 36
2111 After changing table of Posix names 22
2112
2113 Thanks to the folks working on Gregex for glib for this insight.
2114
2115 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2116 unicode-properties was also set.
2117
2118 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2119
2120 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2121 checked only for CRLF.
2122
2123 14. Added casts to pcretest.c to avoid compiler warnings.
2124
2125 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2126
2127 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2128 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2129 entirely. This removes changes made in 7 above.
2130
2131 17. The CMake files have been updated, and there is now more information about
2132 building with CMake in the NON-UNIX-USE document.
2133
2134
2135 Version 7.3 28-Aug-07
2136 ---------------------
2137
2138 1. In the rejigging of the build system that eventually resulted in 7.1, the
2139 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2140 brackets there is not right, since it causes compilers to look for an
2141 installed pcre.h, not the version that is in the source that is being
2142 compiled (which of course may be different). I have changed it back to:
2143
2144 #include "pcre.h"
2145
2146 I have a vague recollection that the change was concerned with compiling in
2147 different directories, but in the new build system, that is taken care of
2148 by the VPATH setting the Makefile.
2149
2150 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2151 when the subject happened to end in the byte 0x85 (e.g. if the last
2152 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2153 characters but of course it shouldn't be taken as a newline when it is part
2154 of another character. The bug was that, for an unlimited repeat of . in
2155 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2156 characters when looking for a newline.
2157
2158 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2159
2160 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2161 in debug output.
2162
2163 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2164 long printing in the pcrecpp unittest when running under MinGW.
2165
2166 6. ESC_K was left out of the EBCDIC table.
2167
2168 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2169 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2170 limit also applies to "virtual nesting" when a pattern is recursive, and in
2171 this case 1000 isn't so big. I have been able to remove this limit at the
2172 expense of backing off one optimization in certain circumstances. Normally,
2173 when pcre_exec() would call its internal match() function recursively and
2174 immediately return the result unconditionally, it uses a "tail recursion"
2175 feature to save stack. However, when a subpattern that can match an empty
2176 string has an unlimited repetition quantifier, it no longer makes this
2177 optimization. That gives it a stack frame in which to save the data for
2178 checking that an empty string has been matched. Previously this was taken
2179 from the 1000-entry workspace that had been reserved. So now there is no
2180 explicit limit, but more stack is used.
2181
2182 8. Applied Daniel's patches to solve problems with the import/export magic
2183 syntax that is required for Windows, and which was going wrong for the
2184 pcreposix and pcrecpp parts of the library. These were overlooked when this
2185 problem was solved for the main library.
2186
2187 9. There were some crude static tests to avoid integer overflow when computing
2188 the size of patterns that contain repeated groups with explicit upper
2189 limits. As the maximum quantifier is 65535, the maximum group length was
2190 set at 30,000 so that the product of these two numbers did not overflow a
2191 32-bit integer. However, it turns out that people want to use groups that
2192 are longer than 30,000 bytes (though not repeat them that many times).
2193 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2194 made it possible to implement the integer overflow checks in a much more
2195 dynamic way, which I have now done. The artificial limitation on group
2196 length has been removed - we now have only the limit on the total length of
2197 the compiled pattern, which depends on the LINK_SIZE setting.
2198
2199 10. Fixed a bug in the documentation for get/copy named substring when
2200 duplicate names are permitted. If none of the named substrings are set, the
2201 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2202 empty string.
2203
2204 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2205 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2206 because the ] is interpreted as the first data character and the
2207 terminating ] is not found. PCRE has been made compatible with Perl in this
2208 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2209 cause memory overwriting.
2210
2211 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2212 string has been matched (to stop an infinite loop). It was not recognizing
2213 a conditional subpattern that could match an empty string if that
2214 subpattern was within another subpattern. For example, it looped when
2215 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
2216 condition was not nested. This bug has been fixed.
2217
2218 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2219 past the start of the subject in the presence of bytes with the top bit
2220 set, for example "\x8aBCD".
2221
2222 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2223 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2224
2225 14. Optimized (?!) to (*FAIL).
2226
2227 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2228 This restricts code points to be within the range 0 to 0x10FFFF, excluding
2229 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2230 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2231 does: it's just the validity check that is more restrictive.
2232
2233 16. Inserted checks for integer overflows during escape sequence (backslash)
2234 processing, and also fixed erroneous offset values for syntax errors during
2235 backslash processing.
2236
2237 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2238 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2239
2240 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2241 caused an overrun.
2242
2243 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2244 something other than just ASCII characters) inside a group that had an
2245 unlimited repeat caused a loop at compile time (while checking to see
2246 whether the group could match an empty string).
2247
2248 20. Debugging a pattern containing \p or \P could cause a crash. For example,
2249 [\P{Any}] did so. (Error in the code for printing property names.)
2250
2251 21. An orphan \E inside a character class could cause a crash.
2252
2253 22. A repeated capturing bracket such as (A)? could cause a wild memory
2254 reference during compilation.
2255
2256 23. There are several functions in pcre_compile() that scan along a compiled
2257 expression for various reasons (e.g. to see if it's fixed length for look
2258 behind). There were bugs in these functions when a repeated \p or \P was
2259 present in the pattern. These operators have additional parameters compared
2260 with \d, etc, and these were not being taken into account when moving along
2261 the compiled data. Specifically:
2262
2263 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2264 length.
2265
2266 (b) An item such as \pL+ within a repeated group could cause crashes or
2267 loops.
2268
2269 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2270 "reference to non-existent subpattern" error.
2271
2272 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2273
2274 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2275 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2276
2277 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2278
2279 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2280 character were causing crashes (broken optimization).
2281
2282 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2283 \p or \P) caused a compile-time loop.
2284
2285 28. More problems have arisen in unanchored patterns when CRLF is a valid line
2286 break. For example, the unstudied pattern [\r\n]A does not match the string
2287 "\r\nA" because change 7.0/46 below moves the current point on by two
2288 characters after failing to match at the start. However, the pattern \nA
2289 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2290 the same is true. There doesn't seem any very clean way out of this, but
2291 what I have chosen to do makes the common cases work: PCRE now takes note
2292 of whether there can be an explicit match for \r or \n anywhere in the
2293 pattern, and if so, 7.0/46 no longer applies. As part of this change,
2294 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2295 pattern has explicit CR or LF references.
2296
2297 29. Added (*CR) etc for changing newline setting at start of pattern.
2298
2299
2300 Version 7.2 19-Jun-07
2301 ---------------------
2302
2303 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2304 which is apparently normally available under Windows.
2305
2306 2. Re-jig the pcregrep tests with different newline settings in an attempt
2307 to make them independent of the local environment's newline setting.
2308
2309 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2310
2311 4. Some of the "internals" tests were previously cut out when the link size
2312 was not 2, because the output contained actual offsets. The recent new
2313 "Z" feature of pcretest means that these can be cut out, making the tests
2314 usable with all link sizes.
2315
2316 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2317 stack recursion. This gives a massive performance boost under BSD, but just
2318 a small improvement under Linux. However, it saves one field in the frame
2319 in all cases.
2320
2321 6. Added more features from the forthcoming Perl 5.10:
2322
2323 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2324 recursion call. It refers to the nth most recently opened parentheses.
2325
2326 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2327 to be opened parentheses.
2328
2329 (c) Conditions that refer to capturing parentheses can be specified
2330 relatively, for example, (?(-2)... or (?(+3)...
2331
2332 (d) \K resets the start of the current match so that everything before
2333 is not part of it.
2334
2335 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2336
2337 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2338 reference syntax.
2339
2340 (g) (?| introduces a group in which the numbering of parentheses in each
2341 alternative starts with the same number.
2342
2343 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2344
2345 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2346 PCRE_INFO_JCHANGED.
2347
2348 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2349 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2350 for detecting groups that can match an empty string.
2351
2352 9. A pattern with a very large number of alternatives (more than several
2353 hundred) was running out of internal workspace during the pre-compile
2354 phase, where pcre_compile() figures out how much memory will be needed. A
2355 bit of new cunning has reduced the workspace needed for groups with
2356 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2357 workspace instead of running out of the 4096 that are available.
2358
2359 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2360
2361 11. Applied patch from Google to remove an optimization that didn't quite work.
2362 The report of the bug said:
2363
2364 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2365 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2366 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2367
2368 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2369 it matched the wrong number of bytes.
2370
2371
2372 Version 7.1 24-Apr-07
2373 ---------------------
2374
2375 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2376 that is more "standard", making use of automake and other Autotools. There
2377 is some re-arrangement of the files and adjustment of comments consequent
2378 on this.
2379
2380 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2381 for recursive directory scanning broke on some systems because the files
2382 are not scanned in any specific order and on different systems the order
2383 was different. A call to "sort" has been inserted into RunGrepTest for the
2384 approprate test as a short-term fix. In the longer term there may be an
2385 alternative.
2386
2387 3. I had an email from Eric Raymond about problems translating some of PCRE's
2388 man pages to HTML (despite the fact that I distribute HTML pages, some
2389 people do their own conversions for various reasons). The problems
2390 concerned the use of low-level troff macros .br and .in. I have therefore
2391 removed all such uses from the man pages (some were redundant, some could
2392 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2393 HTML has been updated to handle .nf/.fi and to complain if it encounters
2394 .br or .in.
2395
2396 4. Updated comments in configure.ac that get placed in config.h.in and also
2397 arranged for config.h to be included in the distribution, with the name
2398 config.h.generic, for the benefit of those who have to compile without
2399 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2400
2401 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2402 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2403 makevp.bat for the current PCRE, using the additional files
2404 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2405
2406 6. A Windows user reported a minor discrepancy with test 2, which turned out
2407 to be caused by a trailing space on an input line that had got lost in his
2408 copy. The trailing space was an accident, so I've just removed it.
2409
2410 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2411 that is needed.
2412
2413 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2414 as "const" (a) because they are and (b) because it helps the PHP
2415 maintainers who have recently made a script to detect big data structures
2416 in the php code that should be moved to the .rodata section. I remembered
2417 to update Builducptable as well, so it won't revert if ucptable.h is ever
2418 re-created.
2419
2420 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2421 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2422 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2423 support is not required. This saves 1.5-2K of code, which is important in
2424 some applications.
2425
2426 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2427 so as not to refer to the tables, even though these functions will never be
2428 called when UTF-8 support is disabled. Otherwise there are problems with a
2429 shared library.
2430
2431 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2432
2433 (a) It was defining its arguments as char * instead of void *.
2434
2435 (b) It was assuming that all moves were upwards in memory; this was true
2436 a long time ago when I wrote it, but is no longer the case.
2437
2438 The emulated memove() is provided for those environments that have neither
2439 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2440 is clearly not the case, as these two bugs were recently reported.
2441
2442 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2443 and Detrail to create the HTML documentation, the .txt form of the man
2444 pages, and it removes trailing spaces from listed files. It also creates
2445 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2446 case, it wraps all the #defines with #ifndefs. This script should be run
2447 before "make dist".
2448
2449 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2450 with Unicode property support.
2451
2452 (a) For a maximizing quantifier, if the two different cases of the
2453 character were of different lengths in their UTF-8 codings (there are
2454 some cases like this - I found 11), and the matching function had to
2455 back up over a mixture of the two cases, it incorrectly assumed they
2456 were both the same length.
2457
2458 (b) When PCRE was configured to use the heap rather than the stack for
2459 recursion during matching, it was not correctly preserving the data for
2460 the other case of a UTF-8 character when checking ahead for a match
2461 while processing a minimizing repeat. If the check also involved
2462 matching a wide character, but failed, corruption could cause an
2463 erroneous result when trying to check for a repeat of the original
2464 character.
2465
2466 13. Some tidying changes to the testing mechanism:
2467
2468 (a) The RunTest script now detects the internal link size and whether there
2469 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2470 values substituted by "configure". (The RunGrepTest script already did
2471 this for UTF-8.) The configure.ac script no longer substitutes the
2472 relevant variables.
2473
2474 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2475 with length and offset values. This means that the output is different
2476 for different internal link sizes. Test 2 is skipped for link sizes
2477 other than 2 because of this, bypassing the problem. Unfortunately,
2478 there was also a test in test 3 (the locale tests) that used /B and
2479 failed for link sizes other than 2. Rather than cut the whole test out,
2480 I have added a new /Z option to pcretest that replaces the length and
2481 offset values with spaces. This is now used to make test 3 independent
2482 of link size. (Test 2 will be tidied up later.)
2483
2484 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2485 segmentation fault instead of returning the appropriate error message.
2486
2487 15. In multiline mode when the newline sequence was set to "any", the pattern
2488 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2489 This doesn't seem right; it now treats the CRLF combination as the line
2490 ending, and so does not match in that case. It's only a pattern such as ^$
2491 that would hit this one: something like ^ABC$ would have failed after \r
2492 and then tried again after \r\n.
2493
2494 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2495 in an attempt to make files that differ only in their line terminators
2496 compare equal. This works on Linux.
2497
2498 17. Under certain error circumstances pcregrep might try to free random memory
2499 as it exited. This is now fixed, thanks to valgrind.
2500
2501 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2502 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2503 was because its rules for how to advance for /g after matching an empty
2504 string at the end of a line did not allow for this case. They now check for
2505 it specially.
2506
2507 20. pcretest is supposed to handle patterns and data of any length, by
2508 extending its buffers when necessary. It was getting this wrong when the
2509 buffer for a data line had to be extended.
2510
2511 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2512 CRLF as a newline sequence.
2513
2514 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2515 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2516 I have nevertheless tidied it up.
2517
2518 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2519
2520 24. Added a man page for pcre-config.
2521
2522
2523 Version 7.0 19-Dec-06
2524 ---------------------
2525
2526 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2527 moving to gcc 4.1.1.
2528
2529 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2530 sys/time.h, which is documented as needed for this function. It doesn't
2531 seem to matter on Linux, but it showed up on some releases of OS X.
2532
2533 3. It seems that there are systems where bytes whose values are greater than
2534 127 match isprint() in the "C" locale. The "C" locale should be the
2535 default when a C program starts up. In most systems, only ASCII printing
2536 characters match isprint(). This difference caused the output from pcretest
2537 to vary, making some of the tests fail. I have changed pcretest so that:
2538
2539 (a) When it is outputting text in the compiled version of a pattern, bytes
2540 other than 32-126 are always shown as hex escapes.
2541
2542 (b) When it is outputting text that is a matched part of a subject string,
2543 it does the same, unless a different locale has been set for the match
2544 (using the /L modifier). In this case, it uses isprint() to decide.
2545
2546 4. Fixed a major bug that caused incorrect computation of the amount of memory
2547 required for a compiled pattern when options that changed within the
2548 pattern affected the logic of the preliminary scan that determines the
2549 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2550 that the computed length was too small. The symptoms of this bug were
2551 either the PCRE error "internal error: code overflow" from pcre_compile(),
2552 or a glibc crash with a message such as "pcretest: free(): invalid next
2553 size (fast)". Examples of patterns that provoked this bug (shown in
2554 pcretest format) are:
2555
2556 /(?-x: )/x
2557 /(?x)(?-x: \s*#\s*)/
2558 /((?i)[\x{c0}])/8
2559 /(?i:[\x{c0}])/8
2560
2561 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2562 is now done differently.
2563
2564 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2565 wrapper classes; (b) implement a new function in the C++ scanner that is
2566 more efficient than the old way of doing things because it avoids levels of
2567 recursion in the regex matching; (c) add a paragraph to the documentation
2568 for the FullMatch() function.
2569
2570 6. The escape sequence \n was being treated as whatever was defined as
2571 "newline". Not only was this contrary to the documentation, which states
2572 that \n is character 10 (hex 0A), but it also went horribly wrong when
2573 "newline" was defined as CRLF. This has been fixed.
2574
2575 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2576 was being set to -1 for the "end of line" case (supposedly a value that no
2577 character can have). Though this value is never used (the check for end of
2578 line is "zero bytes in current character"), it caused compiler complaints.
2579 I've changed it to 0xffffffff.
2580
2581 8. In pcre_version.c, the version string was being built by a sequence of
2582 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2583 string (as it is for production releases) called a macro with an empty
2584 argument. The C standard says the result of this is undefined. The gcc
2585 compiler treats it as an empty string (which was what was wanted) but it is
2586 reported that Visual C gives an error. The source has been hacked around to
2587 avoid this problem.
2588
2589 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2590 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2591 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2592 of them did).
2593
2594 10. Originally, pcretest opened its input and output without "b"; then I was
2595 told that "b" was needed in some environments, so it was added for release
2596 5.0 to both the input and output. (It makes no difference on Unix-like
2597 systems.) Later I was told that it is wrong for the input on Windows. I've
2598 now abstracted the modes into two macros, to make it easier to fiddle with
2599 them, and removed "b" from the input mode under Windows.
2600
2601 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2602
2603 12. Added -help and --help to pcretest as an official way of being reminded
2604 of the options.
2605
2606 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2607 and pcrecpp.cc because they annoy compilers at high warning levels.
2608
2609 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2610
2611 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2612 scripts are not C programs :-) and which was not noticed because it works
2613 on Linux.
2614
2615 16. pcretest is supposed to handle any length of pattern and data line (as one
2616 line or as a continued sequence of lines) by extending its input buffer if
2617 necessary. This feature was broken for very long pattern lines, leading to
2618 a string of junk being passed to pcre_compile() if the pattern was longer
2619 than about 50K.
2620
2621 17. I have done a major re-factoring of the way pcre_compile() computes the
2622 amount of memory needed for a compiled pattern. Previously, there was code
2623 that made a preliminary scan of the pattern in order to do this. That was
2624 OK when PCRE was new, but as the facilities have expanded, it has become
2625 harder and harder to keep it in step with the real compile phase, and there
2626 have been a number of bugs (see for example, 4 above). I have now found a
2627 cunning way of running the real compile function in a "fake" mode that
2628 enables it to compute how much memory it would need, while actually only
2629 ever using a few hundred bytes of working memory and without too many
2630 tests of the mode. This should make future maintenance and development
2631 easier. A side effect of this work is that the limit of 200 on the nesting
2632 depth of parentheses has been removed (though this was never a serious
2633 limitation, I suspect). However, there is a downside: pcre_compile() now
2634 runs more slowly than before (30% or more, depending on the pattern). I
2635 hope this isn't a big issue. There is no effect on runtime performance.
2636
2637 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2638 newline (only possible for the last line of a file) and it was a
2639 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2640
2641 19. Added additional timing features to pcretest. (1) The -tm option now times
2642 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2643 separate command line item, by a number that specifies the number of
2644 repeats to use when timing. The default is 50000; this gives better
2645 precision, but takes uncomfortably long for very large patterns.
2646
2647 20. Extended pcre_study() to be more clever in cases where a branch of a
2648 subpattern has no definite first character. For example, (a*|b*)[cd] would
2649 previously give no result from pcre_study(). Now it recognizes that the
2650 first character must be a, b, c, or d.
2651
2652 21. There was an incorrect error "recursive call could loop indefinitely" if
2653 a subpattern (or the entire pattern) that was being tested for matching an
2654 empty string contained only one non-empty item after a nested subpattern.
2655 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2656 incorrectly, because the \d was being skipped in the check.
2657
2658 22. The pcretest program now has a new pattern option /B and a command line
2659 option -b, which is equivalent to adding /B to every pattern. This causes
2660 it to show the compiled bytecode, without the additional information that
2661 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2662 is the same as /B/I).
2663
2664 23. A new optimization is now able automatically to treat some sequences such
2665 as a*b as a*+b. More specifically, if something simple (such as a character
2666 or a simple class like \d) has an unlimited quantifier, and is followed by
2667 something that cannot possibly match the quantified thing, the quantifier
2668 is automatically "possessified".
2669
2670 24. A recursive reference to a subpattern whose number was greater than 39
2671 went wrong under certain circumstances in UTF-8 mode. This bug could also
2672 have affected the operation of pcre_study().
2673
2674 25. Realized that a little bit of performance could be had by replacing
2675 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2676
2677 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2678
2679 27. Possessive quantifiers such as a++ were previously implemented by turning
2680 them into atomic groups such as ($>a+). Now they have their own opcodes,
2681 which improves performance. This includes the automatically created ones
2682 from 23 above.
2683
2684 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2685 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2686 the first matched character to be a colon. This applied both to named and
2687 numbered groups.
2688
2689 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2690
2691 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2692 building PCRE on Windows easier, so I have included it in the distribution.
2693
2694 31. There is now a check in pcretest against a ridiculously large number being
2695 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2696 loop, the loop is abandoned.
2697
2698 32. Forward references to subpatterns in conditions such as (?(2)...) where
2699 subpattern 2 is defined later cause pcre_compile() to search forwards in
2700 the pattern for the relevant set of parentheses. This search went wrong
2701 when there were unescaped parentheses in a character class, parentheses
2702 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2703
2704 33. "Subroutine" calls and backreferences were previously restricted to
2705 referencing subpatterns earlier in the regex. This restriction has now
2706 been removed.
2707
2708 34. Added a number of extra features that are going to be in Perl 5.10. On the
2709 whole, these are just syntactic alternatives for features that PCRE had
2710 previously implemented using the Python syntax or my own invention. The
2711 other formats are all retained for compatibility.
2712
2713 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2714 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2715 also .NET compatible.
2716
2717 (b) A recursion or subroutine call to a named group can now be defined as
2718 (?&name) as well as (?P>name).
2719
2720 (c) A backreference to a named group can now be defined as \k<name> or
2721 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2722 5.10, are also .NET compatible.
2723
2724 (d) A conditional reference to a named group can now use the syntax
2725 (?(<name>) or (?('name') as well as (?(name).
2726
2727 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2728 groups (named and numbered) that are never evaluated inline, but can be
2729 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2730 is always false. There may be only one alternative in such a group.
2731
2732 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2733 as the simple (?(R). The condition is true only if the most recent
2734 recursion is that of the given number or name. It does not search out
2735 through the entire recursion stack.
2736
2737 (g) The escape \gN or \g{N} has been added, where N is a positive or
2738 negative number, specifying an absolute or relative reference.
2739
2740 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2741 some "unreachable code" warnings.
2742
2743 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2744 things, this adds five new scripts.
2745
2746 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2747 There were also incompatibilities regarding the handling of \Q..\E inside
2748 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2749 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2750
2751 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2752 matches an empty string, and forcibly breaks the loop. There were bugs in
2753 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2754 against aaaa the result was just "a" rather than "aaaa", for example. Two
2755 separate and independent bugs (that affected different cases) have been
2756 fixed.
2757
2758 39. Refactored the code to abolish the use of different opcodes for small
2759 capturing bracket numbers. This is a tidy that I avoided doing when I
2760 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2761 The new approach is not only tidier, it makes it possible to reduce the
2762 memory needed to fix the previous bug (38).
2763
2764 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2765 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2766 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2767 mode.
2768
2769 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2770 report.
2771
2772 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2773 copy construction and assignment in the C++ wrapper.
2774
2775 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2776 couple of bugs that could have given wrong results in the "--newline=crlf"
2777 case.
2778
2779 44. Added a number of casts and did some reorganization of signed/unsigned int
2780 variables following suggestions from Dair Grant. Also renamed the variable
2781 "this" as "item" because it is a C++ keyword.
2782
2783 45. Arranged for dftables to add
2784
2785 #include "pcre_internal.h"
2786
2787 to pcre_chartables.c because without it, gcc 4.x may remove the array
2788 definition from the final binary if PCRE is built into a static library and
2789 dead code stripping is activated.
2790
2791 46. For an unanchored pattern, if a match attempt fails at the start of a
2792 newline sequence, and the newline setting is CRLF or ANY, and the next two
2793 characters are CRLF, advance by two characters instead of one.
2794
2795
2796 Version 6.7 04-Jul-06
2797 ---------------------
2798
2799 1. In order to handle tests when input lines are enormously long, pcretest has
2800 been re-factored so that it automatically extends its buffers when
2801 necessary. The code is crude, but this _is_ just a test program. The
2802 default size has been increased from 32K to 50K.
2803
2804 2. The code in pcre_study() was using the value of the re argument before
2805 testing it for NULL. (Of course, in any sensible call of the function, it
2806 won't be NULL.)
2807
2808 3. The memmove() emulation function in pcre_internal.h, which is used on
2809 systems that lack both memmove() and bcopy() - that is, hardly ever -
2810 was missing a "static" storage class specifier.
2811
2812 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2813 containing an extended class (one that cannot be represented by a bitmap
2814 because it contains high-valued characters or Unicode property items, e.g.
2815 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2816 pattern, but PCRE should not loop if you do not (it no longer does).
2817 [Detail: two cases were found: (a) a repeated subpattern containing an
2818 extended class; (b) a recursive reference to a subpattern that followed a
2819 previous extended class. It wasn't skipping over the extended class
2820 correctly when UTF-8 mode was not set.]
2821
2822 5. A negated single-character class was not being recognized as fixed-length
2823 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2824 compile error "lookbehind assertion is not fixed length".
2825
2826 6. The RunPerlTest auxiliary script was showing an unexpected difference
2827 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2828 write a Perl script that can interpret lines of an input file either as
2829 byte characters or as UTF-8, which is what "perltest" was being required to
2830 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2831 can't do is switch easily at run time between having the "use utf8;" pragma
2832 or not. In the end, I fudged it by using the RunPerlTest script to insert
2833 "use utf8;" explicitly for the UTF-8 tests.
2834
2835 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2836 the end of the subject string, contrary to the documentation and to what
2837 Perl does. This was true of both matching functions. Now it matches only at
2838 the start of the subject and immediately after *internal* newlines.
2839
2840 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2841 a pointer to an int instead of a pointer to an unsigned long int. This
2842 caused problems on 64-bit systems.
2843
2844 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2845 instance of the 'standard' template library not being so standard".
2846
2847 10. There was no check on the number of named subpatterns nor the maximum
2848 length of a subpattern name. The product of these values is used to compute
2849 the size of the memory block for a compiled pattern. By supplying a very
2850 long subpattern name and a large number of named subpatterns, the size
2851 computation could be caused to overflow. This is now prevented by limiting
2852 the length of names to 32 characters, and the number of named subpatterns
2853 to 10,000.
2854
2855 11. Subpatterns that are repeated with specific counts have to be replicated in
2856 the compiled pattern. The size of memory for this was computed from the
2857 length of the subpattern and the repeat count. The latter is limited to
2858 65535, but there was no limit on the former, meaning that integer overflow
2859 could in principle occur. The compiled length of a repeated subpattern is
2860 now limited to 30,000 bytes in order to prevent this.
2861
2862 12. Added the optional facility to have named substrings with the same name.
2863
2864 13. Added the ability to use a named substring as a condition, using the
2865 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2866 are numbers (not recommended). Forward references are permitted.
2867
2868 14. Added forward references in named backreferences (if you see what I mean).
2869
2870 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2871 pattern could run off the end of the subject. For example, the pattern
2872 "(?s)(.{1,5})"8 did this with the subject "ab".
2873
2874 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2875 PCRE_CASELESS was set when matching characters that were quantified with ?
2876 or *.
2877
2878 17. A character class other than a single negated character that had a minimum
2879 but no maximum quantifier - for example [ab]{6,} - was not handled
2880 correctly by pce_dfa_exec(). It would match only one character.
2881
2882 18. A valid (though odd) pattern that looked like a POSIX character
2883 class but used an invalid character after [ (for example [[,abc,]]) caused
2884 pcre_compile() to give the error "Failed: internal error: code overflow" or
2885 in some cases to crash with a glibc free() error. This could even happen if
2886 the pattern terminated after [[ but there just happened to be a sequence of
2887 letters, a binary zero, and a closing ] in the memory that followed.
2888
2889 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2890 over the years. Originally (before any Unicode support), just the bottom 8
2891 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2892 output from "man perlunicode" includes this:
2893
2894 The regular expression compiler produces polymorphic opcodes. That
2895 is, the pattern adapts to the data and automatically switches to
2896 the Unicode character scheme when presented with Unicode data--or
2897 instead uses a traditional byte scheme when presented with byte
2898 data.
2899
2900 Sadly, a wide octal escape does not cause a switch, and in a string with
2901 no other multibyte characters, these octal escapes are treated as before.
2902 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2903 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2904 Unicode string.
2905
2906 I have not perpetrated such confusion in PCRE. Up till now, it took just
2907 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2908 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2909 translate to the appropriate multibyte character.
2910
2911 29. Applied some refactoring to reduce the number of warnings from Microsoft
2912 and Borland compilers. This has included removing the fudge introduced
2913 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2914 a warning about an unused variable.
2915
2916 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2917 characters since release 4.0, because Perl (from release 5.004) does not.
2918 [Or at least, is documented not to: some releases seem to be in conflict
2919 with the documentation.] However, when a pattern was studied with
2920 pcre_study() and all its branches started with \s, PCRE still included VT
2921 as a possible starting character. Of course, this did no harm; it just
2922 caused an unnecessary match attempt.
2923
2924 22. Removed a now-redundant internal flag bit that recorded the fact that case
2925 dependency changed within the pattern. This was once needed for "required
2926 byte" processing, but is no longer used. This recovers a now-scarce options
2927 bit. Also moved the least significant internal flag bit to the most-
2928 significant bit of the word, which was not previously used (hangover from
2929 the days when it was an int rather than a uint) to free up another bit for
2930 the future.
2931
2932 23. Added support for CRLF line endings as well as CR and LF. As well as the
2933 default being selectable at build time, it can now be changed at runtime
2934 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2935 specify that it is scanning data with non-default line endings.
2936
2937 24. Changed the definition of CXXLINK to make it agree with the definition of
2938 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2939
2940 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2941 recursions. This makes a big different to stack usage for some patterns.
2942
2943 26. If a subpattern containing a named recursion or subroutine reference such
2944 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2945 the space required for the compiled pattern went wrong and gave too small a
2946 value. Depending on the environment, this could lead to "Failed: internal
2947 error: code overflow at offset 49" or "glibc detected double free or
2948 corruption" errors.
2949
2950 27. Applied patches from Google (a) to support the new newline modes and (b) to
2951 advance over multibyte UTF-8 characters in GlobalReplace.
2952
2953 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2954 difference for some implementation of PCRE in some Windows version.
2955
2956 29. Added some extra testing facilities to pcretest:
2957
2958 \q<number> in a data line sets the "match limit" value
2959 \Q<number> in a data line sets the "match recursion limt" value
2960 -S <number> sets the stack size, where <number> is in megabytes
2961
2962 The -S option isn't available for Windows.
2963
2964
2965 Version 6.6 06-Feb-06
2966 ---------------------
2967
2968 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2969 in pcreposix.h. I have copied the definition from pcre.h.
2970
2971 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2972 because pcre.h is no longer a built file.
2973
2974 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2975 not normally included in the compiled code.
2976
2977
2978 Version 6.5 01-Feb-06
2979 ---------------------
2980
2981 1. When using the partial match feature with pcre_dfa_exec(), it was not
2982 anchoring the second and subsequent partial matches at the new starting
2983 point. This could lead to incorrect results. For example, with the pattern
2984 /1234/, partially matching against "123" and then "a4" gave a match.
2985
2986 2. Changes to pcregrep:
2987
2988 (a) All non-match returns from pcre_exec() were being treated as failures
2989 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2990 error message is output. Some extra information is given for the
2991 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2992 probably the only errors that are likely to be caused by users (by
2993 specifying a regex that has nested indefinite repeats, for instance).
2994 If there are more than 20 of these errors, pcregrep is abandoned.
2995
2996 (b) A binary zero was treated as data while matching, but terminated the
2997 output line if it was written out. This has been fixed: binary zeroes
2998 are now no different to any other data bytes.
2999
3000 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3001 used to set a locale for matching. The --locale=xxxx long option has
3002 been added (no short equivalent) to specify a locale explicitly on the
3003 pcregrep command, overriding the environment variables.
3004
3005 (d) When -B was used with -n, some line numbers in the output were one less
3006 than they should have been.
3007
3008 (e) Added the -o (--only-matching) option.
3009
3010 (f) If -A or -C was used with -c (count only), some lines of context were
3011 accidentally printed for the final match.
3012
3013 (g) Added the -H (--with-filename) option.
3014
3015 (h) The combination of options -rh failed to suppress file names for files
3016 that were found from directory arguments.
3017
3018 (i) Added the -D (--devices) and -d (--directories) options.
3019
3020 (j) Added the -F (--fixed-strings) option.
3021
3022 (k) Allow "-" to be used as a file name for -f as well as for a data file.
3023
3024 (l) Added the --colo(u)r option.
3025
3026 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3027 is not present by default.
3028
3029 3. A nasty bug was discovered in the handling of recursive patterns, that is,
3030 items such as (?R) or (?1), when the recursion could match a number of
3031 alternatives. If it matched one of the alternatives, but subsequently,
3032 outside the recursion, there was a failure, the code tried to back up into
3033 the recursion. However, because of the way PCRE is implemented, this is not
3034 possible, and the result was an incorrect result from the match.
3035
3036 In order to prevent this happening, the specification of recursion has
3037 been changed so that all such subpatterns are automatically treated as
3038 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3039
3040 4. I had overlooked the fact that, in some locales, there are characters for
3041 which isalpha() is true but neither isupper() nor islower() are true. In
3042 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3043 and ordfeminine) are like this. This affected the treatment of \w and \W
3044 when they appeared in character classes, but not when they appeared outside
3045 a character class. The bit map for "word" characters is now created
3046 separately from the results of isalnum() instead of just taking it from the
3047 upper, lower, and digit maps. (Plus the underscore character, of course.)
3048
3049 5. The above bug also affected the handling of POSIX character classes such as
3050 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3051 permanent tables. Instead, the bit maps for such a class were previously
3052 created as the appropriate unions of the upper, lower, and digit bitmaps.
3053 Now they are created by subtraction from the [[:word:]] class, which has
3054 its own bitmap.
3055
3056 6. The [[:blank:]] character class matches horizontal, but not vertical space.
3057 It is created by subtracting the vertical space characters (\x09, \x0a,
3058 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3059 subtraction was done in the overall bitmap for a character class, meaning
3060 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3061 be recognized. This bug has been fixed.
3062
3063 7. Patches from the folks at Google:
3064
3065 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3066 real life, but is still worth protecting against".
3067
3068 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3069 regular expressions".
3070
3071 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3072 have it.
3073
3074 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3075 "configure" and the latter not, in order to fix a problem somebody had
3076 with compiling the Arg class on HP-UX.
3077
3078 (e) Improve the error-handling of the C++ wrapper a little bit.
3079
3080 (f) New tests for checking recursion limiting.
3081
3082 8. The pcre_memmove() function, which is used only if the environment does not
3083 have a standard memmove() function (and is therefore rarely compiled),
3084 contained two bugs: (a) use of int instead of size_t, and (b) it was not
3085 returning a result (though PCRE never actually uses the result).
3086
3087 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3088 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3089 returned instead of calling malloc() with an overflowing number that would
3090 most likely cause subsequent chaos.
3091
3092 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3093
3094 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3095 with this option is matched, the nmatch and pmatch options of regexec() are
3096 ignored.
3097
3098 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3099 provided in case anyone wants to the the POSIX interface with UTF-8
3100 strings.
3101
3102 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3103 C++ linking (needed for some HP-UX environments).
3104
3105 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3106 (unused parameter) and in the pcre_printint() function (omitted "default"
3107 switch label when the default is to do nothing).
3108
3109 15. Added some code to make it possible, when PCRE is compiled as a C++
3110 library, to replace subject pointers for pcre_exec() with a smart pointer
3111 class, thus making it possible to process discontinuous strings.
3112
3113 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3114 much the same function. They were added by different people who were trying
3115 to make PCRE easy to compile on non-Unix systems. It has been suggested
3116 that PCRE_EXPORT be abolished now that there is more automatic apparatus
3117 for compiling on Windows systems. I have therefore replaced it with
3118 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3119 defaults to "extern" for C or "extern C" for C++, which works fine on
3120 Unix-like systems. It is now possible to override the value of PCRE_DATA_
3121 SCOPE with something explicit in config.h. In addition:
3122
3123 (a) pcreposix.h still had just "extern" instead of either of these macros;
3124 I have replaced it with PCRE_DATA_SCOPE.
3125
3126 (b) Functions such as _pcre_xclass(), which are internal to the library,
3127 but external in the C sense, all had PCRE_EXPORT in their definitions.
3128 This is apparently wrong for the Windows case, so I have removed it.
3129 (It makes no difference on Unix-like systems.)
3130
3131 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3132 of recursive calls to match(). This is different to MATCH_LIMIT because
3133 that limits the total number of calls to match(), not all of which increase
3134 the depth of recursion. Limiting the recursion depth limits the amount of
3135 stack (or heap if NO_RECURSE is set) that is used. The default can be set
3136 when PCRE is compiled, and changed at run time. A patch from Google adds
3137 this functionality to the C++ interface.
3138
3139 18. Changes to the handling of Unicode character properties:
3140
3141 (a) Updated the table to Unicode 4.1.0.
3142
3143 (b) Recognize characters that are not in the table as "Cn" (undefined).
3144
3145 (c) I revised the way the table is implemented to a much improved format
3146 which includes recognition of ranges. It now supports the ranges that
3147 are defined in UnicodeData.txt, and it also amalgamates other
3148 characters into ranges. This has reduced the number of entries in the
3149 table from around 16,000 to around 3,000, thus reducing its size
3150 considerably. I realized I did not need to use a tree structure after
3151 all - a binary chop search is just as efficient. Having reduced the
3152 number of entries, I extended their size from 6 bytes to 8 bytes to
3153 allow for more data.
3154
3155 (d) Added support for Unicode script names via properties such as \p{Han}.
3156
3157 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3158 matching that character.
3159
3160 20. When matching a repeated Unicode property with a minimum greater than zero,
3161 (for example \pL{2,}), PCRE could look past the end of the subject if it
3162 reached it while seeking the minimum number of characters. This could
3163 happen only if some of the characters were more than one byte long, because
3164 there is a check for at least the minimum number of bytes.
3165
3166 21. Refactored the implementation of \p and \P so as to be more general, to
3167 allow for more different types of property in future. This has changed the
3168 compiled form incompatibly. Anybody with saved compiled patterns that use
3169 \p or \P will have to recompile them.
3170
3171 22. Added "Any" and "L&" to the supported property types.
3172
3173 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3174 but give a compile time error if the value is greater than 0xff.
3175
3176 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3177 accidentally not being installed or uninstalled.
3178
3179 25. The pcre.h file was built from pcre.h.in, but the only changes that were
3180 made were to insert the current release number. This seemed silly, because
3181 it made things harder for people building PCRE on systems that don't run
3182 "configure". I have turned pcre.h into a distributed file, no longer built
3183 by "configure", with the version identification directly included. There is
3184 no longer a pcre.h.in file.
3185
3186 However, this change necessitated a change to the pcre-config script as
3187 well. It is built from pcre-config.in, and one of the substitutions was the
3188 release number. I have updated configure.ac so that ./configure now finds
3189 the release number by grepping pcre.h.
3190
3191 26. Added the ability to run the tests under valgrind.
3192
3193
3194 Version 6.4 05-Sep-05
3195 ---------------------
3196
3197 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3198 "--" to be printed when multiple files were scanned, even when none of the
3199 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3200 consider it to be a bug, and have restored the previous behaviour.
3201
3202 2. A couple of code tidies to get rid of compiler warnings.
3203
3204 3. The pcretest program used to cheat by referring to symbols in the library
3205 whose names begin with _pcre_. These are internal symbols that are not
3206 really supposed to be visible externally, and in some environments it is
3207 possible to suppress them. The cheating is now confined to including
3208 certain files from the library's source, which is a bit cleaner.
3209
3210 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3211 file's purpose clearer.
3212
3213 5. Reorganized pcre_ucp_findchar().
3214
3215
3216 Version 6.3 15-Aug-05
3217 ---------------------
3218
3219 1. The file libpcre.pc.in did not have general read permission in the tarball.
3220
3221 2. There were some problems when building without C++ support:
3222
3223 (a) If C++ support was not built, "make install" and "make test" still
3224 tried to test it.
3225
3226 (b) There were problems when the value of CXX was explicitly set. Some
3227 changes have been made to try to fix these, and ...
3228
3229 (c) --disable-cpp can now be used to explicitly disable C++ support.
3230
3231 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3232 backslash in a target when C++ was disabled. This confuses some
3233 versions of "make", apparently. Using an intermediate variable solves
3234 this. (Same for CPP_LOBJ.)
3235
3236 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3237 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3238 necessary on certain architectures.
3239
3240 4. Added a setting of -export-symbols-regex to the link command to remove
3241 those symbols that are exported in the C sense, but actually are local
3242 within the library, and not documented. Their names all begin with
3243 "_pcre_". This is not a perfect job, because (a) we have to except some
3244 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3245 available (and never for static libraries). I have made a note to try to
3246 find a way round (a) in the future.
3247
3248
3249 Version 6.2 01-Aug-05
3250 ---------------------
3251
3252 1. There was no test for integer overflow of quantifier values. A construction
3253 such as {1111111111111111} would give undefined results. What is worse, if
3254 a minimum quantifier for a parenthesized subpattern overflowed and became
3255 negative, the calculation of the memory size went wrong. This could have
3256 led to memory overwriting.
3257
3258 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3259
3260 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3261 operating environments where this matters.
3262
3263 4. Applied Giuseppe Maxia's patch to add additional features for controlling
3264 PCRE options from within the C++ wrapper.
3265
3266 5. Named capturing subpatterns were not being correctly counted when a pattern
3267 was compiled. This caused two problems: (a) If there were more than 100
3268 such subpatterns, the calculation of the memory needed for the whole
3269 compiled pattern went wrong, leading to an overflow error. (b) Numerical
3270 back references of the form \12, where the number was greater than 9, were
3271 not recognized as back references, even though there were sufficient
3272 previous subpatterns.
3273
3274 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3275 versions of gcc, e.g. 2.95.4.
3276
3277
3278 Version 6.1 21-Jun-05
3279 ---------------------
3280
3281 1. There was one reference to the variable "posix" in pcretest.c that was not
3282 surrounded by "#if !defined NOPOSIX".
3283
3284 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3285 the cross-check on the old pcre_info() function, for the benefit of the
3286 cut-down version of PCRE that is currently imported into Exim.
3287
3288 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3289 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3290 patterns that start (?i) but I don't think that matters. The use of (?i) is
3291 just an example; this all applies to the other options as well.
3292
3293 4. Since libtool seems to echo the compile commands it is issuing, the output
3294 from "make" can be reduced a bit by putting "@" in front of each libtool
3295 compile command.
3296
3297 5. Patch from the folks at Google for configure.in to be a bit more thorough
3298 in checking for a suitable C++ installation before trying to compile the
3299 C++ stuff. This should fix a reported problem when a compiler was present,
3300 but no suitable headers.
3301
3302 6. The man pages all had just "PCRE" as their title. I have changed them to
3303 be the relevant file name. I have also arranged that these names are
3304 retained in the file doc/pcre.txt, which is a concatenation in text format
3305 of all the man pages except the little individual ones for each function.
3306
3307 7. The NON-UNIX-USE file had not been updated for the different set of source
3308 files that come with release 6. I also added a few comments about the C++
3309 wrapper.
3310
3311
3312 Version 6.0 07-Jun-05
3313 ---------------------
3314
3315 1. Some minor internal re-organization to help with my DFA experiments.
3316
3317 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3318 didn't matter for the library itself when fully configured, but did matter
3319 when compiling without UCP support, or within Exim, where the ucp files are
3320 not imported.
3321
3322 3. Refactoring of the library code to split up the various functions into
3323 different source modules. The addition of the new DFA matching code (see
3324 below) to a single monolithic source would have made it really too
3325 unwieldy, quite apart from causing all the code to be include in a
3326 statically linked application, when only some functions are used. This is
3327 relevant even without the DFA addition now that patterns can be compiled in
3328 one application and matched in another.
3329
3330 The downside of splitting up is that there have to be some external
3331 functions and data tables that are used internally in different modules of
3332 the library but which are not part of the API. These have all had their
3333 names changed to start with "_pcre_" so that they are unlikely to clash
3334 with other external names.
3335
3336 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3337 a different (DFA) algorithm. Although it is slower than the original
3338 function, it does have some advantages for certain types of matching
3339 problem.
3340
3341 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3342 including restarting after a partial match.
3343
3344 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3345 defined when compiling for Windows was sent to me. I have put it into the
3346 code, though I have no means of testing or verifying it.
3347
3348 7. Added the pcre_refcount() auxiliary function.
3349
3350 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3351 match before or at the first newline in the subject string. In pcretest,
3352 the /f option on a pattern can be used to set this.
3353
3354 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3355 would behave wrongly. This has been present in PCRE since release 4.0.
3356
3357 10. A number of changes to the pcregrep command:
3358
3359 (a) Refactored how -x works; insert ^(...)$ instead of setting
3360 PCRE_ANCHORED and checking the length, in preparation for adding
3361 something similar for -w.
3362
3363 (b) Added the -w (match as a word) option.
3364
3365 (c) Refactored the way lines are read and buffered so as to have more
3366 than one at a time available.
3367
3368 (d) Implemented a pcregrep test script.
3369
3370 (e) Added the -M (multiline match) option. This allows patterns to match
3371 over several lines of the subject. The buffering ensures that at least
3372 8K, or the rest of the document (whichever is the shorter) is available
3373 for matching (and similarly the previous 8K for lookbehind assertions).
3374
3375 (f) Changed the --help output so that it now says
3376
3377 -w, --word-regex(p)
3378
3379 instead of two lines, one with "regex" and the other with "regexp"
3380 because that confused at least one person since the short forms are the
3381 same. (This required a bit of code, as the output is generated
3382 automatically from a table. It wasn't just a text change.)
3383
3384 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3385 option but starts with a hyphen. Could be a pattern or a path name
3386 starting with a hyphen, for instance.
3387
3388 (h) "-" can be given as a file name to represent stdin.
3389
3390 (i) When file names are being printed, "(standard input)" is used for
3391 the standard input, for compatibility with GNU grep. Previously
3392 "<stdin>" was used.
3393
3394 (j) The option --label=xxx can be used to supply a name to be used for
3395 stdin when file names are being printed. There is no short form.
3396
3397 (k) Re-factored the options decoding logic because we are going to add
3398 two more options that take data. Such options can now be given in four
3399 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3400
3401 (l) Added the -A, -B, and -C options for requesting that lines of context
3402 around matches be printed.
3403
3404 (m) Added the -L option to print the names of files that do not contain
3405 any matching lines, that is, the complement of -l.
3406
3407 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3408 continue to scan other files.
3409
3410 (o) The -s option was incorrectly implemented. For compatibility with other
3411 greps, it now suppresses the error message for a non-existent or non-
3412 accessible file (but not the return code). There is a new option called
3413 -q that suppresses the output of matching lines, which was what -s was
3414 previously doing.
3415
3416 (p) Added --include and --exclude options to specify files for inclusion
3417 and exclusion when recursing.
3418
3419 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3420 Hopefully, it now does.
3421
3422 12. Missing cast in pcre_study().
3423
3424 13. Added an "uninstall" target to the makefile.
3425
3426 14. Replaced "extern" in the function prototypes in Makefile.in with
3427 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3428 world, but is set differently for Windows.
3429
3430 15. Added a second compiling function called pcre_compile2(). The only
3431 difference is that it has an extra argument, which is a pointer to an
3432 integer error code. When there is a compile-time failure, this is set
3433 non-zero, in addition to the error test pointer being set to point to an
3434 error message. The new argument may be NULL if no error number is required
3435 (but then you may as well call pcre_compile(), which is now just a
3436 wrapper). This facility is provided because some applications need a
3437 numeric error indication, but it has also enabled me to tidy up the way
3438 compile-time errors are handled in the POSIX wrapper.
3439
3440 16. Added VPATH=.libs to the makefile; this should help when building with one
3441 prefix path and installing with another. (Or so I'm told by someone who
3442 knows more about this stuff than I do.)
3443
3444 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3445 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3446 match everything, including newlines. This is not POSIX-compatible, but
3447 somebody wanted the feature. From pcretest it can be activated by using
3448 both the P and the s flags.
3449
3450 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3451
3452 19. libpcre.pc was being incorrectly installed as executable.
3453
3454 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3455 it now also looks for '\r' so that it will work unmodified on Windows.
3456
3457 21. Added Google's contributed C++ wrapper to the distribution.
3458
3459 22. Added some untidy missing memory free() calls in pcretest, to keep
3460 Electric Fence happy when testing.
3461
3462
3463
3464 Version 5.0 13-Sep-04
3465 ---------------------
3466
3467 1. Internal change: literal characters are no longer packed up into items
3468 containing multiple characters in a single byte-string. Each character
3469 is now matched using a separate opcode. However, there may be more than one
3470 byte in the character in UTF-8 mode.
3471
3472 2. The pcre_callout_block structure has two new fields: pattern_position and
3473 next_item_length. These contain the offset in the pattern to the next match
3474 item, and its length, respectively.
3475
3476 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3477 insertion of callouts before each pattern item. Added the /C option to
3478 pcretest to make use of this.
3479
3480 4. On the advice of a Windows user, the lines
3481
3482 #if defined(_WIN32) || defined(WIN32)
3483 _setmode( _fileno( stdout ), 0x8000 );
3484 #endif /* defined(_WIN32) || defined(WIN32) */
3485
3486 have been added to the source of pcretest. This apparently does useful
3487 magic in relation to line terminators.
3488
3489 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3490 for the benefit of those environments where the "b" makes a difference.
3491
3492 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3493 to know about it. I have put a hack into configure.in that adds in code
3494 to set GCC=yes if CC=icc. This seems to end up at a point in the
3495 generated configure script that is early enough to affect the setting of
3496 compiler options, which is what is needed, but I have no means of testing
3497 whether it really works. (The user who reported this had patched the
3498 generated configure script, which of course I cannot do.)
3499
3500 LATER: After change 22 below (new libtool files), the configure script
3501 seems to know about icc (and also ecc). Therefore, I have commented out
3502 this hack in configure.in.
3503
3504 7. Added support for pkg-config (2 patches were sent in).
3505
3506 8. Negated POSIX character classes that used a combination of internal tables
3507 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3508 [[:^ascii]]. Typically, they would match almost any characters. The other
3509 POSIX classes were not broken in this way.
3510
3511 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3512 to find the match, as PCRE was deluded into thinking that the match had to
3513 start at the start point or following a newline. The same bug applied to
3514 patterns with negative forward assertions or any backward assertions
3515 preceding ".*" at the start, unless the pattern required a fixed first
3516 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3517
3518 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3519 starting at the last subject character, bytes beyond the end of the subject
3520 string were read.
3521
3522 11. Renamed the variable "class" as "classbits" to make life easier for C++
3523 users. (Previously there was a macro definition, but it apparently wasn't
3524 enough.)
3525
3526 12. Added the new field "tables" to the extra data so that tables can be passed
3527 in at exec time, or the internal tables can be re-selected. This allows
3528 a compiled regex to be saved and re-used at a later time by a different
3529 program that might have everything at different addresses.
3530
3531 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3532 -R library as well as a -L library.
3533
3534 14. The debugging options of pcretest (-d on the command line or D on a
3535 pattern) showed incorrect output for anything following an extended class
3536 that contained multibyte characters and which was followed by a quantifier.
3537
3538 15. Added optional support for general category Unicode character properties
3539 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3540 support. It adds about 90K to the size of the library. The meanings of the
3541 inbuilt class escapes such as \d and \s have NOT been changed.
3542
3543 16. Updated pcredemo.c to include calls to free() to release the memory for the
3544 compiled pattern.
3545
3546 17. The generated file chartables.c was being created in the source directory
3547 instead of in the building directory. This caused the build to fail if the
3548 source directory was different from the building directory, and was
3549 read-only.
3550
3551 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3552 file. No doubt somebody will tell me if they don't make sense... Also added
3553 Dan Mooney's comments about building on OpenVMS.
3554
3555 19. Added support for partial matching via the PCRE_PARTIAL option for
3556 pcre_exec() and the \P data escape in pcretest.
3557
3558 20. Extended pcretest with 3 new pattern features:
3559
3560 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3561 write the compiled pattern to the file whose name is "rest-of-line".
3562 This is a straight binary dump of the data, with the saved pointer to
3563 the character tables forced to be NULL. The study data, if any, is
3564 written too. After writing, pcretest reads a new pattern.
3565
3566 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3567 compiled pattern from the given file. There must not be any
3568 occurrences of "<" in the file name (pretty unlikely); if there are,
3569 pcretest will instead treat the initial "<" as a pattern delimiter.
3570 After reading in the pattern, pcretest goes on to read data lines as
3571 usual.
3572
3573 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3574 and 16-bit fields in a compiled pattern, to simulate a pattern that
3575 was compiled on a host of opposite endianness.
3576
3577 21. The pcre-exec() function can now cope with patterns that were compiled on
3578 hosts of opposite endianness, with this restriction:
3579
3580 As for any compiled expression that is saved and used later, the tables
3581 pointer field cannot be preserved; the extra_data field in the arguments
3582 to pcre_exec() should be used to pass in a tables address if a value
3583 other than the default internal tables were used at compile time.
3584
3585 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3586 now diagnosed as an error. Previously, most of the time, a negative number
3587 would have been treated as zero, but if in addition "ovector" was passed as
3588 NULL, a crash could occur.
3589
3590 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3591 new versions from the libtool 1.5 distribution (the last one is a copy of
3592 a file called libtool.m4). This seems to have fixed the need to patch
3593 "configure" to support Darwin 1.3 (which I used to do). However, I still
3594 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3595 workstation).
3596
3597 24. Changed the PCRE licence to be the more standard "BSD" licence.
3598
3599
3600 Version 4.5 01-Dec-03
3601 ---------------------
3602
3603 1. There has been some re-arrangement of the code for the match() function so
3604 that it can be compiled in a version that does not call itself recursively.
3605 Instead, it keeps those local variables that need separate instances for
3606 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3607 needs to "recurse". Keeping track of where control must go is done by means
3608 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3609 hide most of the details from the main code, and operates only if
3610 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3611 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3612 operating.
3613
3614 To make it easier for callers to provide specially tailored get/free
3615 functions for this usage, two new functions, pcre_stack_malloc, and
3616 pcre_stack_free, are used. They are always called in strict stacking order,
3617 and the size of block requested is always the same.
3618
3619 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3620 PCRE has been compiled to use the stack or the heap for recursion. The
3621 -C option of pcretest uses this to show which version is compiled.
3622
3623 A new data escape \S, is added to pcretest; it causes the amounts of store
3624 obtained and freed by both kinds of malloc/free at match time to be added
3625 to the output.
3626
3627 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3628 what's available on my current Linux desktop machine.
3629
3630 3. When matching a UTF-8 string, the test for a valid string at the start has
3631 been extended. If start_offset is not zero, PCRE now checks that it points
3632 to a byte that is the start of a UTF-8 character. If not, it returns
3633 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3634 this is necessary because there may be backward assertions in the pattern.
3635 When matching the same subject several times, it may save resources to use
3636 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3637
3638 4. The code for checking the validity of UTF-8 strings has been tightened so
3639 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3640 containing "overlong sequences".
3641
3642 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3643 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3644 should have been "&", but it just so happened that all the cases this let
3645 through by mistake were picked up later in the function.
3646
3647 6. I had used a variable called "isblank" - this is a C99 function, causing
3648 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3649
3650 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3651 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3652 so the version is shown; (c) stop "make test" echoing "./RunTest".
3653
3654 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3655
3656 9. The prototype for memmove() for systems that don't have it was using
3657 size_t, but the inclusion of the header that defines size_t was later. I've
3658 moved the #includes for the C headers earlier to avoid this.
3659
3660 10. Added some adjustments to the code to make it easier to compiler on certain
3661 special systems:
3662
3663 (a) Some "const" qualifiers were missing.
3664 (b) Added the macro EXPORT before all exported functions; by default this
3665 is defined to be empty.
3666 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3667 that it reads its output file name as an argument instead of writing
3668 to the standard output and assuming this can be redirected.
3669
3670 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3671 class containing characters with values greater than 255, PCRE compilation
3672 went into a loop.
3673
3674 12. A recursive reference to a subpattern that was within another subpattern
3675 that had a minimum quantifier of zero caused PCRE to crash. For example,
3676 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3677 recursion. If the recursively-called subpattern itself had a zero repeat,
3678 that was OK.
3679
3680 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3681 buffer into which it was copied (for escape processing) was still set at
3682 1024, so long lines caused crashes.
3683
3684 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3685 "internal error: code overflow...". This applied to any character class
3686 that was followed by a possessive quantifier.
3687
3688 15. Modified the Makefile to add libpcre.la as a prerequisite for
3689 libpcreposix.la because I was told this is needed for a parallel build to
3690 work.
3691
3692 16. If a pattern that contained .* following optional items at the start was
3693 studied, the wrong optimizing data was generated, leading to matching
3694 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3695 matching string must start with a or b or c. The correct conclusion for
3696 this pattern is that a match can start with any character.
3697
3698
3699 Version 4.4 13-Aug-03
3700 ---------------------
3701
3702 1. In UTF-8 mode, a character class containing characters with values between
3703 127 and 255 was not handled correctly if the compiled pattern was studied.
3704 In fixing this, I have also improved the studying algorithm for such
3705 classes (slightly).
3706
3707 2. Three internal functions had redundant arguments passed to them. Removal
3708 might give a very teeny performance improvement.
3709
3710 3. Documentation bug: the value of the capture_top field in a callout is *one
3711 more than* the number of the hightest numbered captured substring.
3712
3713 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3714 in incorrectly linking with a previously installed version. They now link
3715 explicitly with libpcre.la.
3716
3717 5. configure.in no longer needs to recognize Cygwin specially.
3718
3719 6. A problem in pcre.in for Windows platforms is fixed.
3720
3721 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3722 pcretest, it used to include the size of the study block as part of its
3723 output. Unfortunately, the structure contains a field that has a different
3724 size on different hardware architectures. This meant that the tests that
3725 showed this size failed. As the block is currently always of a fixed size,
3726 this information isn't actually particularly useful in pcretest output, so
3727 I have just removed it.
3728
3729 8. Three pre-processor statements accidentally did not start in column 1.
3730 Sadly, there are *still* compilers around that complain, even though
3731 standard C has not required this for well over a decade. Sigh.
3732
3733 9. In pcretest, the code for checking callouts passed small integers in the
3734 callout_data field, which is a void * field. However, some picky compilers
3735 complained about the casts involved for this on 64-bit systems. Now
3736 pcretest passes the address of the small integer instead, which should get
3737 rid of the warnings.
3738
3739 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3740 both compile and run time, and gives an error if an invalid UTF-8 sequence
3741 is found. There is a option for disabling this check in cases where the
3742 string is known to be correct and/or the maximum performance is wanted.
3743
3744 11. In response to a bug report, I changed one line in Makefile.in from
3745
3746 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3747 to
3748 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3749
3750 to look similar to other lines, but I have no way of telling whether this
3751 is the right thing to do, as I do not use Windows. No doubt I'll get told
3752 if it's wrong...
3753
3754
3755 Version 4.3 21-May-03
3756 ---------------------
3757
3758 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3759 Makefile.
3760
3761 2. Some refactoring to improve the quality of the code:
3762
3763 (i) The utf8_table... variables are now declared "const".
3764
3765 (ii) The code for \cx, which used the "case flipping" table to upper case
3766 lower case letters, now just substracts 32. This is ASCII-specific,
3767 but the whole concept of \cx is ASCII-specific, so it seems
3768 reasonable.
3769
3770 (iii) PCRE was using its character types table to recognize decimal and
3771 hexadecimal digits in the pattern. This is silly, because it handles
3772 only 0-9, a-f, and A-F, but the character types table is locale-
3773 specific, which means strange things might happen. A private
3774 table is now used for this - though it costs 256 bytes, a table is
3775 much faster than multiple explicit tests. Of course, the standard
3776 character types table is still used for matching digits in subject
3777 strings against \d.
3778
3779 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3780 ending in _t are). So I've renamed it as ESC_tee.
3781
3782 3. The first argument for regexec() in the POSIX wrapper should have been
3783 defined as "const".
3784
3785 4. Changed pcretest to use malloc() for its buffers so that they can be
3786 Electric Fenced for debugging.
3787
3788 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3789 to read one or more bytes before the start of the subject string. Often this
3790 had no effect on PCRE's behaviour, but in some circumstances it could
3791 provoke a segmentation fault.
3792
3793 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3794 to try to read one or more bytes before the start of the subject string.
3795
3796 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3797 UTF-8 support could misbehave in various ways if the subject string
3798 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3799 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3800 back over UTF-8 characters.)
3801
3802
3803 Version 4.2 14-Apr-03
3804 ---------------------
3805
3806 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3807
3808 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3809 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3810 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3811 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3812 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3813 and BUILD_EXEEXT
3814 Note: automatic setting of the BUILD variables is not yet working
3815 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3816 compile-time but not at link-time
3817 [LINK]: use for linking executables only
3818 make different versions for Windows and non-Windows
3819 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3820 libraries
3821 [LINK_FOR_BUILD]: new variable
3822 [OBJEXT]: use throughout
3823 [EXEEXT]: use throughout
3824 <winshared>: new target
3825 <wininstall>: new target
3826 <dftables.o>: use native compiler
3827 <dftables>: use native linker
3828 <install>: handle Windows platform correctly
3829 <clean>: ditto
3830 <check>: ditto
3831 copy DLL to top builddir before testing
3832
3833 As part of these changes, -no-undefined was removed again. This was reported
3834 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3835 in any case.
3836
3837 3. Some tidies to get rid of compiler warnings:
3838
3839 . In the match_data structure, match_limit was an unsigned long int, whereas
3840 match_call_count was an int. I've made them both unsigned long ints.
3841
3842 . In pcretest the fact that a const uschar * doesn't automatically cast to
3843 a void * provoked a warning.
3844
3845 . Turning on some more compiler warnings threw up some "shadow" variables
3846 and a few more missing casts.
3847
3848 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3849 option, a class that contained a single character with a value between 128
3850 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3851
3852 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3853 option, a class that contained several characters, but with at least one
3854 whose value was between 128 and 255 caused PCRE to crash.
3855
3856
3857 Version 4.1 12-Mar-03
3858 ---------------------
3859
3860 1. Compiling with gcc -pedantic found a couple of places where casts were
3861 needed, and a string in dftables.c that was longer than standard compilers are
3862 required to support.
3863
3864 2. Compiling with Sun's compiler found a few more places where the code could
3865 be tidied up in order to avoid warnings.
3866
3867 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3868 first of these names is deprecated in the latest Autoconf in favour of the name
3869 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3870 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3871 analogy I have changed it to CFLAGS_FOR_BUILD.
3872
3873 4. Added -no-undefined to the linking command in the Makefile, because this is
3874 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3875 linking step for the pcreposix library.
3876
3877 5. PCRE was failing to diagnose the case of two named groups with the same
3878 name.
3879
3880 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3881 literal character that is needed in the subject for a match, and scans along to
3882 ensure that it is present before embarking on the full matching process. This
3883 saves time in cases of nested unlimited repeats that are never going to match.
3884 Problem: the scan can take a lot of time if the subject is very long (e.g.
3885 megabytes), thus penalizing straightforward matches. It is now done only if the
3886 amount of subject to be scanned is less than 1000 bytes.
3887
3888 7. A lesser problem with the same optimization is that it was recording the
3889 first character of an anchored pattern as "needed", thus provoking a search
3890 right along the subject, even when the first match of the pattern was going to
3891 fail. The "needed" character is now not set for anchored patterns, unless it
3892 follows something in the pattern that is of non-fixed length. Thus, it still
3893 fulfils its original purpose of finding quick non-matches in cases of nested
3894 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3895
3896
3897 Version 4.0 17-Feb-03
3898 ---------------------
3899
3900 1. If a comment in an extended regex that started immediately after a meta-item
3901 extended to the end of string, PCRE compiled incorrect data. This could lead to
3902 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3903
3904 2. Moved to autoconf 2.53 and libtool 1.4.2.
3905
3906 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3907 the special perltest8 script is no longer needed - all the tests can be run
3908 from a single perltest script.
3909
3910 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3911 by \s. It has now been removed in PCRE. This means it isn't recognized as
3912 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3913 class [:space:] *does* include VT, thereby creating a mess.
3914
3915 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3916 space and tab.
3917
3918 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3919 its new features into the main test script, reducing the number of scripts.
3920
3921 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3922 were backward compatible, and made the (?i) apply to the whole pattern, as if
3923 /i were given. Now it behaves more logically, and applies the option setting
3924 only to what follows. PCRE has been changed to follow suit. However, if it
3925 finds options settings right at the start of the pattern, it extracts them into
3926 the global options, as before. Thus, they show up in the info data.
3927
3928 8. Added support for the \Q...\E escape sequence. Characters in between are
3929 treated as literals. This is slightly different from Perl in that $ and @ are
3930 also handled as literals inside the quotes. In Perl, they will cause variable
3931 interpolation. Note the following examples:
3932
3933 Pattern PCRE matches Perl matches
3934
3935 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
3936 \Qabc\$xyz\E abc\$xyz abc\$xyz
3937 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3938
3939 For compatibility with Perl, \Q...\E sequences are recognized inside character
3940 classes as well as outside them.
3941
3942 9. Re-organized 3 code statements in pcretest to avoid "overflow in
3943 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3944 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3945 signed/unsigned warnings.
3946
3947 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3948 option for pcretest, so I've replaced it by a simple function that does just
3949 that job.
3950
3951 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3952 "pcregrep -".
3953
3954 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3955 Java package. This provides some syntactic sugar for simple cases of what my
3956 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3957 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3958 item, you can use this simplified notation. Note that only makes sense with
3959 greedy quantifiers. Consequently, the use of the possessive quantifier forces
3960 greediness, whatever the setting of the PCRE_UNGREEDY option.
3961
3962 13. A change of greediness default within a pattern was not taking effect at
3963 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3964 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3965 was abstracted outside.
3966
3967 14. PCRE now supports the \G assertion. It is true when the current matching
3968 position is at the start point of the match. This differs from \A when the
3969 starting offset is non-zero. Used with the /g option of pcretest (or similar
3970 code), it works in the same way as it does for Perl's /g option. If all
3971 alternatives of a regex begin with \G, the expression is anchored to the start
3972 match position, and the "anchored" flag is set in the compiled expression.
3973
3974 15. Some bugs concerning the handling of certain option changes within patterns
3975 have been fixed. These applied to options other than (?ims). For example,
3976 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3977 been the other way round. Some of this was related to change 7 above.
3978
3979 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3980 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3981 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3982 POSIX classes only within a class (e.g. /[[:alpha:]]/).
3983
3984 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3985 mode. Unlike ".", it always matches newline, whatever the setting of
3986 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3987 assertions. Perl allows it, but it doesn't (in general) work because it can't
3988 calculate the length of the lookbehind. At least, that's the case for Perl
3989 5.8.0 - I've been told they are going to document that it doesn't work in
3990 future.
3991
3992 18. Added an error diagnosis for escapes that PCRE does not support: these are
3993 \L, \l, \N, \P, \p, \U, \u, and \X.
3994
3995 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3996 reading past the end of the pattern in cases such as /[abcd/.
3997
3998 20. PCRE was getting more memory than necessary for patterns with classes that
3999 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
4000
4001 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
4002 compiling PCRE for use with Virtual Pascal.
4003
4004 22. Small fix to the Makefile to make it work properly if the build is done
4005 outside the source tree.
4006
4007 23. Added a new extension: a condition to go with recursion. If a conditional
4008 subpattern starts with (?(R) the "true" branch is used if recursion has
4009 happened, whereas the "false" branch is used only at the top level.
4010
4011 24. When there was a very long string of literal characters (over 255 bytes
4012 without UTF support, over 250 bytes with UTF support), the computation of how
4013 much memory was required could be incorrect, leading to segfaults or other
4014 strange effects.
4015
4016 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
4017 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
4018 there was a subsequent back reference to those brackets. This meant that, for
4019 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
4020 possible to check for precisely this case. All we can do is abandon the
4021 optimization if .* occurs inside capturing brackets when there are any back
4022 references whatsoever. (See below for a better fix that came later.)
4023
4024 26. The handling of the optimization for finding the first character of a
4025 non-anchored pattern, and for finding a character that is required later in the
4026 match were failing in some cases. This didn't break the matching; it just
4027 failed to optimize when it could. The way this is done has been re-implemented.
4028
4029 27. Fixed typo in error message for invalid (?R item (it said "(?p").
4030
4031 28. Added a new feature that provides some of the functionality that Perl
4032 provides with (?{...}). The facility is termed a "callout". The way it is done
4033 in PCRE is for the caller to provide an optional function, by setting
4034 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
4035 global variable. By default it is unset, which disables all calling out. To get
4036 the function called, the regex must include (?C) at appropriate points. This
4037 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
4038 This provides a means of identifying different callout points. When PCRE
4039 reaches such a point in the regex, if pcre_callout has been set, the external
4040 function is called. It is provided with data in a structure called
4041 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
4042 matching continues; if it returns a non-zero value, the match at the current
4043 point fails. However, backtracking will occur if possible. [This was changed
4044 later and other features added - see item 49 below.]
4045
4046 29. pcretest is upgraded to test the callout functionality. It provides a
4047 callout function that displays information. By default, it shows the start of
4048 the match and the current position in the text. There are some new data escapes
4049 to vary what happens:
4050
4051 \C+ in addition, show current contents of captured substrings
4052 \C- do not supply a callout function
4053 \C!n return 1 when callout number n is reached
4054 \C!n!m return 1 when callout number n is reached for the mth time
4055
4056 30. If pcregrep was called with the -l option and just a single file name, it
4057 output "<stdin>" if a match was found, instead of the file name.
4058
4059 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
4060 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
4061 pcre_exec(). This saves a malloc/free per call. The default value of
4062 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
4063 when configuring.
4064
4065 32. The default maximum size of a compiled pattern is 64K. There have been a
4066 few cases of people hitting this limit. The code now uses macros to handle the
4067 storing of links as offsets within the compiled pattern. It defaults to 2-byte
4068 links, but this can be changed to 3 or 4 bytes by --with-link-size when
4069 configuring. Tests 2 and 5 work only with 2-byte links because they output
4070 debugging information about compiled patterns.
4071
4072 33. Internal code re-arrangements:
4073
4074 (a) Moved the debugging function for printing out a compiled regex into
4075 its own source file (printint.c) and used #include to pull it into
4076 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
4077 separate copies.
4078
4079 (b) Defined the list of op-code names for debugging as a macro in
4080 internal.h so that it is next to the definition of the opcodes.
4081
4082 (c) Defined a table of op-code lengths for simpler skipping along compiled
4083 code. This is again a macro in internal.h so that it is next to the
4084 definition of the opcodes.
4085
4086 34. Added support for recursive calls to individual subpatterns, along the
4087 lines of Robin Houston's patch (but implemented somewhat differently).
4088
4089 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
4090 allow it to read and process whole directories in Win32. This code was
4091 contributed by Lionel Fourquaux; it has not been tested by me.
4092
4093 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
4094 used to name a group. Names consist of alphanumerics and underscores, and must
4095 be unique. Back references use the syntax (?P=name) and recursive calls use
4096 (?P>name) which is a PCRE extension to the Python extension. Groups still have
4097 numbers. The function pcre_fullinfo() can be used after compilation to extract
4098 a name/number map. There are three relevant calls:
4099
4100 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
4101 PCRE_INFO_NAMECOUNT yields the number of entries
4102 PCRE_INFO_NAMETABLE yields a pointer to the map.
4103
4104 The map is a vector of fixed-size entries. The size of each entry depends on
4105 the length of the longest name used. The first two bytes of each entry are the
4106 group number, most significant byte first. There follows the corresponding
4107 name, zero terminated. The names are in alphabetical order.
4108
4109 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
4110 case instead of 255. Making it the same both with and without UTF-8 support
4111 means that the same test output works with both.
4112
4113 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
4114 calling malloc() with a zero argument.
4115
4116 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
4117 optimization. I've improved things by keeping a bitmap of backreferences with
4118 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
4119 fact referenced, the optimization can be applied. It is unlikely that a
4120 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
4121 the match to follow \n) will appear inside brackets with a number greater than
4122 31, but if it does, any back reference > 31 suppresses the optimization.
4123
4124 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
4125 of disabling numbered capturing parentheses. Any opening parenthesis that is
4126 not followed by ? behaves as if it were followed by ?: but named parentheses
4127 can still be used for capturing (and they will acquire numbers in the usual
4128 way).
4129
4130 41. Redesigned the return codes from the match() function into yes/no/error so
4131 that errors can be passed back from deep inside the nested calls. A malloc
4132 failure while inside a recursive subpattern call now causes the
4133 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
4134
4135 42. It is now possible to set a limit on the number of times the match()
4136 function is called in a call to pcre_exec(). This facility makes it possible to
4137 limit the amount of recursion and backtracking, though not in a directly
4138 obvious way, because the match() function is used in a number of different
4139 circumstances. The count starts from zero for each position in the subject
4140 string (for non-anchored patterns). The default limit is, for compatibility, a
4141 large number, namely 10 000 000. You can change this in two ways:
4142
4143 (a) When configuring PCRE before making, you can use --with-match-limit=n
4144 to set a default value for the compiled library.
4145
4146 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
4147 a different value is set. See 45 below.
4148
4149 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
4150
4151 43. Added a new function pcre_config(int, void *) to enable run-time extraction
4152 of things that can be changed at compile time. The first argument specifies
4153 what is wanted and the second points to where the information is to be placed.
4154 The current list of available information is:
4155
4156 PCRE_CONFIG_UTF8
4157
4158 The output is an integer that is set to one if UTF-8 support is available;
4159 otherwise it is set to zero.
4160
4161 PCRE_CONFIG_NEWLINE
4162
4163 The output is an integer that it set to the value of the code that is used for
4164 newline. It is either LF (10) or CR (13).
4165
4166 PCRE_CONFIG_LINK_SIZE
4167
4168 The output is an integer that contains the number of bytes used for internal
4169 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
4170
4171 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
4172
4173 The output is an integer that contains the threshold above which the POSIX
4174 interface uses malloc() for output vectors. See item 31 above.
4175
4176 PCRE_CONFIG_MATCH_LIMIT
4177
4178 The output is an unsigned integer that contains the default limit of the number
4179 of match() calls in a pcre_exec() execution. See 42 above.
4180
4181 44. pcretest has been upgraded by the addition of the -C option. This causes it
4182 to extract all the available output from the new pcre_config() function, and to
4183 output it. The program then exits immediately.
4184
4185 45. A need has arisen to pass over additional data with calls to pcre_exec() in
4186 order to support additional features. One way would have been to define
4187 pcre_exec2() (for example) with extra arguments, but this would not have been
4188 extensible, and would also have required all calls to the original function to
4189 be mapped to the new one. Instead, I have chosen to extend the mechanism that
4190 is used for passing in "extra" data from pcre_study().
4191
4192 The pcre_extra structure is now exposed and defined in pcre.h. It currently
4193 contains the following fields:
4194
4195 flags a bitmap indicating which of the following fields are set
4196 study_data opaque data from pcre_study()
4197 match_limit a way of specifying a limit on match() calls for a specific
4198 call to pcre_exec()
4199 callout_data data for callouts (see 49 below)
4200
4201 The flag bits are also defined in pcre.h, and are
4202
4203 PCRE_EXTRA_STUDY_DATA
4204 PCRE_EXTRA_MATCH_LIMIT
4205 PCRE_EXTRA_CALLOUT_DATA
4206
4207 The pcre_study() function now returns one of these new pcre_extra blocks, with
4208 the actual study data pointed to by the study_data field, and the
4209 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
4210 before. That is, this change is entirely upwards-compatible and requires no
4211 change to existing code.
4212
4213 If you want to pass in additional data to pcre_exec(), you can either place it
4214 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
4215 block.
4216
4217 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
4218 data string contains the escape sequence \M, pcretest calls pcre_exec() several
4219 times with different match limits, until it finds the minimum value needed for
4220 pcre_exec() to complete. The value is then output. This can be instructive; for
4221 most simple matches the number is quite small, but for pathological cases it
4222 gets very large very quickly.
4223
4224 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
4225 returns the size of the data block pointed to by the study_data field in a
4226 pcre_extra block, that is, the value that was passed as the argument to
4227 pcre_malloc() when PCRE was getting memory in which to place the information
4228 created by pcre_study(). The fourth argument should point to a size_t variable.
4229 pcretest has been extended so that this information is shown after a successful
4230 pcre_study() call when information about the compiled regex is being displayed.
4231
4232 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
4233 because what follows is always an absolute path. (Later: it turns out that this
4234 is more than cosmetic for MinGW, because it doesn't like empty path
4235 components.)
4236
4237 49. Some changes have been made to the callout feature (see 28 above):
4238
4239 (i) A callout function now has three choices for what it returns:
4240
4241 0 => success, carry on matching
4242 > 0 => failure at this point, but backtrack if possible
4243 < 0 => serious error, return this value from pcre_exec()
4244
4245 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
4246 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
4247 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
4248 use by callout functions. It will never be used by PCRE itself.
4249
4250 (ii) The pcre_extra structure (see 45 above) has a void * field called
4251 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
4252 pcre_callout_block structure has a field of the same name. The contents of
4253 the field passed in the pcre_extra structure are passed to the callout
4254 function in the corresponding field in the callout block. This makes it
4255 easier to use the same callout-containing regex from multiple threads. For
4256 testing, the pcretest program has a new data escape
4257
4258 \C*n pass the number n (may be negative) as callout_data
4259
4260 If the callout function in pcretest receives a non-zero value as
4261 callout_data, it returns that value.
4262
4263 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
4264 there were some redundant $(CFLAGS) in commands that are now specified as
4265 $(LINK), which already includes $(CFLAGS).
4266
4267 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
4268 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
4269 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
4270 one-byte characters throughout. Note that case-insensitive matching applies
4271 only to characters whose values are less than 256. PCRE doesn't support the
4272 notion of cases for higher-valued characters.
4273
4274 (i) A character class whose characters are all within 0-255 is handled as
4275 a bit map, and the map is inverted for negative classes. Previously, a
4276 character > 255 always failed to match such a class; however it should
4277 match if the class was a negative one (e.g. [^ab]). This has been fixed.
4278
4279 (ii) A negated character class with a single character < 255 is coded as
4280 "not this character" (OP_NOT). This wasn't working properly when the test
4281 character was multibyte, either singly or repeated.
4282
4283 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
4284 mode, for example: \x{100}{2,3}.
4285
4286 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
4287 singly or repeated) now correctly test multibyte characters. However,
4288 PCRE doesn't recognize any characters with values greater than 255 as
4289 digits, spaces, or word characters. Such characters always match \D, \S,
4290 and \W, and never match \d, \s, or \w.
4291
4292 (v) Classes may now contain characters and character ranges with values
4293 greater than 255. For example: [ab\x{100}-\x{400}].
4294
4295 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
4296 PCRE in UTF-8 mode.
4297
4298 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
4299 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
4300 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
4301 value.)
4302
4303 53. The single man page has become too large. I have therefore split it up into
4304 a number of separate man pages. These also give rise to individual HTML pages;
4305 these are now put in a separate directory, and there is an index.html page that
4306 lists them all. Some hyperlinking between the pages has been installed.
4307
4308 54. Added convenience functions for handling named capturing parentheses.
4309
4310 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4311 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4312 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4313 are faulted.
4314
4315 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4316 calling configure. These values are used when compiling the dftables.c program
4317 which is run to generate the source of the default character tables. They
4318 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4319 you will need to set these values.
4320
4321 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4322
4323
4324 Version 3.9 02-Jan-02
4325 ---------------------
4326
4327 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4328
4329 2. If --disable-static was given, the building process failed when trying to
4330 build pcretest and pcregrep. (For some reason it was using libtool to compile
4331 them, which is not right, as they aren't part of the library.)
4332
4333
4334 Version 3.8 18-Dec-01
4335 ---------------------
4336
4337 1. The experimental UTF-8 code was completely screwed up. It was packing the
4338 bytes in the wrong order. How dumb can you get?
4339
4340
4341 Version 3.7 29-Oct-01
4342 ---------------------
4343
4344 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4345 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4346 this didn't happen under Solaris 8, where I normally test things.
4347
4348 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4349 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4350 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4351 this fix an hour or so after the initial 3.7 release.)
4352
4353
4354 Version 3.6 23-Oct-01
4355 ---------------------
4356
4357 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4358 offsets passed as NULL with zero offset count.
4359
4360 2. The config.guess and config.sub files had not been updated when I moved to
4361 the latest autoconf.
4362
4363
4364 Version 3.5 15-Aug-01
4365 ---------------------
4366
4367 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4368 had been forgotten.
4369
4370 2. By using declared but undefined structures, we can avoid using "void"
4371 definitions in pcre.h while keeping the internal definitions of the structures
4372 private.
4373
4374 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4375 user point of view, this means that both static and shared libraries are built
4376 by default, but this can be individually controlled. More of the work of
4377 handling this static/shared cases is now inside libtool instead of PCRE's make
4378 file.
4379
4380 4. The pcretest utility is now installed along with pcregrep because it is
4381 useful for users (to test regexs) and by doing this, it automatically gets
4382 relinked by libtool. The documentation has been turned into a man page, so
4383 there are now .1, .txt, and .html versions in /doc.
4384
4385 5. Upgrades to pcregrep:
4386 (i) Added long-form option names like gnu grep.
4387 (ii) Added --help to list all options with an explanatory phrase.
4388 (iii) Added -r, --recursive to recurse into sub-directories.
4389 (iv) Added -f, --file to read patterns from a file.
4390
4391 6. pcre_exec() was referring to its "code" argument before testing that
4392 argument for NULL (and giving an error if it was NULL).
4393
4394 7. Upgraded Makefile.in to allow for compiling in a different directory from
4395 the source directory.
4396
4397 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4398 options bits, the pointer it was passed was to an int instead of to an unsigned
4399 long int. This mattered only on 64-bit systems.
4400
4401 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4402 generated) instead of pcre.in, which it its source. Also made the same change
4403 in several of the .c files.
4404
4405 10. A new release of gcc defines printf() as a macro, which broke pcretest
4406 because it had an ifdef in the middle of a string argument for printf(). Fixed
4407 by using separate calls to printf().
4408
4409 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4410 script, to force use of CR or LF instead of \n in the source. On non-Unix
4411 systems, the value can be set in config.h.
4412
4413 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4414 absolute limit. Changed the text of the error message to make this clear, and
4415 likewise updated the man page.
4416
4417 13. The limit of 99 on the number of capturing subpatterns has been removed.
4418 The new limit is 65535, which I hope will not be a "real" limit.
4419
4420
4421 Version 3.4 22-Aug-00
4422 ---------------------
4423
4424 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4425
4426 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4427
4428
4429 Version 3.3 01-Aug-00
4430 ---------------------
4431
4432 1. If an octal character was given, but the value was greater than \377, it
4433 was not getting masked to the least significant bits, as documented. This could
4434 lead to crashes in some systems.
4435
4436 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4437 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4438
4439 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4440 These just pass their arguments on to (pcre_free)(), but they are provided
4441 because some uses of PCRE bind it to non-C systems that can call its functions,
4442 but cannot call free() or pcre_free() directly.
4443
4444 4. Add "make test" as a synonym for "make check". Corrected some comments in
4445 the Makefile.
4446
4447 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4448 Makefile.
4449
4450 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4451 command called pgrep for grepping around the active processes.
4452
4453 7. Added the beginnings of support for UTF-8 character strings.
4454
4455 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4456 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4457 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4458 out for the ar command.)
4459
4460
4461 Version 3.2 12-May-00
4462 ---------------------
4463
4464 This is purely a bug fixing release.
4465
4466 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4467 of ZA. This was just one example of several cases that could provoke this bug,
4468 which was introduced by change 9 of version 2.00. The code for breaking
4469 infinite loops after an iteration that matches an empty string was't working
4470 correctly.
4471
4472 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4473 when matched against abbab (for example). After matching an empty string, it
4474 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4475 caused it to match further down the string than it should.
4476
4477 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4478 was there because it doesn't seem to be needed, and it causes trouble on some
4479 systems, as it is not a Standard C header. It has been removed.
4480
4481 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4482 were reported on the Macintosh. The changes were from
4483
4484 while ((c = *(++ptr)) != 0 && c != '\n');
4485 to
4486 while ((c = *(++ptr)) != 0 && c != '\n') ;
4487
4488 Totally extraordinary, but if that's what it takes...
4489
4490 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4491 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4492 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4493 assumes the way PCRE uses memmove() (always moving upwards).
4494
4495 6. PCRE is being used in one environment where strchr() is not available. There
4496 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4497 faster code anyway.
4498
4499
4500 Version 3.1 09-Feb-00
4501 ---------------------
4502
4503 The only change in this release is the fixing of some bugs in Makefile.in for
4504 the "install" target:
4505
4506 (1) It was failing to install pcreposix.h.
4507
4508 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4509
4510
4511 Version 3.0 01-Feb-00
4512 ---------------------
4513
4514 1. Add support for the /+ modifier to perltest (to output $` like it does in
4515 pcretest).
4516
4517 2. Add support for the /g modifier to perltest.
4518
4519 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4520 matches null strings.
4521
4522 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4523 pattern. Perl treats empty patterns specially - it reuses the most recent
4524 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4525 effect.
4526
4527 5. The POSIX interface was broken in that it was just handing over the POSIX
4528 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4529 required a bigger vector, with some working space on the end. This means that
4530 the POSIX wrapper now has to get and free some memory, and copy the results.
4531
4532 6. Added some simple autoconf support, placing the test data and the
4533 documentation in separate directories, re-organizing some of the
4534 information files, and making it build pcre-config (a GNU standard). Also added
4535 libtool support for building PCRE as a shared library, which is now the
4536 default.
4537
4538 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4539 09 are not valid octal constants. Single digits will be used for minor values
4540 less than 10.
4541
4542 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4543 existing programs that set these in the POSIX interface can use PCRE without
4544 modification.
4545
4546 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4547 return all that pcre_info() returns, plus additional data. The pcre_info()
4548 function is retained for compatibility, but is considered to be obsolete.
4549
4550 10. Added experimental recursion feature (?R) to handle one common case that
4551 Perl 5.6 will be able to do with (?p{...}).
4552
4553 11. Added support for POSIX character classes like [:alpha:], which Perl is
4554 adopting.
4555
4556
4557 Version 2.08 31-Aug-99
4558 ----------------------
4559
4560 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4561 trying to match at the startoffset position, but instead was moving forward to
4562 the next newline as if a previous match had failed.
4563
4564 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4565 and could get into a loop if a null string was matched other than at the start
4566 of the subject.
4567
4568 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4569 be distinguished at compile time, and for completeness also added PCRE_DATE.
4570
4571 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4572 in GnuWin32 environments.
4573
4574
4575 Version 2.07 29-Jul-99
4576 ----------------------
4577
4578 1. The documentation is now supplied in plain text form and HTML as well as in
4579 the form of man page sources.
4580
4581 2. C++ compilers don't like assigning (void *) values to other pointer types.
4582 In particular this affects malloc(). Although there is no problem in Standard
4583 C, I've put in casts to keep C++ compilers happy.
4584
4585 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4586 should be (const char *).
4587
4588 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4589 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4590 However, I haven't made this a standard facility. The documentation doesn't
4591 mention it, and the Makefile doesn't support it.
4592
4593 5. The Makefile now contains an "install" target, with editable destinations at
4594 the top of the file. The pcretest program is not installed.
4595
4596 6. pgrep -V now gives the PCRE version number and date.
4597
4598 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4599 causing the entire string to be ignored, instead of just the last character.
4600
4601 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4602 non-matching string, it can take a very, very long time, even for strings of
4603 quite modest length, because of the nested recursion. PCRE now does better in
4604 some of these cases. It does this by remembering the last required literal
4605 character in the pattern, and pre-searching the subject to ensure it is present
4606 before running the real match. In other words, it applies a heuristic to detect
4607 some types of certain failure quickly, and in the above example, if presented
4608 with a string that has no trailing " it gives "no match" very quickly.
4609
4610 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4611 other alternatives are tried instead.
4612
4613
4614 Version 2.06 09-Jun-99
4615 ----------------------
4616
4617 1. Change pcretest's output for amount of store used to show just the code
4618 space, because the remainder (the data block) varies in size between 32-bit and
4619 64-bit systems.
4620
4621 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4622 start matching at. This allows lookbehinds to work when searching for multiple
4623 occurrences in a string.
4624
4625 3. Added additional options to pcretest for testing multiple occurrences:
4626
4627 /+ outputs the rest of the string that follows a match
4628 /g loops for multiple occurrences, using the new startoffset argument
4629 /G loops for multiple occurrences by passing an incremented pointer
4630
4631 4. PCRE wasn't doing the "first character" optimization for patterns starting
4632 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4633 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4634 the letter 'x'. On long subject strings, this gives a significant speed-up.
4635
4636
4637 Version 2.05 21-Apr-99
4638 ----------------------
4639
4640 1. Changed the type of magic_number from int to long int so that it works
4641 properly on 16-bit systems.
4642
4643 2. Fixed a bug which caused patterns starting with .* not to work correctly
4644 when the subject string contained newline characters. PCRE was assuming
4645 anchoring for such patterns in all cases, which is not correct because .* will
4646 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4647 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4648 must be retried after every newline in the subject.
4649
4650
4651 Version 2.04 18-Feb-99
4652 ----------------------
4653
4654 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4655 computation of the store needed to hold the pattern was incorrect (too large).
4656 If such patterns were nested a few deep, this could multiply and become a real
4657 problem.
4658
4659 2. Added /M option to pcretest to show the memory requirement of a specific
4660 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4661
4662 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4663 compiled in such a way that the backtracking after subsequent failure was
4664 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4665 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4666
4667
4668 Version 2.03 02-Feb-99
4669 ----------------------
4670
4671 1. Fixed typo and small mistake in man page.
4672
4673 2. Added 4th condition (GPL supersedes if conflict) and created separate
4674 LICENCE file containing the conditions.
4675
4676 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4677 Perl, that is the internal \ allows the delimiter to be included in the
4678 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4679 the final delimiter, add \ to the end of the pattern (to test the error).
4680
4681 4. Added the convenience functions for extracting substrings after a successful
4682 match. Updated pcretest to make it able to test these functions.
4683
4684
4685 Version 2.02 14-Jan-99
4686 ----------------------
4687
4688 1. Initialized the working variables associated with each extraction so that
4689 their saving and restoring doesn't refer to uninitialized store.
4690
4691 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4692 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4693 fix the problem.
4694
4695 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4696 calls, and wasn't printing the correct value for compiling calls. Increased the
4697 default value of LOOPREPEAT, and the number of significant figures in the
4698 times.
4699
4700 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4701
4702 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4703 a building problem on Windows NT with a FAT file system.
4704
4705
4706 Version 2.01 21-Oct-98
4707 ----------------------
4708
4709 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4710 to character tables built by pcre_maketables() in the current locale. If NULL
4711 is passed, the default tables are used.
4712
4713
4714 Version 2.00 24-Sep-98
4715 ----------------------
4716
4717 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4718 it any more.
4719
4720 2. Allow quantification of (?>) groups, and make it work correctly.
4721
4722 3. The first character computation wasn't working for (?>) groups.
4723
4724 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4725 end of the subject) and add 5.005's \z, which really does match only at the
4726 very end of the subject.
4727
4728 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4729
4730 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4731 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4732 localized options. All options to pcre_study() were also removed.
4733
4734 7. Add other new features from 5.005:
4735
4736 $(?<= positive lookbehind
4737 $(?<! negative lookbehind
4738 (?imsx-imsx) added the unsetting capability
4739 such a setting is global if at outer level; local otherwise
4740 (?imsx-imsx:) non-capturing groups with option setting
4741 (?(cond)re|re) conditional pattern matching
4742
4743 A backreference to itself in a repeated group matches the previous
4744 captured string.
4745
4746 8. General tidying up of studying (both automatic and via "study")
4747 consequential on the addition of new assertions.
4748
4749 9. As in 5.005, unlimited repeated groups that could match an empty substring
4750 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4751 runtime if any iteration does actually match an empty substring.
4752
4753 10. Include the RunTest script in the distribution.
4754
4755 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4756 discrepancies, some of which were old and were also with respect to 5.004. They
4757 have now been fixed.
4758
4759
4760 Version 1.09 28-Apr-98
4761 ----------------------
4762
4763 1. A negated single character class followed by a quantifier with a minimum
4764 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4765 program crashes, or just wrong answers. This did not apply to negated classes
4766 containing more than one character, or to minima other than one.
4767
4768
4769 Version 1.08 27-Mar-98
4770 ----------------------
4771
4772 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4773
4774 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4775 latter must appear before anything that relies on it in the pattern.
4776
4777
4778 Version 1.07 16-Feb-98
4779 ----------------------
4780
4781 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4782 repeat of a potentially empty string).
4783
4784
4785 Version 1.06 23-Jan-98
4786 ----------------------
4787
4788 1. Added Markus Oberhumer's little patches for C++.
4789
4790 2. Literal strings longer than 255 characters were broken.
4791
4792
4793 Version 1.05 23-Dec-97
4794 ----------------------
4795
4796 1. Negated character classes containing more than one character were failing if
4797 PCRE_CASELESS was set at run time.
4798
4799
4800 Version 1.04 19-Dec-97
4801 ----------------------
4802
4803 1. Corrected the man page, where some "const" qualifiers had been omitted.
4804
4805 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4806 input syntax.
4807
4808 3. Fixed memory leak which occurred when a regex with back references was
4809 matched with an offsets vector that wasn't big enough. The temporary memory
4810 that is used in this case wasn't being freed if the match failed.
4811
4812 4. Tidied pcretest to ensure it frees memory that it gets.
4813
4814 5. Temporary memory was being obtained in the case where the passed offsets
4815 vector was exactly big enough.
4816
4817 6. Corrected definition of offsetof() from change 5 below.
4818
4819 7. I had screwed up change 6 below and broken the rules for the use of
4820 setjmp(). Now fixed.
4821
4822
4823 Version 1.03 18-Dec-97
4824 ----------------------
4825
4826 1. A erroneous regex with a missing opening parenthesis was correctly
4827 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4828 on some systems.
4829
4830 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4831 it was reported that one broken compiler failed on the former because "code" is
4832 also an independent variable.
4833
4834 3. The erroneous regex a[]b caused an array overrun reference.
4835
4836 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4837 fail on data ending with that character. (It was going on too far, and checking
4838 the next character, typically a binary zero.) This was specific to the
4839 optimized code for single-character negative classes.
4840
4841 5. Added a contributed patch from the TIN world which does the following:
4842
4843 + Add an undef for memmove, in case the the system defines a macro for it.
4844
4845 + Add a definition of offsetof(), in case there isn't one. (I don't know
4846 the reason behind this - offsetof() is part of the ANSI standard - but
4847 it does no harm).
4848
4849 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4850 most of the places where whitespace preceded '#'. I have given up and
4851 allowed the remaining 2 cases to be at the margin.
4852
4853 + Rename some variables in pcre to eliminate shadowing. This seems very
4854 pedantic, but does no harm, of course.
4855
4856 6. Moved the call to setjmp() into its own function, to get rid of warnings
4857 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4858
4859 7. Constructs such as \d{8,} were compiling into the equivalent of
4860 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4861 outcome, but in this particular case used more store than had been allocated,
4862 which caused the bug to be discovered because it threw up an internal error.
4863
4864 8. The debugging code in both pcre and pcretest for outputting the compiled
4865 form of a regex was going wrong in the case of back references followed by
4866 curly-bracketed repeats.
4867
4868
4869 Version 1.02 12-Dec-97
4870 ----------------------
4871
4872 1. Typos in pcre.3 and comments in the source fixed.
4873
4874 2. Applied a contributed patch to get rid of places where it used to remove
4875 'const' from variables, and fixed some signed/unsigned and uninitialized
4876 variable warnings.
4877
4878 3. Added the "runtest" target to Makefile.
4879
4880 4. Set default compiler flag to -O2 rather than just -O.
4881
4882
4883 Version 1.01 19-Nov-97
4884 ----------------------
4885
4886 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4887 like /([ab]*)*/, that is, for classes with more than one character in them.
4888
4889 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4890 as /((?>a*))*/ (a PCRE_EXTRA facility).
4891
4892
4893 Version 1.00 18-Nov-97
4894 ----------------------
4895
4896 1. Added compile-time macros to support systems such as SunOS4 which don't have
4897 memmove() or strerror() but have other things that can be used instead.
4898
4899 2. Arranged that "make clean" removes the executables.
4900
4901
4902 Version 0.99 27-Oct-97
4903 ----------------------
4904
4905 1. Fixed bug in code for optimizing classes with only one character. It was
4906 initializing a 32-byte map regardless, which could cause it to run off the end
4907 of the memory it had got.
4908
4909 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4910
4911
4912 Version 0.98 22-Oct-97
4913 ----------------------
4914
4915 1. Fixed bug in code for handling temporary memory usage when there are more
4916 back references than supplied space in the ovector. This could cause segfaults.
4917
4918
4919 Version 0.97 21-Oct-97
4920 ----------------------
4921
4922 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4923
4924 2. Optimized negated single characters not to use a bit map.
4925
4926 3. Brought error texts together as macro definitions; clarified some of them;
4927 fixed one that was wrong - it said "range out of order" when it meant "invalid
4928 escape sequence".
4929
4930 4. Changed some char * arguments to const char *.
4931
4932 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4933
4934 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
4935 pcretest.
4936
4937
4938 Version 0.96 16-Oct-97
4939 ----------------------
4940
4941 1. Added a simple "pgrep" utility to the distribution.
4942
4943 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
4944 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
4945 where "ddd" means "one or more decimal digits".
4946
4947 3. Fixed serious bug. If a pattern had a back reference, but the call to
4948 pcre_exec() didn't supply a large enough ovector to record the related
4949 identifying subpattern, the match always failed. PCRE now remembers the number
4950 of the largest back reference, and gets some temporary memory in which to save
4951 the offsets during matching if necessary, in order to ensure that
4952 backreferences always work.
4953
4954 4. Increased the compatibility with Perl in a number of ways:
4955
4956 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
4957 to request this handling. The option can be set at compile or exec time.
4958
4959 (b) $ matches before a terminating newline by default; an option
4960 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
4961 mode). The option can be set at compile or exec time.
4962
4963 (c) The handling of \ followed by a digit other than 0 is now supposed to be
4964 the same as Perl's. If the decimal number it represents is less than 10
4965 or there aren't that many previous left capturing parentheses, an octal
4966 escape is read. Inside a character class, it's always an octal escape,
4967 even if it is a single digit.
4968
4969 (d) An escaped but undefined alphabetic character is taken as a literal,
4970 unless PCRE_EXTRA is set. Currently this just reserves the remaining
4971 escapes.
4972
4973 (e) {0} is now permitted. (The previous item is removed from the compiled
4974 pattern).
4975
4976 5. Changed all the names of code files so that the basic parts are no longer
4977 than 10 characters, and abolished the teeny "globals.c" file.
4978
4979 6. Changed the handling of character classes; they are now done with a 32-byte
4980 bit map always.
4981
4982 7. Added the -d and /D options to pcretest to make it possible to look at the
4983 internals of compilation without having to recompile pcre.
4984
4985
4986 Version 0.95 23-Sep-97
4987 ----------------------
4988
4989 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
4990 \x20 at the start of a run of normal characters. These were being treated as
4991 real characters, instead of the source characters being re-checked.
4992
4993
4994 Version 0.94 18-Sep-97
4995 ----------------------
4996
4997 1. The functions are now thread-safe, with the caveat that the global variables
4998 containing pointers to malloc() and free() or alternative functions are the
4999 same for all threads.
5000
5001 2. Get pcre_study() to generate a bitmap of initial characters for non-
5002 anchored patterns when this is possible, and use it if passed to pcre_exec().
5003
5004
5005 Version 0.93 15-Sep-97
5006 ----------------------
5007
5008 1. /(b)|(:+)/ was computing an incorrect first character.
5009
5010 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
5011 but not actually doing anything yet.
5012
5013 3. Treat "-" characters in classes that cannot be part of ranges as literals,
5014 as Perl does (e.g. [-az] or [az-]).
5015
5016 4. Set the anchored flag if a branch starts with .* or .*? because that tests
5017 all possible positions.
5018
5019 5. Split up into different modules to avoid including unneeded functions in a
5020 compiled binary. However, compile and exec are still in one module. The "study"
5021 function is split off.
5022
5023 6. The character tables are now in a separate module whose source is generated
5024 by an auxiliary program - but can then be edited by hand if required. There are
5025 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
5026 toupper() in the code.
5027
5028 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
5029 make them global. Abolish the function for setting them, as the caller can now
5030 set them directly.
5031
5032
5033 Version 0.92 11-Sep-97
5034 ----------------------
5035
5036 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
5037 (e.g. /a{1,3}/) was broken (I mis-optimized it).
5038
5039 2. Caseless matching was not working in character classes if the characters in
5040 the pattern were in upper case.
5041
5042 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
5043
5044 4. Make PCRE_ANCHORED public and accept as a compile option.
5045
5046 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
5047 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
5048 pass them.
5049
5050 6. Give an error if bad option bits passed at compile or run time.
5051
5052 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
5053 pcretest to cause it to pass that flag.
5054
5055 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
5056 options, and the first character, if set.
5057
5058 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
5059
5060
5061 Version 0.91 10-Sep-97
5062 ----------------------
5063
5064 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
5065 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
5066
5067 2. PCRE was looping on encountering an indefinitely repeated back reference to
5068 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
5069 Perl does - treats the match as successful.
5070
5071 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5