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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1392 - (show annotations)
Wed Nov 6 18:00:09 2013 UTC (6 years ago) by ph10
File size: 256042 byte(s)
Error occurred while calculating annotation data.
Give errors for [A-\d] and [a-[:digit:]] etc.
1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.34 xx-xxxx-201x
5 --------------------------
6
7 1. Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
8 executable memory. Patch inspired by Carsten Klein.
9
10 2. ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
11 this macro is never tested and has no effect, because the work to support
12 coverage involves only compiling and linking options and special targets in
13 the Makefile. The comment in config.h implied that defining the macro would
14 enable coverage support, which is totally false. There was also support for
15 setting this macro in the CMake files (my fault, I just copied it from
16 configure). SUPPORT_GCOV has now been removed.
17
18 3. Make a small performance improvement in strlen16() and strlen32() in
19 pcretest.
20
21 4. Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
22 detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
23 these cases). There was also one in pcretest.c.
24
25 5. Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
26
27 6. In UTF mode, the code for checking whether a group could match an empty
28 string (which is used for indefinitely repeated groups to allow for
29 breaking an infinite loop) was broken when the group contained a repeated
30 negated single-character class with a character that occupied more than one
31 data item and had a minimum repetition of zero (for example, [^\x{100}]* in
32 UTF-8 mode). The effect was undefined: the group might or might not be
33 deemed as matching an empty string, or the program might have crashed.
34
35 7. The code for checking whether a group could match an empty string was not
36 recognizing that \h, \H, \v, \V, and \R must match a character.
37
38 8. Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
39 an empty string. If it can, pcretest shows this in its information output.
40
41 9. Fixed two related bugs that applied to Unicode extended grapheme clusters
42 that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
43 matched by pcre_exec() without using JIT:
44
45 (a) If the rest of the pattern did not match after a maximal run of
46 grapheme clusters, the code for backing up to try with fewer of them
47 did not always back up over a full grapheme when characters that do not
48 have the modifier quality were involved, e.g. Hangul syllables.
49
50 (b) If the match point in a subject started with modifier character, and
51 there was no match, the code could incorrectly back up beyond the match
52 point, and potentially beyond the first character in the subject,
53 leading to a segfault or an incorrect match result.
54
55 10. A conditional group with an assertion condition could lead to PCRE
56 recording an incorrect first data item for a match if no other first data
57 item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
58 first data item, and therefore matched "ca" after "c" instead of at the
59 start.
60
61 11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
62 bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
63
64 12. The source of pcregrep now includes z/OS-specific code so that it can be
65 compiled for z/OS as part of the special z/OS distribution.
66
67 13. Added the -T and -TM options to pcretest.
68
69 14. The code in pcre_compile.c for creating the table of named capturing groups
70 has been refactored. Instead of creating the table dynamically during the
71 actual compiling pass, the information is remembered during the pre-compile
72 pass (on the stack unless there are more than 20 named groups, in which
73 case malloc() is used) and the whole table is created before the actual
74 compile happens. This has simplified the code (it is now nearly 150 lines
75 shorter) and prepared the way for better handling of references to groups
76 with duplicate names.
77
78 15. A back reference to a named subpattern when there is more than one of the
79 same name now checks them in the order in which they appear in the pattern.
80 The first one that is set is used for the reference. Previously only the
81 first one was inspected. This change makes PCRE more compatible with Perl.
82
83 16. Unicode character properties were updated from Unicode 6.3.0.
84
85 17. The compile-time code for auto-possessification has been refactored, based
86 on a patch by Zoltan Herczeg. It now happens after instead of during
87 compilation. The code is cleaner, and more cases are handled. The option
88 PCRE_NO_AUTO_POSSESSIFY is added for testing purposes, and the -O and /O
89 options in pcretest are provided to set it.
90
91 18. The character VT has been added to the set of characters that match \s and
92 are generally treated as white space, following this same change in Perl
93 5.18. There is now no difference between "Perl space" and "POSIX space".
94
95 19. The code for checking named groups as conditions, either for being set or
96 for being recursed, has been refactored (this is related to 14 and 15
97 above). Processing unduplicated named groups should now be as fast at
98 numerical groups, and processing duplicated groups should be faster than
99 before.
100
101 20. Two patches to the CMake build system, by Alexander Barkov:
102
103 (1) Replace the "source" command by "." in CMakeLists.txt because
104 "source" is a bash-ism.
105
106 (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
107 without these the CMake build does not work on Solaris.
108
109 21. Perl has changed its handling of \8 and \9. If there is no previously
110 encountered capturing group of those numbers, they are treated as the
111 literal characters 8 and 9 instead of a binary zero followed by the
112 literals. PCRE now does the same.
113
114 22. Following Perl, added \o{} to specify codepoints in octal, making it
115 possible to specify values greater than 0777 and also making them
116 unambiguous.
117
118 23. Perl now gives an error for missing closing braces after \x{... instead of
119 treating the string as literal. PCRE now does the same.
120
121 24. RunTest used to grumble if an inappropriate test was selected explicitly,
122 but just skip it when running all tests. This make it awkward to run ranges
123 of tests when one of them was inappropriate. Now it just skips any
124 inappropriate tests, as it always did when running all tests.
125
126 25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
127 character types such as \d or \w, too many callouts were inserted, and the
128 data that they returned was rubbish.
129
130 26. In UCP mode, \s was not matching two of the characters that Perl matches,
131 namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
132 were matched by \h. The code has now been refactored so that the lists of
133 the horizontal and vertical whitespace characters used for \h and \v (which
134 are defined only in one place) are now also used for \s.
135
136 27. Add JIT support for the 64 bit TileGX architecture.
137 Patch by Jiong Wang (Tilera Corporation).
138
139 28. Possessive quantifiers for classes (both explicit and automatically
140 generated) now use special opcodes instead of wrapping in ONCE brackets.
141
142 29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
143 (because it's meaningless), this was not happening when PCRE_CASELESS was
144 set. Not wrong, but inefficient.
145
146 30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
147 properties for \w, \d, etc) is present in a test regex. Otherwise if the
148 test contains no characters greater than 255, Perl doesn't realise it
149 should be using Unicode semantics.
150
151 31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
152 [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
153 does in Unicode mode.
154
155 32. Added the "forbid" facility to pcretest so that putting tests into the
156 wrong test files can sometimes be quickly detected.
157
158 33. There is now a limit (default 250) on the depth of nesting of parentheses.
159 This limit is imposed to control the amount of system stack used at compile
160 time. It can be changed at build time by --with-parens-nest-limit=xxx or
161 the equivalent in CMake.
162
163 34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
164 errors. Perl warns for these when in warning mode, but PCRE has no facility
165 for giving warnings.
166
167
168 Version 8.33 28-May-2013
169 ------------------------
170
171 1. Added 'U' to some constants that are compared to unsigned integers, to
172 avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
173 variables that are added to signed variables, to ensure the result is
174 signed and can be negated.
175
176 2. Applied patch by Daniel Richard G for quashing MSVC warnings to the
177 CMake config files.
178
179 3. Revise the creation of config.h.generic so that all boolean macros are
180 #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
181 overriding via -D on the command line possible.
182
183 4. Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
184 to unsigned int is reported to make a quite noticeable speed difference in
185 a specific Windows environment. Testing on Linux did also appear to show
186 some benefit (and it is clearly not harmful). Also fixed the definition of
187 Xop which should be unsigned.
188
189 5. Related to (4), changing the definition of the intermediate variable cc
190 in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
191 improvements.
192
193 6. Fix forward search in JIT when link size is 3 or greater. Also removed some
194 unnecessary spaces.
195
196 7. Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
197 and later.
198
199 8. Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
200
201 9. Optimizing fast_forward_start_bits in JIT.
202
203 10. Adding support for callouts in JIT, and fixing some issues revealed
204 during this work. Namely:
205
206 (a) Unoptimized capturing brackets incorrectly reset on backtrack.
207
208 (b) Minimum length was not checked before the matching is started.
209
210 11. The value of capture_last that is passed to callouts was incorrect in some
211 cases when there was a capture on one path that was subsequently abandoned
212 after a backtrack. Also, the capture_last value is now reset after a
213 recursion, since all captures are also reset in this case.
214
215 12. The interpreter no longer returns the "too many substrings" error in the
216 case when an overflowing capture is in a branch that is subsequently
217 abandoned after a backtrack.
218
219 13. In the pathological case when an offset vector of size 2 is used, pcretest
220 now prints out the matched string after a yield of 0 or 1.
221
222 14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
223 Only supported by the JIT compiler at the moment.
224
225 15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
226
227 16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
228 offset of the starting point of the matching process, provided the offsets
229 vector is large enough.
230
231 17. The \A escape now records a lookbehind value of 1, though its execution
232 does not actually inspect the previous character. This is to ensure that,
233 in partial multi-segment matching, at least one character from the old
234 segment is retained when a new segment is processed. Otherwise, if there
235 are no lookbehinds in the pattern, \A might match incorrectly at the start
236 of a new segment.
237
238 18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
239
240 19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
241 gives some modest performance improvement in 8-bit mode.
242
243 20. Added the PCRE-specific property \p{Xuc} for matching characters that can
244 be expressed in certain programming languages using Universal Character
245 Names.
246
247 21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
248 which points out that "non characters" are not "characters that may not
249 appear in Unicode strings" but rather "characters that are reserved for
250 internal use and have only local meaning".
251
252 22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
253 there was a conditional group that depended on an assertion, if the
254 assertion was false, the callout that immediately followed the alternation
255 in the condition was skipped when pcre_exec() was used for matching.
256
257 23. Allow an explicit callout to be inserted before an assertion that is the
258 condition for a conditional group, for compatibility with automatic
259 callouts, which always insert a callout at this point.
260
261 24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
262 confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
263
264 25. (*PRUNE) is now supported by the JIT compiler.
265
266 26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
267
268 27. Fix the case where there are two or more SKIPs with arguments that may be
269 ignored.
270
271 28. (*SKIP) is now supported by the JIT compiler.
272
273 29. (*THEN) is now supported by the JIT compiler.
274
275 30. Update RunTest with additional test selector options.
276
277 31. The way PCRE handles backtracking verbs has been changed in two ways.
278
279 (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
280 SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
281 backtracking. In some cases this makes it more Perl-compatible, but Perl's
282 rather obscure rules do not always do the same thing.
283
284 (2) Previously, backtracking verbs were confined within assertions. This is
285 no longer the case for positive assertions, except for (*ACCEPT). Again,
286 this sometimes improves Perl compatibility, and sometimes does not.
287
288 32. A number of tests that were in test 2 because Perl did things differently
289 have been moved to test 1, because either Perl or PCRE has changed, and
290 these tests are now compatible.
291
292 32. Backtracking control verbs are now handled in the same way in JIT and
293 interpreter.
294
295 33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
296 contained a forward subroutine reference caused a compile error.
297
298 34. Auto-detect and optimize limited repetitions in JIT.
299
300 35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
301 blocking (*UTF) etc.
302
303 36. In the interpreter, maximizing pattern repetitions for characters and
304 character types now use tail recursion, which reduces stack usage.
305
306 37. The value of the max lookbehind was not correctly preserved if a compiled
307 and saved regex was reloaded on a host of different endianness.
308
309 38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
310 of the compiled pattern block, expand the flags field from 16 to 32 bits
311 because it was almost full.
312
313 39. Try madvise first before posix_madvise.
314
315 40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
316 with a pattern such as ^$. It has taken 4 years for anybody to notice! The
317 original change locked out all matches of empty strings. This has been
318 changed so that one match of an empty string per line is recognized.
319 Subsequent searches on the same line (for colouring or for --only-matching,
320 for example) do not recognize empty strings.
321
322 41. Applied a user patch to fix a number of spelling mistakes in comments.
323
324 42. Data lines longer than 65536 caused pcretest to crash.
325
326 43. Clarified the data type for length and startoffset arguments for pcre_exec
327 and pcre_dfa_exec in the function-specific man pages, where they were
328 explicitly stated to be in bytes, never having been updated. I also added
329 some clarification to the pcreapi man page.
330
331 44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
332 a segmentation fault.
333
334
335 Version 8.32 30-November-2012
336 -----------------------------
337
338 1. Improved JIT compiler optimizations for first character search and single
339 character iterators.
340
341 2. Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
342 Patch by Daniel Richard G.
343
344 3. Single character iterator optimizations in the JIT compiler.
345
346 4. Improved JIT compiler optimizations for character ranges.
347
348 5. Rename the "leave" variable names to "quit" to improve WinCE compatibility.
349 Reported by Giuseppe D'Angelo.
350
351 6. The PCRE_STARTLINE bit, indicating that a match can occur only at the start
352 of a line, was being set incorrectly in cases where .* appeared inside
353 atomic brackets at the start of a pattern, or where there was a subsequent
354 *PRUNE or *SKIP.
355
356 7. Improved instruction cache flush for POWER/PowerPC.
357 Patch by Daniel Richard G.
358
359 8. Fixed a number of issues in pcregrep, making it more compatible with GNU
360 grep:
361
362 (a) There is now no limit to the number of patterns to be matched.
363
364 (b) An error is given if a pattern is too long.
365
366 (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
367 are now supported.
368
369 (d) --exclude-from and --include-from (multiple use) have been added.
370
371 (e) Exclusions and inclusions now apply to all files and directories, not
372 just to those obtained from scanning a directory recursively.
373
374 (f) Multiple uses of -f and --file-list are now supported.
375
376 (g) In a Windows environment, the default for -d has been changed from
377 "read" (the GNU grep default) to "skip", because otherwise the presence
378 of a directory in the file list provokes an error.
379
380 (h) The documentation has been revised and clarified in places.
381
382 9. Improve the matching speed of capturing brackets.
383
384 10. Changed the meaning of \X so that it now matches a Unicode extended
385 grapheme cluster.
386
387 11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
388 out POSIX threads when JIT support is configured.
389
390 12. Added support for PCRE_STUDY_EXTRA_NEEDED.
391
392 13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
393 structure could go wrong in environments where size_t is not the same size
394 as int.
395
396 14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
397 set.
398
399 15. The EBCDIC support had decayed; later updates to the code had included
400 explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
401 general tidy up of EBCDIC-related issues, and the documentation was also
402 not quite right. There is now a test that can be run on ASCII systems to
403 check some of the EBCDIC-related things (but is it not a full test).
404
405 16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
406 in a small tidy to the code.
407
408 17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
409
410 18. If the --only-matching (-o) option in pcregrep is specified multiple
411 times, each one causes appropriate output. For example, -o1 -o2 outputs the
412 substrings matched by the 1st and 2nd capturing parentheses. A separating
413 string can be specified by --om-separator (default empty).
414
415 19. Improving the first n character searches.
416
417 20. Turn case lists for horizontal and vertical white space into macros so that
418 they are defined only once.
419
420 21. This set of changes together give more compatible Unicode case-folding
421 behaviour for characters that have more than one other case when UCP
422 support is available.
423
424 (a) The Unicode property table now has offsets into a new table of sets of
425 three or more characters that are case-equivalent. The MultiStage2.py
426 script that generates these tables (the pcre_ucd.c file) now scans
427 CaseFolding.txt instead of UnicodeData.txt for character case
428 information.
429
430 (b) The code for adding characters or ranges of characters to a character
431 class has been abstracted into a generalized function that also handles
432 case-independence. In UTF-mode with UCP support, this uses the new data
433 to handle characters with more than one other case.
434
435 (c) A bug that is fixed as a result of (b) is that codepoints less than 256
436 whose other case is greater than 256 are now correctly matched
437 caselessly. Previously, the high codepoint matched the low one, but not
438 vice versa.
439
440 (d) The processing of \h, \H, \v, and \ in character classes now makes use
441 of the new class addition function, using character lists defined as
442 macros alongside the case definitions of 20 above.
443
444 (e) Caseless back references now work with characters that have more than
445 one other case.
446
447 (f) General caseless matching of characters with more than one other case
448 is supported.
449
450 22. Unicode character properties were updated from Unicode 6.2.0
451
452 23. Improved CMake support under Windows. Patch by Daniel Richard G.
453
454 24. Add support for 32-bit character strings, and UTF-32
455
456 25. Major JIT compiler update (code refactoring and bugfixing).
457 Experimental Sparc 32 support is added.
458
459 26. Applied a modified version of Daniel Richard G's patch to create
460 pcre.h.generic and config.h.generic by "make" instead of in the
461 PrepareRelease script.
462
463 27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
464 pcre_compile.c when checking for a zero character.
465
466 28. Introducing a native interface for JIT. Through this interface, the compiled
467 machine code can be directly executed. The purpose of this interface is to
468 provide fast pattern matching, so several sanity checks are not performed.
469 However, feature tests are still performed. The new interface provides
470 1.4x speedup compared to the old one.
471
472 29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
473 the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
474 was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
475
476 30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
477 greater than 0x7fffffff (which cannot be represented in UTF-8, even under
478 the "old" RFC 2279). Instead, it ended up passing a negative length to
479 pcre_exec().
480
481 31. Add support for GCC's visibility feature to hide internal functions.
482
483 32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
484 "unknown -C option" after outputting 0 or 1.
485
486 33. There is now support for generating a code coverage report for the test
487 suite in environments where gcc is the compiler and lcov is installed. This
488 is mainly for the benefit of the developers.
489
490 34. If PCRE is built with --enable-valgrind, certain memory regions are marked
491 unaddressable using valgrind annotations, allowing valgrind to detect
492 invalid memory accesses. This is mainly for the benefit of the developers.
493
494 25. (*UTF) can now be used to start a pattern in any of the three libraries.
495
496 26. Give configure error if --enable-cpp but no C++ compiler found.
497
498
499 Version 8.31 06-July-2012
500 -------------------------
501
502 1. Fixing a wrong JIT test case and some compiler warnings.
503
504 2. Removed a bashism from the RunTest script.
505
506 3. Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
507 to unsigned type, result still unsigned" that was given by an MS compiler
508 on encountering the code "-sizeof(xxx)".
509
510 4. Partial matching support is added to the JIT compiler.
511
512 5. Fixed several bugs concerned with partial matching of items that consist
513 of more than one character:
514
515 (a) /^(..)\1/ did not partially match "aba" because checking references was
516 done on an "all or nothing" basis. This also applied to repeated
517 references.
518
519 (b) \R did not give a hard partial match if \r was found at the end of the
520 subject.
521
522 (c) \X did not give a hard partial match after matching one or more
523 characters at the end of the subject.
524
525 (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
526 a partial match for the string "\r".
527
528 (e) When newline was set to CRLF, the metacharacter "." did not recognize
529 a partial match for a CR character at the end of the subject string.
530
531 6. If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
532 running pcretest, the text "(JIT)" added to the output whenever JIT is
533 actually used to run the match.
534
535 7. Individual JIT compile options can be set in pcretest by following -s+[+]
536 or /S+[+] with a digit between 1 and 7.
537
538 8. OP_NOT now supports any UTF character not just single-byte ones.
539
540 9. (*MARK) control verb is now supported by the JIT compiler.
541
542 10. The command "./RunTest list" lists the available tests without actually
543 running any of them. (Because I keep forgetting what they all are.)
544
545 11. Add PCRE_INFO_MAXLOOKBEHIND.
546
547 12. Applied a (slightly modified) user-supplied patch that improves performance
548 when the heap is used for recursion (compiled with --disable-stack-for-
549 recursion). Instead of malloc and free for each heap frame each time a
550 logical recursion happens, frames are retained on a chain and re-used where
551 possible. This sometimes gives as much as 30% improvement.
552
553 13. As documented, (*COMMIT) is now confined to within a recursive subpattern
554 call.
555
556 14. As documented, (*COMMIT) is now confined to within a positive assertion.
557
558 15. It is now possible to link pcretest with libedit as an alternative to
559 libreadline.
560
561 16. (*COMMIT) control verb is now supported by the JIT compiler.
562
563 17. The Unicode data tables have been updated to Unicode 6.1.0.
564
565 18. Added --file-list option to pcregrep.
566
567 19. Added binary file support to pcregrep, including the -a, --binary-files,
568 -I, and --text options.
569
570 20. The madvise function is renamed for posix_madvise for QNX compatibility
571 reasons. Fixed by Giuseppe D'Angelo.
572
573 21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
574
575 22. Changed the option for creating symbolic links for 16-bit man pages from
576 -s to -sf so that re-installing does not cause issues.
577
578 23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
579
580 24. Fixed a very old bug in pcretest that caused errors with restarted DFA
581 matches in certain environments (the workspace was not being correctly
582 retained). Also added to pcre_dfa_exec() a simple plausibility check on
583 some of the workspace data at the beginning of a restart.
584
585 25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
586 was not doing so when it should - probably a typo introduced by SVN 528
587 (change 8.10/14).
588
589 26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
590 \w+ when the character tables indicated that \x{c4} was a word character.
591 There were several related cases, all because the tests for doing a table
592 lookup were testing for characters less than 127 instead of 255.
593
594 27. If a pattern contains capturing parentheses that are not used in a match,
595 their slots in the ovector are set to -1. For those that are higher than
596 any matched groups, this happens at the end of processing. In the case when
597 there were back references that the ovector was too small to contain
598 (causing temporary malloc'd memory to be used during matching), and the
599 highest capturing number was not used, memory off the end of the ovector
600 was incorrectly being set to -1. (It was using the size of the temporary
601 memory instead of the true size.)
602
603 28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
604 ovector size, it uses memory at the end of the block that it has got.
605
606 29. Check for an overlong MARK name and give an error at compile time. The
607 limit is 255 for the 8-bit library and 65535 for the 16-bit library.
608
609 30. JIT compiler update.
610
611 31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
612 Rill for the patch.
613
614 32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
615
616 33. Variable renamings in the PCRE-JIT compiler. No functionality change.
617
618 34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
619 SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
620 was enabled.
621
622 35. Improve JIT code generation for greedy plus quantifier.
623
624 36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
625 1 to "aa" instead of to an empty string. The bug affected repeated groups
626 that could potentially match an empty string.
627
628 37. Optimizing single character iterators in JIT.
629
630 38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
631 the same checks as \x{...} characters in non-JavaScript mode. Specifically,
632 codepoints that are too big for the mode are faulted, and in a UTF mode,
633 disallowed codepoints are also faulted.
634
635 39. If PCRE was compiled with UTF support, in three places in the DFA
636 matcher there was code that should only have been obeyed in UTF mode, but
637 was being obeyed unconditionally. In 8-bit mode this could cause incorrect
638 processing when bytes with values greater than 127 were present. In 16-bit
639 mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
640 both cases the values are those that cannot be the first data item in a UTF
641 character. The three items that might have provoked this were recursions,
642 possessively repeated groups, and atomic groups.
643
644 40. Ensure that libpcre is explicitly listed in the link commands for pcretest
645 and pcregrep, because some OS require shared objects to be explicitly
646 passed to ld, causing the link step to fail if they are not.
647
648 41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
649 mode, patterns that started with \h* or \R* might be incorrectly matched.
650
651
652 Version 8.30 04-February-2012
653 -----------------------------
654
655 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
656 name is defined in ctype.h.
657
658 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
659 only in quite long subpatterns.
660
661 3. Removed the function pcre_info(), which has been obsolete and deprecated
662 since it was replaced by pcre_fullinfo() in February 2000.
663
664 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
665 match a (*MARK), and the match failed at the start of the subject, a
666 reference to memory before the start of the subject could occur. This bug
667 was introduced by fix 17 of release 8.21.
668
669 5. A reference to an unset group with zero minimum repetition was giving
670 totally wrong answers (in non-JavaScript-compatibility mode). For example,
671 /(another)?(\1?)test/ matched against "hello world test". This bug was
672 introduced in release 8.13.
673
674 6. Add support for 16-bit character strings (a large amount of work involving
675 many changes and refactorings).
676
677 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
678 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
679 from a file.
680
681 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
682 rounding is not applied in this particular case).
683
684 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
685 if they appear, or are escaped, in patterns.
686
687 10. Get rid of a number of -Wunused-but-set-variable warnings.
688
689 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
690 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
691 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
692 also returns the mark "x". This bug applied to capturing parentheses,
693 non-capturing parentheses, and atomic parentheses. It also applied to some
694 assertions.
695
696 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
697 information out of configure.ac instead of relying on pcre.h.generic, which
698 is not stored in the repository.
699
700 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
701 -lreadline.
702
703 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
704
705 15. Applied Graycode's patch to put the top-level frame on the stack rather
706 than the heap when not using the stack for recursion. This gives a
707 performance improvement in many cases when recursion is not deep.
708
709 16. Experimental code added to "pcretest -C" to output the stack frame size.
710
711
712 Version 8.21 12-Dec-2011
713 ------------------------
714
715 1. Updating the JIT compiler.
716
717 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
718 are added as well.
719
720 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
721 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
722 calling _pcre_jit_exec. Some extra comments are added.
723
724 4. (*MARK) settings inside atomic groups that do not contain any capturing
725 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
726 was introduced by change 18 for 8.20.
727
728 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
729 ECMA-262 standard.
730
731 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
732 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
733 This bug was probably introduced by change 9 of 8.13.
734
735 7. While fixing 6 above, I noticed that a number of other items were being
736 incorrectly rejected as "not fixed length". This arose partly because newer
737 opcodes had not been added to the fixed-length checking code. I have (a)
738 corrected the bug and added tests for these items, and (b) arranged for an
739 error to occur if an unknown opcode is encountered while checking for fixed
740 length instead of just assuming "not fixed length". The items that were
741 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
742 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
743 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
744
745 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
746 being incorrectly compiled and would have given unpredicatble results.
747
748 9. A possessively repeated subpattern with minimum repeat count greater than
749 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
750 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
751 the first (A) could occur when it should not.
752
753 10. Add a cast and remove a redundant test from the code.
754
755 11. JIT should use pcre_malloc/pcre_free for allocation.
756
757 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
758 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
759 is anything other than /usr, -L is still shown).
760
761 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
762
763 14. Perl does not support \N without a following name in a [] class; PCRE now
764 also gives an error.
765
766 15. If a forward reference was repeated with an upper limit of around 2000,
767 it caused the error "internal error: overran compiling workspace". The
768 maximum number of forward references (including repeats) was limited by the
769 internal workspace, and dependent on the LINK_SIZE. The code has been
770 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
771 the default depends on LINK_SIZE. There is a new upper limit (for safety)
772 of around 200,000 forward references. While doing this, I also speeded up
773 the filling in of repeated forward references.
774
775 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
776 incorrectly expecting the subject to contain another "a" after the start.
777
778 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
779 in the match, the SKIP should be ignored. This was not happening; instead
780 the SKIP was being treated as NOMATCH. For patterns such as
781 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
782 tested.
783
784 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
785 now much more compatible with Perl, in particular in cases where the result
786 is a non-match for a non-anchored pattern. For example, if
787 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
788 "m", where previously it did not return a name. A side effect of this
789 change is that for partial matches, the last encountered mark name is
790 returned, as for non matches. A number of tests that were previously not
791 Perl-compatible have been moved into the Perl-compatible test files. The
792 refactoring has had the pleasing side effect of removing one argument from
793 the match() function, thus reducing its stack requirements.
794
795 19. If the /S+ option was used in pcretest to study a pattern using JIT,
796 subsequent uses of /S (without +) incorrectly behaved like /S+.
797
798 21. Retrieve executable code size support for the JIT compiler and fixing
799 some warnings.
800
801 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
802 not work when the shorter character appeared right at the end of the
803 subject string.
804
805 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
806 systems.
807
808 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
809 output it when the /M option is used in pcretest.
810
811 25. The CheckMan script was not being included in the distribution. Also, added
812 an explicit "perl" to run Perl scripts from the PrepareRelease script
813 because this is reportedly needed in Windows.
814
815 26. If study data was being save in a file and studying had not found a set of
816 "starts with" bytes for the pattern, the data written to the file (though
817 never used) was taken from uninitialized memory and so caused valgrind to
818 complain.
819
820 27. Updated RunTest.bat as provided by Sheri Pierce.
821
822 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
823
824 29. Computation of memory usage for the table of capturing group names was
825 giving an unnecessarily large value.
826
827
828 Version 8.20 21-Oct-2011
829 ------------------------
830
831 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
832 a POSIX class. After further experiments with Perl, which convinced me that
833 Perl has bugs and confusions, a closing square bracket is no longer allowed
834 in a POSIX name. This bug also affected patterns with classes that started
835 with full stops.
836
837 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
838 captured substring, but while checking the failing first alternative,
839 substring 1 is temporarily captured. If the output vector supplied to
840 pcre_exec() was not big enough for this capture, the yield of the function
841 was still zero ("insufficient space for captured substrings"). This cannot
842 be totally fixed without adding another stack variable, which seems a lot
843 of expense for a edge case. However, I have improved the situation in cases
844 such as /(a)(b)x|abc/ matched against "abc", where the return code
845 indicates that fewer than the maximum number of slots in the ovector have
846 been set.
847
848 3. Related to (2) above: when there are more back references in a pattern than
849 slots in the output vector, pcre_exec() uses temporary memory during
850 matching, and copies in the captures as far as possible afterwards. It was
851 using the entire output vector, but this conflicts with the specification
852 that only 2/3 is used for passing back captured substrings. Now it uses
853 only the first 2/3, for compatibility. This is, of course, another edge
854 case.
855
856 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
857 main code base, and can be used by building with --enable-jit. When this is
858 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
859 runtime --no-jit option is given.
860
861 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
862 ovector, the return from the function was zero, implying that there were
863 other matches that did not fit. The correct "exactly full" value is now
864 returned.
865
866 6. If a subpattern that was called recursively or as a subroutine contained
867 (*PRUNE) or any other control that caused it to give a non-standard return,
868 invalid errors such as "Error -26 (nested recursion at the same subject
869 position)" or even infinite loops could occur.
870
871 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
872 computing the minimum length on reaching *ACCEPT, and so ended up with the
873 wrong value of 1 rather than 0. Further investigation indicates that
874 computing a minimum subject length in the presence of *ACCEPT is difficult
875 (think back references, subroutine calls), and so I have changed the code
876 so that no minimum is registered for a pattern that contains *ACCEPT.
877
878 8. If (*THEN) was present in the first (true) branch of a conditional group,
879 it was not handled as intended. [But see 16 below.]
880
881 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
882 Sheri Pierce.
883
884 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
885 the first byte in a match must be "a".
886
887 11. Change 17 for 8.13 increased the recursion depth for patterns like
888 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
889 pattern contains any instances of (*THEN). If it does not, the old
890 optimizations are restored. It would be nice to do this on a per-group
891 basis, but at the moment that is not feasible.
892
893 12. In some environments, the output of pcretest -C is CRLF terminated. This
894 broke RunTest's code that checks for the link size. A single white space
895 character after the value is now allowed for.
896
897 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
898 For "fr", it uses the Windows-specific input and output files.
899
900 14. If (*THEN) appeared in a group that was called recursively or as a
901 subroutine, it did not work as intended. [But see next item.]
902
903 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
904 pattern fragments (but not containing any | characters). If A and B are
905 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
906 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
907 D. In other words, Perl considers parentheses that do not contain any |
908 characters to be part of a surrounding alternative, whereas PCRE was
909 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
910 differently. PCRE now behaves in the same way as Perl, except in the case
911 of subroutine/recursion calls such as (?1) which have in any case always
912 been different (but PCRE had them first :-).
913
914 16. Related to 15 above: Perl does not treat the | in a conditional group as
915 creating alternatives. Such a group is treated in the same way as an
916 ordinary group without any | characters when processing (*THEN). PCRE has
917 been changed to match Perl's behaviour.
918
919 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
920 RunGrepTest script failed.
921
922 18. Change 22 for version 13 caused atomic groups to use more stack. This is
923 inevitable for groups that contain captures, but it can lead to a lot of
924 stack use in large patterns. The old behaviour has been restored for atomic
925 groups that do not contain any capturing parentheses.
926
927 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
928 suppress the check for a minimum subject length at run time. (If it was
929 given to pcre_exec() or pcre_dfa_exec() it did work.)
930
931 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
932 fail to work when decoding hex characters in data strings in EBCDIC
933 environments.
934
935 21. It appears that in at least one Mac OS environment, the isxdigit() function
936 is implemented as a macro that evaluates to its argument more than once,
937 contravening the C 90 Standard (I haven't checked a later standard). There
938 was an instance in pcretest which caused it to go wrong when processing
939 \x{...} escapes in subject strings. The has been rewritten to avoid using
940 things like p++ in the argument of isxdigit().
941
942
943 Version 8.13 16-Aug-2011
944 ------------------------
945
946 1. The Unicode data tables have been updated to Unicode 6.0.0.
947
948 2. Two minor typos in pcre_internal.h have been fixed.
949
950 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
951 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
952 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
953
954 4. There were a number of related bugs in the code for matching backrefences
955 caselessly in UTF-8 mode when codes for the characters concerned were
956 different numbers of bytes. For example, U+023A and U+2C65 are an upper
957 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
958 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
959 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
960 2-byte code at the end of the subject (it thought there wasn't enough data
961 left).
962
963 5. Comprehensive information about what went wrong is now returned by
964 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
965 as the output vector has at least 2 elements. The offset of the start of
966 the failing character and a reason code are placed in the vector.
967
968 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
969 now returned is for the first byte of the failing character, instead of the
970 last byte inspected. This is an incompatible change, but I hope it is small
971 enough not to be a problem. It makes the returned offset consistent with
972 pcre_exec() and pcre_dfa_exec().
973
974 7. pcretest now gives a text phrase as well as the error number when
975 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
976 failure, the offset and reason code are output.
977
978 8. When \R was used with a maximizing quantifier it failed to skip backwards
979 over a \r\n pair if the subsequent match failed. Instead, it just skipped
980 back over a single character (\n). This seems wrong (because it treated the
981 two characters as a single entity when going forwards), conflicts with the
982 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
983 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
984 has been changed.
985
986 9. Some internal refactoring has changed the processing so that the handling
987 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
988 time (the PCRE_DOTALL option was changed this way some time ago: version
989 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
990 which was always a bit of a fudge. It also means that there is one less
991 argument for the match() function, which reduces its stack requirements
992 slightly. This change also fixes an incompatibility with Perl: the pattern
993 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
994
995 10. More internal refactoring has drastically reduced the number of recursive
996 calls to match() for possessively repeated groups such as (abc)++ when
997 using pcre_exec().
998
999 11. While implementing 10, a number of bugs in the handling of groups were
1000 discovered and fixed:
1001
1002 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1003 (a|)*(?1) gave a compile-time internal error.
1004 ((a|)+)+ did not notice that the outer group could match an empty string.
1005 (^a|^)+ was not marked as anchored.
1006 (.*a|.*)+ was not marked as matching at start or after a newline.
1007
1008 12. Yet more internal refactoring has removed another argument from the match()
1009 function. Special calls to this function are now indicated by setting a
1010 value in a variable in the "match data" data block.
1011
1012 13. Be more explicit in pcre_study() instead of relying on "default" for
1013 opcodes that mean there is no starting character; this means that when new
1014 ones are added and accidentally left out of pcre_study(), testing should
1015 pick them up.
1016
1017 14. The -s option of pcretest has been documented for ages as being an old
1018 synonym of -m (show memory usage). I have changed it to mean "force study
1019 for every regex", that is, assume /S for every regex. This is similar to -i
1020 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1021 using it. It makes it easier to run collections of tests with and without
1022 study enabled, and thereby test pcre_study() more easily. All the standard
1023 tests are now run with and without -s (but some patterns can be marked as
1024 "never study" - see 20 below).
1025
1026 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1027 restoration of the capturing data to the outer values was not happening
1028 correctly.
1029
1030 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1031 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1032 pattern had matched an empty string, and so incorrectly returned a no
1033 match.
1034
1035 17. There was optimizing code for the last branch of non-capturing parentheses,
1036 and also for the obeyed branch of a conditional subexpression, which used
1037 tail recursion to cut down on stack usage. Unfortunately, now that there is
1038 the possibility of (*THEN) occurring in these branches, tail recursion is
1039 no longer possible because the return has to be checked for (*THEN). These
1040 two optimizations have therefore been removed. [But see 8.20/11 above.]
1041
1042 18. If a pattern containing \R was studied, it was assumed that \R always
1043 matched two bytes, thus causing the minimum subject length to be
1044 incorrectly computed because \R can also match just one byte.
1045
1046 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1047 was incorrectly computed.
1048
1049 20. If /S is present twice on a test pattern in pcretest input, it now
1050 *disables* studying, thereby overriding the use of -s on the command line
1051 (see 14 above). This is necessary for one or two tests to keep the output
1052 identical in both cases.
1053
1054 21. When (*ACCEPT) was used in an assertion that matched an empty string and
1055 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1056
1057 22. When an atomic group that contained a capturing parenthesis was
1058 successfully matched, but the branch in which it appeared failed, the
1059 capturing was not being forgotten if a higher numbered group was later
1060 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1061 group 1 to "a", when in fact it should be unset. This applied to multi-
1062 branched capturing and non-capturing groups, repeated or not, and also to
1063 positive assertions (capturing in negative assertions does not happen
1064 in PCRE) and also to nested atomic groups.
1065
1066 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1067 subject after a captured substring, to make it easier to tell which of a
1068 number of identical substrings has been captured.
1069
1070 24. The way atomic groups are processed by pcre_exec() has been changed so that
1071 if they are repeated, backtracking one repetition now resets captured
1072 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1073 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1074 "aaa". Previously, it would have been "a". As part of this code
1075 refactoring, the way recursive calls are handled has also been changed.
1076
1077 25. If an assertion condition captured any substrings, they were not passed
1078 back unless some other capturing happened later. For example, if
1079 (?(?=(a))a) was matched against "a", no capturing was returned.
1080
1081 26. When studying a pattern that contained subroutine calls or assertions,
1082 the code for finding the minimum length of a possible match was handling
1083 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1084 group 1 called group 2 while simultaneously a separate group 2 called group
1085 1). A stack overflow occurred in this case. I have fixed this by limiting
1086 the recursion depth to 10.
1087
1088 27. Updated RunTest.bat in the distribution to the version supplied by Tom
1089 Fortmann. This supports explicit test numbers on the command line, and has
1090 argument validation and error reporting.
1091
1092 28. An instance of \X with an unlimited repeat could fail if at any point the
1093 first character it looked at was a mark character.
1094
1095 29. Some minor code refactoring concerning Unicode properties and scripts
1096 should reduce the stack requirement of match() slightly.
1097
1098 30. Added the '=' option to pcretest to check the setting of unused capturing
1099 slots at the end of the pattern, which are documented as being -1, but are
1100 not included in the return count.
1101
1102 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1103 compiled something random. Now it gives a compile-time error (as does
1104 Perl).
1105
1106 32. A *MARK encountered during the processing of a positive assertion is now
1107 recorded and passed back (compatible with Perl).
1108
1109 33. If --only-matching or --colour was set on a pcregrep call whose pattern
1110 had alternative anchored branches, the search for a second match in a line
1111 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1112 matched the line "0102" twice. The same bug affected patterns that started
1113 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1114 twice.
1115
1116 34. Previously, PCRE did not allow quantification of assertions. However, Perl
1117 does, and because of capturing effects, quantifying parenthesized
1118 assertions may at times be useful. Quantifiers are now allowed for
1119 parenthesized assertions.
1120
1121 35. A minor code tidy in pcre_compile() when checking options for \R usage.
1122
1123 36. \g was being checked for fancy things in a character class, when it should
1124 just be a literal "g".
1125
1126 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1127 appearance of a nested POSIX class supersedes an apparent external class.
1128 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1129 unescaped square brackets may also appear as part of class names. For
1130 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1131 more like Perl. (But see 8.20/1 above.)
1132
1133 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1134 was because it thought it was \N{name}, which is not supported).
1135
1136 39. Add minix to OS list not supporting the -S option in pcretest.
1137
1138 40. PCRE tries to detect cases of infinite recursion at compile time, but it
1139 cannot analyze patterns in sufficient detail to catch mutual recursions
1140 such as ((?1))((?2)). There is now a runtime test that gives an error if a
1141 subgroup is called recursively as a subpattern for a second time at the
1142 same position in the subject string. In previous releases this might have
1143 been caught by the recursion limit, or it might have run out of stack.
1144
1145 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1146 happen only once. PCRE was, however incorrectly giving a compile time error
1147 "recursive call could loop indefinitely" because it cannot analyze the
1148 pattern in sufficient detail. The compile time test no longer happens when
1149 PCRE is compiling a conditional subpattern, but actual runaway loops are
1150 now caught at runtime (see 40 above).
1151
1152 42. It seems that Perl allows any characters other than a closing parenthesis
1153 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1154 has been changed to be the same.
1155
1156 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1157 as not to get warnings when autogen.sh is called. Also changed
1158 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1159
1160 44. To help people who use pcregrep to scan files containing exceedingly long
1161 lines, the following changes have been made:
1162
1163 (a) The default value of the buffer size parameter has been increased from
1164 8K to 20K. (The actual buffer used is three times this size.)
1165
1166 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1167 PCRE is built.
1168
1169 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1170 to be set at run time.
1171
1172 (d) Numerical values in pcregrep options can be followed by K or M, for
1173 example --buffer-size=50K.
1174
1175 (e) If a line being scanned overflows pcregrep's buffer, an error is now
1176 given and the return code is set to 2.
1177
1178 45. Add a pointer to the latest mark to the callout data block.
1179
1180 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1181 partial match of an empty string instead of no match. This was specific to
1182 the use of ".".
1183
1184 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1185 complete match instead of a partial match. This bug was dependent on both
1186 the PCRE_UTF8 and PCRE_DOTALL options being set.
1187
1188 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1189 starting byte set, because \b was not being ignored.
1190
1191
1192 Version 8.12 15-Jan-2011
1193 ------------------------
1194
1195 1. Fixed some typos in the markup of the man pages, and wrote a script that
1196 checks for such things as part of the documentation building process.
1197
1198 2. On a big-endian 64-bit system, pcregrep did not correctly process the
1199 --match-limit and --recursion-limit options (added for 8.11). In
1200 particular, this made one of the standard tests fail. (The integer value
1201 went into the wrong half of a long int.)
1202
1203 3. If the --colour option was given to pcregrep with -v (invert match), it
1204 did strange things, either producing crazy output, or crashing. It should,
1205 of course, ignore a request for colour when reporting lines that do not
1206 match.
1207
1208 4. Another pcregrep bug caused similar problems if --colour was specified with
1209 -M (multiline) and the pattern match finished with a line ending.
1210
1211 5. In pcregrep, when a pattern that ended with a literal newline sequence was
1212 matched in multiline mode, the following line was shown as part of the
1213 match. This seems wrong, so I have changed it.
1214
1215 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
1216 the check for further matches in the same line (so they could be coloured)
1217 to overrun the end of the current line. If another match was found, it was
1218 incorrectly shown (and then shown again when found in the next line).
1219
1220 7. If pcregrep was compiled under Windows, there was a reference to the
1221 function pcregrep_exit() before it was defined. I am assuming this was
1222 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1223 reported by a user. I've moved the definition above the reference.
1224
1225
1226 Version 8.11 10-Dec-2010
1227 ------------------------
1228
1229 1. (*THEN) was not working properly if there were untried alternatives prior
1230 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1231 backtracked to try for "b" instead of moving to the next alternative branch
1232 at the same level (in this case, to look for "c"). The Perl documentation
1233 is clear that when (*THEN) is backtracked onto, it goes to the "next
1234 alternative in the innermost enclosing group".
1235
1236 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1237 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
1238 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1239 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1240 (*THEN).
1241
1242 3. If \s appeared in a character class, it removed the VT character from
1243 the class, even if it had been included by some previous item, for example
1244 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1245 of \s, but is part of the POSIX "space" class.)
1246
1247 4. A partial match never returns an empty string (because you can always
1248 match an empty string at the end of the subject); however the checking for
1249 an empty string was starting at the "start of match" point. This has been
1250 changed to the "earliest inspected character" point, because the returned
1251 data for a partial match starts at this character. This means that, for
1252 example, /(?<=abc)def/ gives a partial match for the subject "abc"
1253 (previously it gave "no match").
1254
1255 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1256 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1257 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1258 has an implication that the given string is incomplete (because a partial
1259 match is preferred over a full match). For this reason, these items now
1260 give a partial match in this situation. [Aside: previously, the one case
1261 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1262 match rather than a full match, which was wrong by the old rules, but is
1263 now correct.]
1264
1265 6. There was a bug in the handling of #-introduced comments, recognized when
1266 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1267 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1268 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1269 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1270 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1271 places in pcre_compile().
1272
1273 7. Related to (6) above, when pcre_compile() was skipping #-introduced
1274 comments when looking ahead for named forward references to subpatterns,
1275 the only newline sequence it recognized was NL. It now handles newlines
1276 according to the set newline convention.
1277
1278 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1279 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1280 cater for a lack of strerror(). These oversights have been fixed.
1281
1282 9. Added --match-limit and --recursion-limit to pcregrep.
1283
1284 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1285
1286 11. When the -o option was used, pcregrep was setting a return code of 1, even
1287 when matches were found, and --line-buffered was not being honoured.
1288
1289 12. Added an optional parentheses number to the -o and --only-matching options
1290 of pcregrep.
1291
1292 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1293 can match an empty string. The code to do it in pcretest and pcredemo
1294 needed fixing:
1295
1296 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1297 only one byte after an empty string match just before CRLF (this case
1298 just got forgotten; "any" and "anycrlf" were OK).
1299
1300 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1301 mode when an empty string match preceded an ASCII character followed by
1302 a non-ASCII character. (The code for advancing by one character rather
1303 than one byte was nonsense.)
1304
1305 (c) The pcredemo.c sample program did not have any code at all to handle
1306 the cases when CRLF is a valid newline sequence.
1307
1308 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1309 as a starting offset was within the subject string. There is now a new
1310 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1311 negative or greater than the length of the string. In order to test this,
1312 pcretest is extended to allow the setting of negative starting offsets.
1313
1314 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1315 starting offset points to the beginning of a UTF-8 character was
1316 unnecessarily clumsy. I tidied it up.
1317
1318 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1319 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1320
1321 17. Nobody had reported that the --include_dir option, which was added in
1322 release 7.7 should have been called --include-dir (hyphen, not underscore)
1323 for compatibility with GNU grep. I have changed it to --include-dir, but
1324 left --include_dir as an undocumented synonym, and the same for
1325 --exclude-dir, though that is not available in GNU grep, at least as of
1326 release 2.5.4.
1327
1328 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1329 characters from a string of bytes) have been redefined so as not to use
1330 loops, in order to improve performance in some environments. At the same
1331 time, I abstracted some of the common code into auxiliary macros to save
1332 repetition (this should not affect the compiled code).
1333
1334 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1335 compile-time error is now given if \c is not followed by an ASCII
1336 character, that is, a byte less than 128. (In EBCDIC mode, the code is
1337 different, and any byte value is allowed.)
1338
1339 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1340 START_OPTIMIZE option, which is now allowed at compile time - but just
1341 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1342 to pcregrep and other applications that have no direct access to PCRE
1343 options. The new /Y option in pcretest sets this option when calling
1344 pcre_compile().
1345
1346 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1347 back references. Groups containing recursive back references were forced to
1348 be atomic by that change, but in the case of named groups, the amount of
1349 memory required was incorrectly computed, leading to "Failed: internal
1350 error: code overflow". This has been fixed.
1351
1352 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1353 pcretest.c, to avoid build problems in some Borland environments.
1354
1355
1356 Version 8.10 25-Jun-2010
1357 ------------------------
1358
1359 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1360 THEN.
1361
1362 2. (*ACCEPT) was not working when inside an atomic group.
1363
1364 3. Inside a character class, \B is treated as a literal by default, but
1365 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1366 causes the error). The code is unchanged, but I tidied the documentation.
1367
1368 4. Inside a character class, PCRE always treated \R and \X as literals,
1369 whereas Perl faults them if its -w option is set. I have changed PCRE so
1370 that it faults them when PCRE_EXTRA is set.
1371
1372 5. Added support for \N, which always matches any character other than
1373 newline. (It is the same as "." when PCRE_DOTALL is not set.)
1374
1375 6. When compiling pcregrep with newer versions of gcc which may have
1376 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1377 declared with attribute warn_unused_result" were given. Just casting the
1378 result to (void) does not stop the warnings; a more elaborate fudge is
1379 needed. I've used a macro to implement this.
1380
1381 7. Minor change to pcretest.c to avoid a compiler warning.
1382
1383 8. Added four artifical Unicode properties to help with an option to make
1384 \s etc use properties (see next item). The new properties are: Xan
1385 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1386
1387 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1388 use Unicode properties. (*UCP) at the start of a pattern can be used to set
1389 this option. Modified pcretest to add /W to test this facility. Added
1390 REG_UCP to make it available via the POSIX interface.
1391
1392 10. Added --line-buffered to pcregrep.
1393
1394 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1395 studied, and the match started with a letter with a code point greater than
1396 127 whose first byte was different to the first byte of the other case of
1397 the letter, the other case of this starting letter was not recognized
1398 (#976).
1399
1400 12. If a pattern that was studied started with a repeated Unicode property
1401 test, for example, \p{Nd}+, there was the theoretical possibility of
1402 setting up an incorrect bitmap of starting bytes, but fortunately it could
1403 not have actually happened in practice until change 8 above was made (it
1404 added property types that matched character-matching opcodes).
1405
1406 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1407 possible starting bytes for non-anchored patterns.
1408
1409 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1410 \R, and also a number of cases that involve Unicode properties, both
1411 explicit and implicit when PCRE_UCP is set.
1412
1413 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1414 input, it could crash or give wrong results if characters with values
1415 greater than 0xc0 were present in the subject string. (Detail: it assumed
1416 UTF-8 input when processing these items.)
1417
1418 16. Added a lot of (int) casts to avoid compiler warnings in systems where
1419 size_t is 64-bit (#991).
1420
1421 17. Added a check for running out of memory when PCRE is compiled with
1422 --disable-stack-for-recursion (#990).
1423
1424 18. If the last data line in a file for pcretest does not have a newline on
1425 the end, a newline was missing in the output.
1426
1427 19. The default pcre_chartables.c file recognizes only ASCII characters (values
1428 less than 128) in its various bitmaps. However, there is a facility for
1429 generating tables according to the current locale when PCRE is compiled. It
1430 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1431 characters, are recognized by isspace() and therefore were getting set in
1432 these tables, and indeed these tables seem to approximate to ISO 8859. This
1433 caused a problem in UTF-8 mode when pcre_study() was used to create a list
1434 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1435 which of course cannot start UTF-8 characters. I have changed the code so
1436 that only real ASCII characters (less than 128) and the correct starting
1437 bytes for UTF-8 encodings are set for characters greater than 127 when in
1438 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1439 altogether.)
1440
1441 20. Added the /T option to pcretest so as to be able to run tests with non-
1442 standard character tables, thus making it possible to include the tests
1443 used for 19 above in the standard set of tests.
1444
1445 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1446 reference to a subpattern the other side of a comment that contains an
1447 opening parenthesis caused either an internal compiling error, or a
1448 reference to the wrong subpattern.
1449
1450
1451 Version 8.02 19-Mar-2010
1452 ------------------------
1453
1454 1. The Unicode data tables have been updated to Unicode 5.2.0.
1455
1456 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
1457 configured.
1458
1459 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
1460 original author of that file, following a query about its status.
1461
1462 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
1463 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1464
1465 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1466 quantifier applied to a forward-referencing subroutine call, could compile
1467 incorrect code or give the error "internal error: previously-checked
1468 referenced subpattern not found".
1469
1470 6. Both MS Visual Studio and Symbian OS have problems with initializing
1471 variables to point to external functions. For these systems, therefore,
1472 pcre_malloc etc. are now initialized to local functions that call the
1473 relevant global functions.
1474
1475 7. There were two entries missing in the vectors called coptable and poptable
1476 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1477 I've fixed the data, and added a kludgy way of testing at compile time that
1478 the lengths are correct (equal to the number of opcodes).
1479
1480 8. Following on from 7, I added a similar kludge to check the length of the
1481 eint vector in pcreposix.c.
1482
1483 9. Error texts for pcre_compile() are held as one long string to avoid too
1484 much relocation at load time. To find a text, the string is searched,
1485 counting zeros. There was no check for running off the end of the string,
1486 which could happen if a new error number was added without updating the
1487 string.
1488
1489 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1490
1491 11. \K was not working if it appeared in an atomic group or in a group that
1492 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1493 \K is "not well defined" if used in an assertion. PCRE now accepts it if
1494 the assertion is positive, but not if it is negative.
1495
1496 12. Change 11 fortuitously reduced the size of the stack frame used in the
1497 "match()" function of pcre_exec.c by one pointer. Forthcoming
1498 implementation of support for (*MARK) will need an extra pointer on the
1499 stack; I have reserved it now, so that the stack frame size does not
1500 decrease.
1501
1502 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1503 item in branch that calls a recursion is a subroutine call - as in the
1504 second branch in the above example - was incorrectly given the compile-
1505 time error "recursive call could loop indefinitely" because pcre_compile()
1506 was not correctly checking the subroutine for matching a non-empty string.
1507
1508 14. The checks for overrunning compiling workspace could trigger after an
1509 overrun had occurred. This is a "should never occur" error, but it can be
1510 triggered by pathological patterns such as hundreds of nested parentheses.
1511 The checks now trigger 100 bytes before the end of the workspace.
1512
1513 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1514
1515
1516 Version 8.01 19-Jan-2010
1517 ------------------------
1518
1519 1. If a pattern contained a conditional subpattern with only one branch (in
1520 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1521 computed the wrong minimum data length (which is of course zero for such
1522 subpatterns). This could cause incorrect "no match" results.
1523
1524 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1525 the pattern is reset in the first branch, pcre_compile() failed with
1526 "internal error: code overflow at offset...". This happened only when
1527 the reset was to the original external option setting. (An optimization
1528 abstracts leading options settings into an external setting, which was the
1529 cause of this.)
1530
1531 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1532 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1533 assertion pattern did not match (meaning that the assertion was true), it
1534 was incorrectly treated as false if the SKIP had been reached during the
1535 matching. This also applied to assertions used as conditions.
1536
1537 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
1538 assertion subpattern, including such a pattern used as a condition,
1539 unpredictable results occurred, instead of the error return
1540 PCRE_ERROR_DFA_UITEM.
1541
1542 5. The C++ GlobalReplace function was not working like Perl for the special
1543 situation when an empty string is matched. It now does the fancy magic
1544 stuff that is necessary.
1545
1546 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1547 removed. (These were left over from very, very early versions of PCRE.)
1548
1549 7. Some cosmetic changes to the code to make life easier when compiling it
1550 as part of something else:
1551
1552 (a) Change DEBUG to PCRE_DEBUG.
1553
1554 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1555 called "current" as "current_branch", to prevent a collision with the
1556 Linux macro when compiled as a kernel module.
1557
1558 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1559 prevent a collision with the Linux macro when compiled as a kernel
1560 module.
1561
1562 8. In pcre_compile() there are some checks for integer overflows that used to
1563 cast potentially large values to (double). This has been changed to that
1564 when building, a check for int64_t is made, and if it is found, it is used
1565 instead, thus avoiding the use of floating point arithmetic. (There is no
1566 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1567 double.
1568
1569 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1570 2005 (difference between two addresses compared to an unsigned value).
1571
1572 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1573 custom one, because of the following reported problem in Windows:
1574
1575 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1576 under Win32.
1577 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1578 therefore missing the function definition.
1579 - The compiler thus generates a "C" signature for the test function.
1580 - The linker fails to find the "C" function.
1581 - PCRE fails to configure if asked to do so against libbz2.
1582
1583 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1584 messages were output:
1585
1586 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1587 rerunning libtoolize, to keep the correct libtool macros in-tree.
1588 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1589
1590 I have done both of these things.
1591
1592 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1593 most of the time, it *can* run out if it is given a pattern that contains a
1594 runaway infinite recursion. I updated the discussion in the pcrestack man
1595 page.
1596
1597 13. Now that we have gone to the x.xx style of version numbers, the minor
1598 version may start with zero. Using 08 or 09 is a bad idea because users
1599 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1600 interpreted as invalid octal numbers. I've updated the previous comment in
1601 configure.ac, and also added a check that gives an error if 08 or 09 are
1602 used.
1603
1604 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1605 causing partial matching to fail when the end of the subject matched \W
1606 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1607
1608 15. There were some discrepancies between the declarations in pcre_internal.h
1609 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1610 their definitions. The declarations used "const uschar *" and the
1611 definitions used USPTR. Even though USPTR is normally defined as "const
1612 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1613 reported that: "This difference in casting confuses some C++ compilers, for
1614 example, SunCC recognizes above declarations as different functions and
1615 generates broken code for hbpcre." I have changed the declarations to use
1616 USPTR.
1617
1618 16. GNU libtool is named differently on some systems. The autogen.sh script now
1619 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1620 (FreeBSD).
1621
1622 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1623 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1624 comment: "Figure out how to create a longlong from a string: strtoll and
1625 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1626 instance, but it only takes 2 args instead of 3!"
1627
1628 18. A subtle bug concerned with back references has been fixed by a change of
1629 specification, with a corresponding code fix. A pattern such as
1630 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1631 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1632 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1633 same bug. Such groups have to be quantified to be useful, or contained
1634 inside another quantified group. (If there's no repetition, the reference
1635 can never match.) The problem arises because, having left the group and
1636 moved on to the rest of the pattern, a later failure that backtracks into
1637 the group uses the captured value from the final iteration of the group
1638 rather than the correct earlier one. I have fixed this in PCRE by forcing
1639 any group that contains a reference to itself to be an atomic group; that
1640 is, there cannot be any backtracking into it once it has completed. This is
1641 similar to recursive and subroutine calls.
1642
1643
1644 Version 8.00 19-Oct-09
1645 ----------------------
1646
1647 1. The table for translating pcre_compile() error codes into POSIX error codes
1648 was out-of-date, and there was no check on the pcre_compile() error code
1649 being within the table. This could lead to an OK return being given in
1650 error.
1651
1652 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1653 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1654 in a Windows environment.
1655
1656 3. The pcregrep --count option prints the count for each file even when it is
1657 zero, as does GNU grep. However, pcregrep was also printing all files when
1658 --files-with-matches was added. Now, when both options are given, it prints
1659 counts only for those files that have at least one match. (GNU grep just
1660 prints the file name in this circumstance, but including the count seems
1661 more useful - otherwise, why use --count?) Also ensured that the
1662 combination -clh just lists non-zero counts, with no names.
1663
1664 4. The long form of the pcregrep -F option was incorrectly implemented as
1665 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1666 but it seems right to fix it, and I didn't think it was worth preserving
1667 the old behaviour.
1668
1669 5. The command line items --regex=pattern and --regexp=pattern were not
1670 recognized by pcregrep, which required --regex pattern or --regexp pattern
1671 (with a space rather than an '='). The man page documented the '=' forms,
1672 which are compatible with GNU grep; these now work.
1673
1674 6. No libpcreposix.pc file was created for pkg-config; there was just
1675 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1676
1677 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1678 when UCP support is not needed, by modifying the Python script that
1679 generates it from Unicode data files. This should not matter if the module
1680 is correctly used as a library, but I received one complaint about 50K of
1681 unwanted data. My guess is that the person linked everything into his
1682 program rather than using a library. Anyway, it does no harm.
1683
1684 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1685 was a minimum greater than 1 for a wide character in a possessive
1686 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1687 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1688 character. Chaos in the form of incorrect output or a compiling loop could
1689 result.
1690
1691 9. The restrictions on what a pattern can contain when partial matching is
1692 requested for pcre_exec() have been removed. All patterns can now be
1693 partially matched by this function. In addition, if there are at least two
1694 slots in the offset vector, the offset of the earliest inspected character
1695 for the match and the offset of the end of the subject are set in them when
1696 PCRE_ERROR_PARTIAL is returned.
1697
1698 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1699 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1700 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1701 and may be more useful for multi-segment matching.
1702
1703 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1704 used to be given if ever the end of the subject was reached; now it is
1705 given only if matching could not proceed because another character was
1706 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1707 string "Z", which now yields "no match" instead of "partial match". In the
1708 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1709 final character ended with (*FAIL).
1710
1711 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1712 if the pattern had a "must contain" character that was already found in the
1713 earlier partial match, unless partial matching was again requested. For
1714 example, with the pattern /dog.(body)?/, the "must contain" character is
1715 "g". If the first part-match was for the string "dog", restarting with
1716 "sbody" failed. This bug has been fixed.
1717
1718 13. The string returned by pcre_dfa_exec() after a partial match has been
1719 changed so that it starts at the first inspected character rather than the
1720 first character of the match. This makes a difference only if the pattern
1721 starts with a lookbehind assertion or \b or \B (\K is not supported by
1722 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1723 matching functions compatible, and I think it's the right thing to do.
1724
1725 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1726 so that the demonstration program is easily available in environments where
1727 PCRE has not been installed from source.
1728
1729 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1730 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1731 library.
1732
1733 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1734 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1735 is not the first non-POSIX option to be added. Clearly some people find
1736 these options useful.
1737
1738 17. If a caller to the POSIX matching function regexec() passes a non-zero
1739 value for nmatch with a NULL value for pmatch, the value of
1740 nmatch is forced to zero.
1741
1742 18. RunGrepTest did not have a test for the availability of the -u option of
1743 the diff command, as RunTest does. It now checks in the same way as
1744 RunTest, and also checks for the -b option.
1745
1746 19. If an odd number of negated classes containing just a single character
1747 interposed, within parentheses, between a forward reference to a named
1748 subpattern and the definition of the subpattern, compilation crashed with
1749 an internal error, complaining that it could not find the referenced
1750 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1751 [The bug was that it was starting one character too far in when skipping
1752 over the character class, thus treating the ] as data rather than
1753 terminating the class. This meant it could skip too much.]
1754
1755 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1756 /g option in pcretest when the pattern contains \K, which makes it possible
1757 to have an empty string match not at the start, even when the pattern is
1758 anchored. Updated pcretest and pcredemo to use this option.
1759
1760 21. If the maximum number of capturing subpatterns in a recursion was greater
1761 than the maximum at the outer level, the higher number was returned, but
1762 with unset values at the outer level. The correct (outer level) value is
1763 now given.
1764
1765 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1766 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1767 make it do so. The string so far is captured, making this feature
1768 compatible with Perl.
1769
1770 23. The tests have been re-organized, adding tests 11 and 12, to make it
1771 possible to check the Perl 5.10 features against Perl 5.10.
1772
1773 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1774 pattern matches a fixed length string. PCRE did not allow this; now it
1775 does. Neither allows recursion.
1776
1777 25. I finally figured out how to implement a request to provide the minimum
1778 length of subject string that was needed in order to match a given pattern.
1779 (It was back references and recursion that I had previously got hung up
1780 on.) This code has now been added to pcre_study(); it finds a lower bound
1781 to the length of subject needed. It is not necessarily the greatest lower
1782 bound, but using it to avoid searching strings that are too short does give
1783 some useful speed-ups. The value is available to calling programs via
1784 pcre_fullinfo().
1785
1786 26. While implementing 25, I discovered to my embarrassment that pcretest had
1787 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1788 study optimizations had never been tested with that matching function.
1789 Oops. What is worse, even when it was passed study data, there was a bug in
1790 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1791 were also very few tests of studied patterns with pcre_dfa_exec().
1792
1793 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1794 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1795 on the other side of the coin, they are no longer allowed to have different
1796 names, because these cannot be distinguished in PCRE, and this has caused
1797 confusion. (This is a difference from Perl.)
1798
1799 28. When duplicate subpattern names are present (necessarily with different
1800 numbers, as required by 27 above), and a test is made by name in a
1801 conditional pattern, either for a subpattern having been matched, or for
1802 recursion in such a pattern, all the associated numbered subpatterns are
1803 tested, and the overall condition is true if the condition is true for any
1804 one of them. This is the way Perl works, and is also more like the way
1805 testing by number works.
1806
1807
1808 Version 7.9 11-Apr-09
1809 ---------------------
1810
1811 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1812 (pcretest), all targets were linked against these libraries. This included
1813 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1814 libraries. This caused unwanted dependencies to be created. This problem
1815 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1816 pcretest is linked with readline.
1817
1818 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1819 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1820 moved outside it again, because FALSE and TRUE are already defined in AIX,
1821 but BOOL is not.
1822
1823 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1824 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1825
1826 4. The pcregrep documentation said spaces were inserted as well as colons (or
1827 hyphens) following file names and line numbers when outputting matching
1828 lines. This is not true; no spaces are inserted. I have also clarified the
1829 wording for the --colour (or --color) option.
1830
1831 5. In pcregrep, when --colour was used with -o, the list of matching strings
1832 was not coloured; this is different to GNU grep, so I have changed it to be
1833 the same.
1834
1835 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1836 each matching line was coloured. Now it goes on to look for further matches
1837 of any of the test patterns, which is the same behaviour as GNU grep.
1838
1839 7. A pattern that could match an empty string could cause pcregrep to loop; it
1840 doesn't make sense to accept an empty string match in pcregrep, so I have
1841 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1842 seems to be how GNU grep behaves. [But see later change 40 for release
1843 8.33.]
1844
1845 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1846 start or after a newline", because the conditional assertion was not being
1847 correctly handled. The rule now is that both the assertion and what follows
1848 in the first alternative must satisfy the test.
1849
1850 9. If auto-callout was enabled in a pattern with a conditional group whose
1851 condition was an assertion, PCRE could crash during matching, both with
1852 pcre_exec() and pcre_dfa_exec().
1853
1854 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1855 used for matching.
1856
1857 11. Unicode property support in character classes was not working for
1858 characters (bytes) greater than 127 when not in UTF-8 mode.
1859
1860 12. Added the -M command line option to pcretest.
1861
1862 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1863
1864 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1865
1866 16. Added comments and documentation about mis-use of no_arg in the C++
1867 wrapper.
1868
1869 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1870 from Martin Jerabek that uses macro names for all relevant character and
1871 string constants.
1872
1873 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1874 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1875 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1876 these, but not everybody uses configure.
1877
1878 19. A conditional group that had only one branch was not being correctly
1879 recognized as an item that could match an empty string. This meant that an
1880 enclosing group might also not be so recognized, causing infinite looping
1881 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1882 with the subject "ab", where knowledge that the repeated group can match
1883 nothing is needed in order to break the loop.
1884
1885 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1886 exec(), but without supplying a callout function, matching went wrong.
1887
1888 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1889 leak if the size of the offset vector was greater than 30. When the vector
1890 is smaller, the saved offsets during recursion go onto a local stack
1891 vector, but for larger vectors malloc() is used. It was failing to free
1892 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1893 error, in fact).
1894
1895 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1896 heapframe that is used only when UTF-8 support is enabled. This caused no
1897 problem, but was untidy.
1898
1899 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1900 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1901 included within another project.
1902
1903 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1904 slightly modified by me:
1905
1906 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1907 not building pcregrep.
1908
1909 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1910 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1911
1912 25. Forward references, both numeric and by name, in patterns that made use of
1913 duplicate group numbers, could behave incorrectly or give incorrect errors,
1914 because when scanning forward to find the reference group, PCRE was not
1915 taking into account the duplicate group numbers. A pattern such as
1916 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1917
1918 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1919 the feature of a custom pointer more persuasive (as requested by a user).
1920
1921 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1922 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1923 pre-defined.
1924
1925 28. Added support for (*UTF8) at the start of a pattern.
1926
1927 29. Arrange for flags added by the "release type" setting in CMake to be shown
1928 in the configuration summary.
1929
1930
1931 Version 7.8 05-Sep-08
1932 ---------------------
1933
1934 1. Replaced UCP searching code with optimized version as implemented for Ad
1935 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1936 stage table and inline lookup instead of a function, giving speed ups of 2
1937 to 5 times on some simple patterns that I tested. Permission was given to
1938 distribute the MultiStage2.py script that generates the tables (it's not in
1939 the tarball, but is in the Subversion repository).
1940
1941 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1942 scripts.
1943
1944 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1945 a group with a zero qualifier. The result of the study could be incorrect,
1946 or the function might crash, depending on the pattern.
1947
1948 4. Caseless matching was not working for non-ASCII characters in back
1949 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1950 It now works when Unicode Property Support is available.
1951
1952 5. In pcretest, an escape such as \x{de} in the data was always generating
1953 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1954 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1955 truncation.
1956
1957 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1958
1959 7. Added two (int) casts to pcregrep when printing the difference of two
1960 pointers, in case they are 64-bit values.
1961
1962 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1963 test 2 if it fails.
1964
1965 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1966 and a #define of that name to empty if it is not externally set. This is to
1967 allow users of MSVC to set it if necessary.
1968
1969 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1970 the convenience functions in the pcre_get.c source file.
1971
1972 11. An option change at the start of a pattern that had top-level alternatives
1973 could cause overwriting and/or a crash. This command provoked a crash in
1974 some environments:
1975
1976 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1977
1978 This potential security problem was recorded as CVE-2008-2371.
1979
1980 12. For a pattern where the match had to start at the beginning or immediately
1981 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1982 pcre_dfa_exec() could read past the end of the passed subject if there was
1983 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1984 pcretest so that it places the subject at the end of its malloc-ed buffer.
1985
1986 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1987 exec() might read past the end of the data buffer in UTF-8 mode.
1988
1989 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1990 the data contained the byte 0x85 as part of a UTF-8 character within its
1991 first line. This applied both to normal and DFA matching.
1992
1993 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1994 /^[^d]*?$/8 failed to match "abc".
1995
1996 16. Added a missing copyright notice to pcrecpp_internal.h.
1997
1998 17. Make it more clear in the documentation that values returned from
1999 pcre_exec() in ovector are byte offsets, not character counts.
2000
2001 18. Tidied a few places to stop certain compilers from issuing warnings.
2002
2003 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2004 supplied by Stefan Weber. I made a further small update for 7.8 because
2005 there is a change of source arrangements: the pcre_searchfuncs.c module is
2006 replaced by pcre_ucd.c.
2007
2008
2009 Version 7.7 07-May-08
2010 ---------------------
2011
2012 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
2013 a string to a long long, pretend we don't even have a long long." This is
2014 done by checking for the strtoq, strtoll, and _strtoi64 functions.
2015
2016 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2017 pre-7.6 versions, which defined a global no_arg variable instead of putting
2018 it in the RE class. (See also #8 below.)
2019
2020 3. Remove a line of dead code, identified by coverity and reported by Nuno
2021 Lopes.
2022
2023 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
2024
2025 (1) The include/exclude patterns were being applied to the whole pathnames
2026 of files, instead of just to the final components.
2027
2028 (2) If there was more than one level of directory, the subdirectories were
2029 skipped unless they satisfied the include/exclude conditions. This is
2030 inconsistent with GNU grep (and could even be seen as contrary to the
2031 pcregrep specification - which I improved to make it absolutely clear).
2032 The action now is always to scan all levels of directory, and just
2033 apply the include/exclude patterns to regular files.
2034
2035 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2036 --exclude_dir in the tests to avoid scanning .svn directories.
2037
2038 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
2039 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2040 doesn't support NULs in patterns.
2041
2042 7. Added some missing "const"s to declarations of static tables in
2043 pcre_compile.c and pcre_dfa_exec.c.
2044
2045 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2046 caused by fix #2 above. (Subsequently also a second patch to fix the
2047 first patch. And a third patch - this was a messy problem.)
2048
2049 9. Applied Craig's patch to remove the use of push_back().
2050
2051 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2052 matching function regexec().
2053
2054 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2055 which, however, unlike Perl's \g{...}, are subroutine calls, not back
2056 references. PCRE supports relative numbers with this syntax (I don't think
2057 Oniguruma does).
2058
2059 12. Previously, a group with a zero repeat such as (...){0} was completely
2060 omitted from the compiled regex. However, this means that if the group
2061 was called as a subroutine from elsewhere in the pattern, things went wrong
2062 (an internal error was given). Such groups are now left in the compiled
2063 pattern, with a new opcode that causes them to be skipped at execution
2064 time.
2065
2066 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2067 to the way PCRE behaves:
2068
2069 (a) A lone ] character is dis-allowed (Perl treats it as data).
2070
2071 (b) A back reference to an unmatched subpattern matches an empty string
2072 (Perl fails the current match path).
2073
2074 (c) A data ] in a character class must be notated as \] because if the
2075 first data character in a class is ], it defines an empty class. (In
2076 Perl it is not possible to have an empty class.) The empty class []
2077 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2078 The negative empty class [^] matches any one character, independently
2079 of the DOTALL setting.
2080
2081 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2082 non-existent subpattern following a character class starting with ']' and
2083 containing () gave an internal compiling error instead of "reference to
2084 non-existent subpattern". Fortunately, when the pattern did exist, the
2085 compiled code was correct. (When scanning forwards to check for the
2086 existence of the subpattern, it was treating the data ']' as terminating
2087 the class, so got the count wrong. When actually compiling, the reference
2088 was subsequently set up correctly.)
2089
2090 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2091 it was being rejected as not supported by pcre_dfa_exec(), even though
2092 other assertions are supported. I have made pcre_dfa_exec() support
2093 (*FAIL).
2094
2095 16. The implementation of 13c above involved the invention of a new opcode,
2096 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2097 cannot be changed at match time, I realized I could make a small
2098 improvement to matching performance by compiling OP_ALLANY instead of
2099 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2100 on the OP_ANY path.
2101
2102 17. Compiling pcretest on Windows with readline support failed without the
2103 following two fixes: (1) Make the unistd.h include conditional on
2104 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2105
2106 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2107 ncurses library to be included for pcretest when ReadLine support is
2108 requested, but also to allow for it to be overridden. This patch came from
2109 Daniel Bergström.
2110
2111 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2112 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2113 any errors with the current Unicode tables. Thanks to Peter Kankowski for
2114 spotting this.
2115
2116
2117 Version 7.6 28-Jan-08
2118 ---------------------
2119
2120 1. A character class containing a very large number of characters with
2121 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2122 overflow.
2123
2124 2. Patch to cut out the "long long" test in pcrecpp_unittest when
2125 HAVE_LONG_LONG is not defined.
2126
2127 3. Applied Christian Ehrlicher's patch to update the CMake build files to
2128 bring them up to date and include new features. This patch includes:
2129
2130 - Fixed PH's badly added libz and libbz2 support.
2131 - Fixed a problem with static linking.
2132 - Added pcredemo. [But later removed - see 7 below.]
2133 - Fixed dftables problem and added an option.
2134 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2135 HAVE_LONG_LONG.
2136 - Added readline support for pcretest.
2137 - Added an listing of the option settings after cmake has run.
2138
2139 4. A user submitted a patch to Makefile that makes it easy to create
2140 "pcre.dll" under mingw when using Configure/Make. I added stuff to
2141 Makefile.am that cause it to include this special target, without
2142 affecting anything else. Note that the same mingw target plus all
2143 the other distribution libraries and programs are now supported
2144 when configuring with CMake (see 6 below) instead of with
2145 Configure/Make.
2146
2147 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2148 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2149 exported in the Windows port". It has not yet been confirmed that the patch
2150 solves the problem, but it does no harm.
2151
2152 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2153 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2154 with CMake, and also correct the comment about stack recursion.
2155
2156 7. Remove the automatic building of pcredemo from the ./configure system and
2157 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2158 of a program that users should build themselves after PCRE is installed, so
2159 building it automatically is not really right. What is more, it gave
2160 trouble in some build environments.
2161
2162 8. Further tidies to CMakeLists.txt from Sheri and Christian.
2163
2164
2165 Version 7.5 10-Jan-08
2166 ---------------------
2167
2168 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
2169 values in parens when parsing an RE using the C++ wrapper."
2170
2171 2. Negative specials like \S did not work in character classes in UTF-8 mode.
2172 Characters greater than 255 were excluded from the class instead of being
2173 included.
2174
2175 3. The same bug as (2) above applied to negated POSIX classes such as
2176 [:^space:].
2177
2178 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2179 defined or documented. It seems to have been a typo for PCRE_STATIC, so
2180 I have changed it.
2181
2182 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
2183 first named subpattern) and a construct such as (?&a) would reference the
2184 first named subpattern whose name started with "a" (in other words, the
2185 length check was missing). Both these problems are fixed. "Subpattern name
2186 expected" is now given for (?&) (a zero-length name), and this patch also
2187 makes it give the same error for \k'' (previously it complained that that
2188 was a reference to a non-existent subpattern).
2189
2190 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
2191 this is right because (?- can be followed by option settings as well as by
2192 digits. I have, however, made the messages clearer.
2193
2194 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2195 than the number used in the conditional) now cause a compile-time error.
2196 This is actually not compatible with Perl, which accepts such patterns, but
2197 treats the conditional as always being FALSE (as PCRE used to), but it
2198 seems to me that giving a diagnostic is better.
2199
2200 8. Change "alphameric" to the more common word "alphanumeric" in comments
2201 and messages.
2202
2203 9. Fix two occurrences of "backslash" in comments that should have been
2204 "backspace".
2205
2206 10. Remove two redundant lines of code that can never be obeyed (their function
2207 was moved elsewhere).
2208
2209 11. The program that makes PCRE's Unicode character property table had a bug
2210 which caused it to generate incorrect table entries for sequences of
2211 characters that have the same character type, but are in different scripts.
2212 It amalgamated them into a single range, with the script of the first of
2213 them. In other words, some characters were in the wrong script. There were
2214 thirteen such cases, affecting characters in the following ranges:
2215
2216 U+002b0 - U+002c1
2217 U+0060c - U+0060d
2218 U+0061e - U+00612
2219 U+0064b - U+0065e
2220 U+0074d - U+0076d
2221 U+01800 - U+01805
2222 U+01d00 - U+01d77
2223 U+01d9b - U+01dbf
2224 U+0200b - U+0200f
2225 U+030fc - U+030fe
2226 U+03260 - U+0327f
2227 U+0fb46 - U+0fbb1
2228 U+10450 - U+1049d
2229
2230 12. The -o option (show only the matching part of a line) for pcregrep was not
2231 compatible with GNU grep in that, if there was more than one match in a
2232 line, it showed only the first of them. It now behaves in the same way as
2233 GNU grep.
2234
2235 13. If the -o and -v options were combined for pcregrep, it printed a blank
2236 line for every non-matching line. GNU grep prints nothing, and pcregrep now
2237 does the same. The return code can be used to tell if there were any
2238 non-matching lines.
2239
2240 14. Added --file-offsets and --line-offsets to pcregrep.
2241
2242 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2243 infinitely looping recursion. The bug was that positive lookaheads were not
2244 being skipped when checking for a possible empty match (negative lookaheads
2245 and both kinds of lookbehind were skipped).
2246
2247 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2248 inclusion of <windows.h> to before rather than after the definition of
2249 INVALID_FILE_ATTRIBUTES (patch from David Byron).
2250
2251 17. Specifying a possessive quantifier with a specific limit for a Unicode
2252 character property caused pcre_compile() to compile bad code, which led at
2253 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2254 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2255 caused the error; without that there was no problem.
2256
2257 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2258
2259 19. Added --enable-pcretest-libreadline.
2260
2261 20. In pcrecpp.cc, the variable 'count' was incremented twice in
2262 RE::GlobalReplace(). As a result, the number of replacements returned was
2263 double what it should be. I removed one of the increments, but Craig sent a
2264 later patch that removed the other one (the right fix) and added unit tests
2265 that check the return values (which was not done before).
2266
2267 21. Several CMake things:
2268
2269 (1) Arranged that, when cmake is used on Unix, the libraries end up with
2270 the names libpcre and libpcreposix, not just pcre and pcreposix.
2271
2272 (2) The above change means that pcretest and pcregrep are now correctly
2273 linked with the newly-built libraries, not previously installed ones.
2274
2275 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2276
2277 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2278 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2279 UTF-8 newline character). The key issue is that the pattern starts .*;
2280 this means that the match must be either at the beginning, or after a
2281 newline. The bug was in the code for advancing after a failed match and
2282 checking that the new position followed a newline. It was not taking
2283 account of UTF-8 characters correctly.
2284
2285 23. PCRE was behaving differently from Perl in the way it recognized POSIX
2286 character classes. PCRE was not treating the sequence [:...:] as a
2287 character class unless the ... were all letters. Perl, however, seems to
2288 allow any characters between [: and :], though of course it rejects as
2289 unknown any "names" that contain non-letters, because all the known class
2290 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2291 for example, whereas PCRE did not - it did not recognize a POSIX character
2292 class. This seemed a bit dangerous, so the code has been changed to be
2293 closer to Perl. The behaviour is not identical to Perl, because PCRE will
2294 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2295 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2296 Perl does, and where it didn't before.
2297
2298 24. Rewrite so as to remove the single use of %n from pcregrep because in some
2299 Windows environments %n is disabled by default.
2300
2301
2302 Version 7.4 21-Sep-07
2303 ---------------------
2304
2305 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
2306 means that a class such as [\s] counted as "explicit reference to CR or
2307 LF". That isn't really right - the whole point of the change was to try to
2308 help when there was an actual mention of one of the two characters. So now
2309 the change happens only if \r or \n (or a literal CR or LF) character is
2310 encountered.
2311
2312 2. The 32-bit options word was also used for 6 internal flags, but the numbers
2313 of both had grown to the point where there were only 3 bits left.
2314 Fortunately, there was spare space in the data structure, and so I have
2315 moved the internal flags into a new 16-bit field to free up more option
2316 bits.
2317
2318 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2319 but did not set the internal JCHANGED flag - either of these is enough to
2320 control the way the "get" function works - but the PCRE_INFO_JCHANGED
2321 facility is supposed to tell if (?J) was ever used, so now (?J) at the
2322 start sets both bits.
2323
2324 4. Added options (at build time, compile time, exec time) to change \R from
2325 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2326
2327 5. doc/pcresyntax.html was missing from the distribution.
2328
2329 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2330 compatibility, even though it is no longer used.
2331
2332 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2333 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2334 windows.h file is present (where different names are used). [This was
2335 reversed later after testing - see 16 below.]
2336
2337 8. Changed all #include <config.h> to #include "config.h". There were also
2338 some further <pcre.h> cases that I changed to "pcre.h".
2339
2340 9. When pcregrep was used with the --colour option, it missed the line ending
2341 sequence off the lines that it output.
2342
2343 10. It was pointed out to me that arrays of string pointers cause lots of
2344 relocations when a shared library is dynamically loaded. A technique of
2345 using a single long string with a table of offsets can drastically reduce
2346 these. I have refactored PCRE in four places to do this. The result is
2347 dramatic:
2348
2349 Originally: 290
2350 After changing UCP table: 187
2351 After changing error message table: 43
2352 After changing table of "verbs" 36
2353 After changing table of Posix names 22
2354
2355 Thanks to the folks working on Gregex for glib for this insight.
2356
2357 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2358 unicode-properties was also set.
2359
2360 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2361
2362 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2363 checked only for CRLF.
2364
2365 14. Added casts to pcretest.c to avoid compiler warnings.
2366
2367 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2368
2369 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2370 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2371 entirely. This removes changes made in 7 above.
2372
2373 17. The CMake files have been updated, and there is now more information about
2374 building with CMake in the NON-UNIX-USE document.
2375
2376
2377 Version 7.3 28-Aug-07
2378 ---------------------
2379
2380 1. In the rejigging of the build system that eventually resulted in 7.1, the
2381 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2382 brackets there is not right, since it causes compilers to look for an
2383 installed pcre.h, not the version that is in the source that is being
2384 compiled (which of course may be different). I have changed it back to:
2385
2386 #include "pcre.h"
2387
2388 I have a vague recollection that the change was concerned with compiling in
2389 different directories, but in the new build system, that is taken care of
2390 by the VPATH setting the Makefile.
2391
2392 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2393 when the subject happened to end in the byte 0x85 (e.g. if the last
2394 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2395 characters but of course it shouldn't be taken as a newline when it is part
2396 of another character. The bug was that, for an unlimited repeat of . in
2397 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2398 characters when looking for a newline.
2399
2400 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2401
2402 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2403 in debug output.
2404
2405 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2406 long printing in the pcrecpp unittest when running under MinGW.
2407
2408 6. ESC_K was left out of the EBCDIC table.
2409
2410 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2411 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2412 limit also applies to "virtual nesting" when a pattern is recursive, and in
2413 this case 1000 isn't so big. I have been able to remove this limit at the
2414 expense of backing off one optimization in certain circumstances. Normally,
2415 when pcre_exec() would call its internal match() function recursively and
2416 immediately return the result unconditionally, it uses a "tail recursion"
2417 feature to save stack. However, when a subpattern that can match an empty
2418 string has an unlimited repetition quantifier, it no longer makes this
2419 optimization. That gives it a stack frame in which to save the data for
2420 checking that an empty string has been matched. Previously this was taken
2421 from the 1000-entry workspace that had been reserved. So now there is no
2422 explicit limit, but more stack is used.
2423
2424 8. Applied Daniel's patches to solve problems with the import/export magic
2425 syntax that is required for Windows, and which was going wrong for the
2426 pcreposix and pcrecpp parts of the library. These were overlooked when this
2427 problem was solved for the main library.
2428
2429 9. There were some crude static tests to avoid integer overflow when computing
2430 the size of patterns that contain repeated groups with explicit upper
2431 limits. As the maximum quantifier is 65535, the maximum group length was
2432 set at 30,000 so that the product of these two numbers did not overflow a
2433 32-bit integer. However, it turns out that people want to use groups that
2434 are longer than 30,000 bytes (though not repeat them that many times).
2435 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2436 made it possible to implement the integer overflow checks in a much more
2437 dynamic way, which I have now done. The artificial limitation on group
2438 length has been removed - we now have only the limit on the total length of
2439 the compiled pattern, which depends on the LINK_SIZE setting.
2440
2441 10. Fixed a bug in the documentation for get/copy named substring when
2442 duplicate names are permitted. If none of the named substrings are set, the
2443 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2444 empty string.
2445
2446 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2447 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2448 because the ] is interpreted as the first data character and the
2449 terminating ] is not found. PCRE has been made compatible with Perl in this
2450 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2451 cause memory overwriting.
2452
2453 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2454 string has been matched (to stop an infinite loop). It was not recognizing
2455 a conditional subpattern that could match an empty string if that
2456 subpattern was within another subpattern. For example, it looped when
2457 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
2458 condition was not nested. This bug has been fixed.
2459
2460 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2461 past the start of the subject in the presence of bytes with the top bit
2462 set, for example "\x8aBCD".
2463
2464 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2465 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2466
2467 14. Optimized (?!) to (*FAIL).
2468
2469 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2470 This restricts code points to be within the range 0 to 0x10FFFF, excluding
2471 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2472 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2473 does: it's just the validity check that is more restrictive.
2474
2475 16. Inserted checks for integer overflows during escape sequence (backslash)
2476 processing, and also fixed erroneous offset values for syntax errors during
2477 backslash processing.
2478
2479 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2480 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2481
2482 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2483 caused an overrun.
2484
2485 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2486 something other than just ASCII characters) inside a group that had an
2487 unlimited repeat caused a loop at compile time (while checking to see
2488 whether the group could match an empty string).
2489
2490 20. Debugging a pattern containing \p or \P could cause a crash. For example,
2491 [\P{Any}] did so. (Error in the code for printing property names.)
2492
2493 21. An orphan \E inside a character class could cause a crash.
2494
2495 22. A repeated capturing bracket such as (A)? could cause a wild memory
2496 reference during compilation.
2497
2498 23. There are several functions in pcre_compile() that scan along a compiled
2499 expression for various reasons (e.g. to see if it's fixed length for look
2500 behind). There were bugs in these functions when a repeated \p or \P was
2501 present in the pattern. These operators have additional parameters compared
2502 with \d, etc, and these were not being taken into account when moving along
2503 the compiled data. Specifically:
2504
2505 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2506 length.
2507
2508 (b) An item such as \pL+ within a repeated group could cause crashes or
2509 loops.
2510
2511 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2512 "reference to non-existent subpattern" error.
2513
2514 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2515
2516 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2517 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2518
2519 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2520
2521 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2522 character were causing crashes (broken optimization).
2523
2524 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2525 \p or \P) caused a compile-time loop.
2526
2527 28. More problems have arisen in unanchored patterns when CRLF is a valid line
2528 break. For example, the unstudied pattern [\r\n]A does not match the string
2529 "\r\nA" because change 7.0/46 below moves the current point on by two
2530 characters after failing to match at the start. However, the pattern \nA
2531 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2532 the same is true. There doesn't seem any very clean way out of this, but
2533 what I have chosen to do makes the common cases work: PCRE now takes note
2534 of whether there can be an explicit match for \r or \n anywhere in the
2535 pattern, and if so, 7.0/46 no longer applies. As part of this change,
2536 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2537 pattern has explicit CR or LF references.
2538
2539 29. Added (*CR) etc for changing newline setting at start of pattern.
2540
2541
2542 Version 7.2 19-Jun-07
2543 ---------------------
2544
2545 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2546 which is apparently normally available under Windows.
2547
2548 2. Re-jig the pcregrep tests with different newline settings in an attempt
2549 to make them independent of the local environment's newline setting.
2550
2551 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2552
2553 4. Some of the "internals" tests were previously cut out when the link size
2554 was not 2, because the output contained actual offsets. The recent new
2555 "Z" feature of pcretest means that these can be cut out, making the tests
2556 usable with all link sizes.
2557
2558 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2559 stack recursion. This gives a massive performance boost under BSD, but just
2560 a small improvement under Linux. However, it saves one field in the frame
2561 in all cases.
2562
2563 6. Added more features from the forthcoming Perl 5.10:
2564
2565 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2566 recursion call. It refers to the nth most recently opened parentheses.
2567
2568 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2569 to be opened parentheses.
2570
2571 (c) Conditions that refer to capturing parentheses can be specified
2572 relatively, for example, (?(-2)... or (?(+3)...
2573
2574 (d) \K resets the start of the current match so that everything before
2575 is not part of it.
2576
2577 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2578
2579 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2580 reference syntax.
2581
2582 (g) (?| introduces a group in which the numbering of parentheses in each
2583 alternative starts with the same number.
2584
2585 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2586
2587 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2588 PCRE_INFO_JCHANGED.
2589
2590 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2591 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2592 for detecting groups that can match an empty string.
2593
2594 9. A pattern with a very large number of alternatives (more than several
2595 hundred) was running out of internal workspace during the pre-compile
2596 phase, where pcre_compile() figures out how much memory will be needed. A
2597 bit of new cunning has reduced the workspace needed for groups with
2598 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2599 workspace instead of running out of the 4096 that are available.
2600
2601 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2602
2603 11. Applied patch from Google to remove an optimization that didn't quite work.
2604 The report of the bug said:
2605
2606 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2607 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2608 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2609
2610 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2611 it matched the wrong number of bytes.
2612
2613
2614 Version 7.1 24-Apr-07
2615 ---------------------
2616
2617 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2618 that is more "standard", making use of automake and other Autotools. There
2619 is some re-arrangement of the files and adjustment of comments consequent
2620 on this.
2621
2622 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2623 for recursive directory scanning broke on some systems because the files
2624 are not scanned in any specific order and on different systems the order
2625 was different. A call to "sort" has been inserted into RunGrepTest for the
2626 approprate test as a short-term fix. In the longer term there may be an
2627 alternative.
2628
2629 3. I had an email from Eric Raymond about problems translating some of PCRE's
2630 man pages to HTML (despite the fact that I distribute HTML pages, some
2631 people do their own conversions for various reasons). The problems
2632 concerned the use of low-level troff macros .br and .in. I have therefore
2633 removed all such uses from the man pages (some were redundant, some could
2634 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2635 HTML has been updated to handle .nf/.fi and to complain if it encounters
2636 .br or .in.
2637
2638 4. Updated comments in configure.ac that get placed in config.h.in and also
2639 arranged for config.h to be included in the distribution, with the name
2640 config.h.generic, for the benefit of those who have to compile without
2641 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2642
2643 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2644 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2645 makevp.bat for the current PCRE, using the additional files
2646 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2647
2648 6. A Windows user reported a minor discrepancy with test 2, which turned out
2649 to be caused by a trailing space on an input line that had got lost in his
2650 copy. The trailing space was an accident, so I've just removed it.
2651
2652 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2653 that is needed.
2654
2655 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2656 as "const" (a) because they are and (b) because it helps the PHP
2657 maintainers who have recently made a script to detect big data structures
2658 in the php code that should be moved to the .rodata section. I remembered
2659 to update Builducptable as well, so it won't revert if ucptable.h is ever
2660 re-created.
2661
2662 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2663 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2664 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2665 support is not required. This saves 1.5-2K of code, which is important in
2666 some applications.
2667
2668 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2669 so as not to refer to the tables, even though these functions will never be
2670 called when UTF-8 support is disabled. Otherwise there are problems with a
2671 shared library.
2672
2673 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2674
2675 (a) It was defining its arguments as char * instead of void *.
2676
2677 (b) It was assuming that all moves were upwards in memory; this was true
2678 a long time ago when I wrote it, but is no longer the case.
2679
2680 The emulated memove() is provided for those environments that have neither
2681 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2682 is clearly not the case, as these two bugs were recently reported.
2683
2684 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2685 and Detrail to create the HTML documentation, the .txt form of the man
2686 pages, and it removes trailing spaces from listed files. It also creates
2687 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2688 case, it wraps all the #defines with #ifndefs. This script should be run
2689 before "make dist".
2690
2691 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2692 with Unicode property support.
2693
2694 (a) For a maximizing quantifier, if the two different cases of the
2695 character were of different lengths in their UTF-8 codings (there are
2696 some cases like this - I found 11), and the matching function had to
2697 back up over a mixture of the two cases, it incorrectly assumed they
2698 were both the same length.
2699
2700 (b) When PCRE was configured to use the heap rather than the stack for
2701 recursion during matching, it was not correctly preserving the data for
2702 the other case of a UTF-8 character when checking ahead for a match
2703 while processing a minimizing repeat. If the check also involved
2704 matching a wide character, but failed, corruption could cause an
2705 erroneous result when trying to check for a repeat of the original
2706 character.
2707
2708 13. Some tidying changes to the testing mechanism:
2709
2710 (a) The RunTest script now detects the internal link size and whether there
2711 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2712 values substituted by "configure". (The RunGrepTest script already did
2713 this for UTF-8.) The configure.ac script no longer substitutes the
2714 relevant variables.
2715
2716 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2717 with length and offset values. This means that the output is different
2718 for different internal link sizes. Test 2 is skipped for link sizes
2719 other than 2 because of this, bypassing the problem. Unfortunately,
2720 there was also a test in test 3 (the locale tests) that used /B and
2721 failed for link sizes other than 2. Rather than cut the whole test out,
2722 I have added a new /Z option to pcretest that replaces the length and
2723 offset values with spaces. This is now used to make test 3 independent
2724 of link size. (Test 2 will be tidied up later.)
2725
2726 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2727 segmentation fault instead of returning the appropriate error message.
2728
2729 15. In multiline mode when the newline sequence was set to "any", the pattern
2730 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2731 This doesn't seem right; it now treats the CRLF combination as the line
2732 ending, and so does not match in that case. It's only a pattern such as ^$
2733 that would hit this one: something like ^ABC$ would have failed after \r
2734 and then tried again after \r\n.
2735
2736 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2737 in an attempt to make files that differ only in their line terminators
2738 compare equal. This works on Linux.
2739
2740 17. Under certain error circumstances pcregrep might try to free random memory
2741 as it exited. This is now fixed, thanks to valgrind.
2742
2743 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2744 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2745 was because its rules for how to advance for /g after matching an empty
2746 string at the end of a line did not allow for this case. They now check for
2747 it specially.
2748
2749 20. pcretest is supposed to handle patterns and data of any length, by
2750 extending its buffers when necessary. It was getting this wrong when the
2751 buffer for a data line had to be extended.
2752
2753 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2754 CRLF as a newline sequence.
2755
2756 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2757 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2758 I have nevertheless tidied it up.
2759
2760 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2761
2762 24. Added a man page for pcre-config.
2763
2764
2765 Version 7.0 19-Dec-06
2766 ---------------------
2767
2768 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2769 moving to gcc 4.1.1.
2770
2771 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2772 sys/time.h, which is documented as needed for this function. It doesn't
2773 seem to matter on Linux, but it showed up on some releases of OS X.
2774
2775 3. It seems that there are systems where bytes whose values are greater than
2776 127 match isprint() in the "C" locale. The "C" locale should be the
2777 default when a C program starts up. In most systems, only ASCII printing
2778 characters match isprint(). This difference caused the output from pcretest
2779 to vary, making some of the tests fail. I have changed pcretest so that:
2780
2781 (a) When it is outputting text in the compiled version of a pattern, bytes
2782 other than 32-126 are always shown as hex escapes.
2783
2784 (b) When it is outputting text that is a matched part of a subject string,
2785 it does the same, unless a different locale has been set for the match
2786 (using the /L modifier). In this case, it uses isprint() to decide.
2787
2788 4. Fixed a major bug that caused incorrect computation of the amount of memory
2789 required for a compiled pattern when options that changed within the
2790 pattern affected the logic of the preliminary scan that determines the
2791 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2792 that the computed length was too small. The symptoms of this bug were
2793 either the PCRE error "internal error: code overflow" from pcre_compile(),
2794 or a glibc crash with a message such as "pcretest: free(): invalid next
2795 size (fast)". Examples of patterns that provoked this bug (shown in
2796 pcretest format) are:
2797
2798 /(?-x: )/x
2799 /(?x)(?-x: \s*#\s*)/
2800 /((?i)[\x{c0}])/8
2801 /(?i:[\x{c0}])/8
2802
2803 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2804 is now done differently.
2805
2806 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2807 wrapper classes; (b) implement a new function in the C++ scanner that is
2808 more efficient than the old way of doing things because it avoids levels of
2809 recursion in the regex matching; (c) add a paragraph to the documentation
2810 for the FullMatch() function.
2811
2812 6. The escape sequence \n was being treated as whatever was defined as
2813 "newline". Not only was this contrary to the documentation, which states
2814 that \n is character 10 (hex 0A), but it also went horribly wrong when
2815 "newline" was defined as CRLF. This has been fixed.
2816
2817 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2818 was being set to -1 for the "end of line" case (supposedly a value that no
2819 character can have). Though this value is never used (the check for end of
2820 line is "zero bytes in current character"), it caused compiler complaints.
2821 I've changed it to 0xffffffff.
2822
2823 8. In pcre_version.c, the version string was being built by a sequence of
2824 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2825 string (as it is for production releases) called a macro with an empty
2826 argument. The C standard says the result of this is undefined. The gcc
2827 compiler treats it as an empty string (which was what was wanted) but it is
2828 reported that Visual C gives an error. The source has been hacked around to
2829 avoid this problem.
2830
2831 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2832 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2833 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2834 of them did).
2835
2836 10. Originally, pcretest opened its input and output without "b"; then I was
2837 told that "b" was needed in some environments, so it was added for release
2838 5.0 to both the input and output. (It makes no difference on Unix-like
2839 systems.) Later I was told that it is wrong for the input on Windows. I've
2840 now abstracted the modes into two macros, to make it easier to fiddle with
2841 them, and removed "b" from the input mode under Windows.
2842
2843 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2844
2845 12. Added -help and --help to pcretest as an official way of being reminded
2846 of the options.
2847
2848 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2849 and pcrecpp.cc because they annoy compilers at high warning levels.
2850
2851 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2852
2853 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2854 scripts are not C programs :-) and which was not noticed because it works
2855 on Linux.
2856
2857 16. pcretest is supposed to handle any length of pattern and data line (as one
2858 line or as a continued sequence of lines) by extending its input buffer if
2859 necessary. This feature was broken for very long pattern lines, leading to
2860 a string of junk being passed to pcre_compile() if the pattern was longer
2861 than about 50K.
2862
2863 17. I have done a major re-factoring of the way pcre_compile() computes the
2864 amount of memory needed for a compiled pattern. Previously, there was code
2865 that made a preliminary scan of the pattern in order to do this. That was
2866 OK when PCRE was new, but as the facilities have expanded, it has become
2867 harder and harder to keep it in step with the real compile phase, and there
2868 have been a number of bugs (see for example, 4 above). I have now found a
2869 cunning way of running the real compile function in a "fake" mode that
2870 enables it to compute how much memory it would need, while actually only
2871 ever using a few hundred bytes of working memory and without too many
2872 tests of the mode. This should make future maintenance and development
2873 easier. A side effect of this work is that the limit of 200 on the nesting
2874 depth of parentheses has been removed (though this was never a serious
2875 limitation, I suspect). However, there is a downside: pcre_compile() now
2876 runs more slowly than before (30% or more, depending on the pattern). I
2877 hope this isn't a big issue. There is no effect on runtime performance.
2878
2879 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2880 newline (only possible for the last line of a file) and it was a
2881 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2882
2883 19. Added additional timing features to pcretest. (1) The -tm option now times
2884 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2885 separate command line item, by a number that specifies the number of
2886 repeats to use when timing. The default is 50000; this gives better
2887 precision, but takes uncomfortably long for very large patterns.
2888
2889 20. Extended pcre_study() to be more clever in cases where a branch of a
2890 subpattern has no definite first character. For example, (a*|b*)[cd] would
2891 previously give no result from pcre_study(). Now it recognizes that the
2892 first character must be a, b, c, or d.
2893
2894 21. There was an incorrect error "recursive call could loop indefinitely" if
2895 a subpattern (or the entire pattern) that was being tested for matching an
2896 empty string contained only one non-empty item after a nested subpattern.
2897 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2898 incorrectly, because the \d was being skipped in the check.
2899
2900 22. The pcretest program now has a new pattern option /B and a command line
2901 option -b, which is equivalent to adding /B to every pattern. This causes
2902 it to show the compiled bytecode, without the additional information that
2903 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2904 is the same as /B/I).
2905
2906 23. A new optimization is now able automatically to treat some sequences such
2907 as a*b as a*+b. More specifically, if something simple (such as a character
2908 or a simple class like \d) has an unlimited quantifier, and is followed by
2909 something that cannot possibly match the quantified thing, the quantifier
2910 is automatically "possessified".
2911
2912 24. A recursive reference to a subpattern whose number was greater than 39
2913 went wrong under certain circumstances in UTF-8 mode. This bug could also
2914 have affected the operation of pcre_study().
2915
2916 25. Realized that a little bit of performance could be had by replacing
2917 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2918
2919 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2920
2921 27. Possessive quantifiers such as a++ were previously implemented by turning
2922 them into atomic groups such as ($>a+). Now they have their own opcodes,
2923 which improves performance. This includes the automatically created ones
2924 from 23 above.
2925
2926 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2927 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2928 the first matched character to be a colon. This applied both to named and
2929 numbered groups.
2930
2931 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2932
2933 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2934 building PCRE on Windows easier, so I have included it in the distribution.
2935
2936 31. There is now a check in pcretest against a ridiculously large number being
2937 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2938 loop, the loop is abandoned.
2939
2940 32. Forward references to subpatterns in conditions such as (?(2)...) where
2941 subpattern 2 is defined later cause pcre_compile() to search forwards in
2942 the pattern for the relevant set of parentheses. This search went wrong
2943 when there were unescaped parentheses in a character class, parentheses
2944 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2945
2946 33. "Subroutine" calls and backreferences were previously restricted to
2947 referencing subpatterns earlier in the regex. This restriction has now
2948 been removed.
2949
2950 34. Added a number of extra features that are going to be in Perl 5.10. On the
2951 whole, these are just syntactic alternatives for features that PCRE had
2952 previously implemented using the Python syntax or my own invention. The
2953 other formats are all retained for compatibility.
2954
2955 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2956 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2957 also .NET compatible.
2958
2959 (b) A recursion or subroutine call to a named group can now be defined as
2960 (?&name) as well as (?P>name).
2961
2962 (c) A backreference to a named group can now be defined as \k<name> or
2963 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2964 5.10, are also .NET compatible.
2965
2966 (d) A conditional reference to a named group can now use the syntax
2967 (?(<name>) or (?('name') as well as (?(name).
2968
2969 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2970 groups (named and numbered) that are never evaluated inline, but can be
2971 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2972 is always false. There may be only one alternative in such a group.
2973
2974 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2975 as the simple (?(R). The condition is true only if the most recent
2976 recursion is that of the given number or name. It does not search out
2977 through the entire recursion stack.
2978
2979 (g) The escape \gN or \g{N} has been added, where N is a positive or
2980 negative number, specifying an absolute or relative reference.
2981
2982 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2983 some "unreachable code" warnings.
2984
2985 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2986 things, this adds five new scripts.
2987
2988 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2989 There were also incompatibilities regarding the handling of \Q..\E inside
2990 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2991 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2992
2993 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2994 matches an empty string, and forcibly breaks the loop. There were bugs in
2995 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2996 against aaaa the result was just "a" rather than "aaaa", for example. Two
2997 separate and independent bugs (that affected different cases) have been
2998 fixed.
2999
3000 39. Refactored the code to abolish the use of different opcodes for small
3001 capturing bracket numbers. This is a tidy that I avoided doing when I
3002 removed the limit on the number of capturing brackets for 3.5 back in 2001.
3003 The new approach is not only tidier, it makes it possible to reduce the
3004 memory needed to fix the previous bug (38).
3005
3006 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
3007 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
3008 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
3009 mode.
3010
3011 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
3012 report.
3013
3014 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
3015 copy construction and assignment in the C++ wrapper.
3016
3017 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
3018 couple of bugs that could have given wrong results in the "--newline=crlf"
3019 case.
3020
3021 44. Added a number of casts and did some reorganization of signed/unsigned int
3022 variables following suggestions from Dair Grant. Also renamed the variable
3023 "this" as "item" because it is a C++ keyword.
3024
3025 45. Arranged for dftables to add
3026
3027 #include "pcre_internal.h"
3028
3029 to pcre_chartables.c because without it, gcc 4.x may remove the array
3030 definition from the final binary if PCRE is built into a static library and
3031 dead code stripping is activated.
3032
3033 46. For an unanchored pattern, if a match attempt fails at the start of a
3034 newline sequence, and the newline setting is CRLF or ANY, and the next two
3035 characters are CRLF, advance by two characters instead of one.
3036
3037
3038 Version 6.7 04-Jul-06
3039 ---------------------
3040
3041 1. In order to handle tests when input lines are enormously long, pcretest has
3042 been re-factored so that it automatically extends its buffers when
3043 necessary. The code is crude, but this _is_ just a test program. The
3044 default size has been increased from 32K to 50K.
3045
3046 2. The code in pcre_study() was using the value of the re argument before
3047 testing it for NULL. (Of course, in any sensible call of the function, it
3048 won't be NULL.)
3049
3050 3. The memmove() emulation function in pcre_internal.h, which is used on
3051 systems that lack both memmove() and bcopy() - that is, hardly ever -
3052 was missing a "static" storage class specifier.
3053
3054 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
3055 containing an extended class (one that cannot be represented by a bitmap
3056 because it contains high-valued characters or Unicode property items, e.g.
3057 [\pZ]). Almost always one would set UTF-8 mode when processing such a
3058 pattern, but PCRE should not loop if you do not (it no longer does).
3059 [Detail: two cases were found: (a) a repeated subpattern containing an
3060 extended class; (b) a recursive reference to a subpattern that followed a
3061 previous extended class. It wasn't skipping over the extended class
3062 correctly when UTF-8 mode was not set.]
3063
3064 5. A negated single-character class was not being recognized as fixed-length
3065 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
3066 compile error "lookbehind assertion is not fixed length".
3067
3068 6. The RunPerlTest auxiliary script was showing an unexpected difference
3069 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
3070 write a Perl script that can interpret lines of an input file either as
3071 byte characters or as UTF-8, which is what "perltest" was being required to
3072 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
3073 can't do is switch easily at run time between having the "use utf8;" pragma
3074 or not. In the end, I fudged it by using the RunPerlTest script to insert
3075 "use utf8;" explicitly for the UTF-8 tests.
3076
3077 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
3078 the end of the subject string, contrary to the documentation and to what
3079 Perl does. This was true of both matching functions. Now it matches only at
3080 the start of the subject and immediately after *internal* newlines.
3081
3082 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
3083 a pointer to an int instead of a pointer to an unsigned long int. This
3084 caused problems on 64-bit systems.
3085
3086 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
3087 instance of the 'standard' template library not being so standard".
3088
3089 10. There was no check on the number of named subpatterns nor the maximum
3090 length of a subpattern name. The product of these values is used to compute
3091 the size of the memory block for a compiled pattern. By supplying a very
3092 long subpattern name and a large number of named subpatterns, the size
3093 computation could be caused to overflow. This is now prevented by limiting
3094 the length of names to 32 characters, and the number of named subpatterns
3095 to 10,000.
3096
3097 11. Subpatterns that are repeated with specific counts have to be replicated in
3098 the compiled pattern. The size of memory for this was computed from the
3099 length of the subpattern and the repeat count. The latter is limited to
3100 65535, but there was no limit on the former, meaning that integer overflow
3101 could in principle occur. The compiled length of a repeated subpattern is
3102 now limited to 30,000 bytes in order to prevent this.
3103
3104 12. Added the optional facility to have named substrings with the same name.
3105
3106 13. Added the ability to use a named substring as a condition, using the
3107 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
3108 are numbers (not recommended). Forward references are permitted.
3109
3110 14. Added forward references in named backreferences (if you see what I mean).
3111
3112 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
3113 pattern could run off the end of the subject. For example, the pattern
3114 "(?s)(.{1,5})"8 did this with the subject "ab".
3115
3116 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
3117 PCRE_CASELESS was set when matching characters that were quantified with ?
3118 or *.
3119
3120 17. A character class other than a single negated character that had a minimum
3121 but no maximum quantifier - for example [ab]{6,} - was not handled
3122 correctly by pce_dfa_exec(). It would match only one character.
3123
3124 18. A valid (though odd) pattern that looked like a POSIX character
3125 class but used an invalid character after [ (for example [[,abc,]]) caused
3126 pcre_compile() to give the error "Failed: internal error: code overflow" or
3127 in some cases to crash with a glibc free() error. This could even happen if
3128 the pattern terminated after [[ but there just happened to be a sequence of
3129 letters, a binary zero, and a closing ] in the memory that followed.
3130
3131 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
3132 over the years. Originally (before any Unicode support), just the bottom 8
3133 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
3134 output from "man perlunicode" includes this:
3135
3136 The regular expression compiler produces polymorphic opcodes. That
3137 is, the pattern adapts to the data and automatically switches to
3138 the Unicode character scheme when presented with Unicode data--or
3139 instead uses a traditional byte scheme when presented with byte
3140 data.
3141
3142 Sadly, a wide octal escape does not cause a switch, and in a string with
3143 no other multibyte characters, these octal escapes are treated as before.
3144 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
3145 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
3146 Unicode string.
3147
3148 I have not perpetrated such confusion in PCRE. Up till now, it took just
3149 the bottom 8 bits, as in old Perl. I have now made octal escapes with
3150 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
3151 translate to the appropriate multibyte character.
3152
3153 29. Applied some refactoring to reduce the number of warnings from Microsoft
3154 and Borland compilers. This has included removing the fudge introduced
3155 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
3156 a warning about an unused variable.
3157
3158 21. PCRE has not included VT (character 0x0b) in the set of whitespace
3159 characters since release 4.0, because Perl (from release 5.004) does not.
3160 [Or at least, is documented not to: some releases seem to be in conflict
3161 with the documentation.] However, when a pattern was studied with
3162 pcre_study() and all its branches started with \s, PCRE still included VT
3163 as a possible starting character. Of course, this did no harm; it just
3164 caused an unnecessary match attempt.
3165
3166 22. Removed a now-redundant internal flag bit that recorded the fact that case
3167 dependency changed within the pattern. This was once needed for "required
3168 byte" processing, but is no longer used. This recovers a now-scarce options
3169 bit. Also moved the least significant internal flag bit to the most-
3170 significant bit of the word, which was not previously used (hangover from
3171 the days when it was an int rather than a uint) to free up another bit for
3172 the future.
3173
3174 23. Added support for CRLF line endings as well as CR and LF. As well as the
3175 default being selectable at build time, it can now be changed at runtime
3176 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
3177 specify that it is scanning data with non-default line endings.
3178
3179 24. Changed the definition of CXXLINK to make it agree with the definition of
3180 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
3181
3182 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
3183 recursions. This makes a big different to stack usage for some patterns.
3184
3185 26. If a subpattern containing a named recursion or subroutine reference such
3186 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
3187 the space required for the compiled pattern went wrong and gave too small a
3188 value. Depending on the environment, this could lead to "Failed: internal
3189 error: code overflow at offset 49" or "glibc detected double free or
3190 corruption" errors.
3191
3192 27. Applied patches from Google (a) to support the new newline modes and (b) to
3193 advance over multibyte UTF-8 characters in GlobalReplace.
3194
3195 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
3196 difference for some implementation of PCRE in some Windows version.
3197
3198 29. Added some extra testing facilities to pcretest:
3199
3200 \q<number> in a data line sets the "match limit" value
3201 \Q<number> in a data line sets the "match recursion limt" value
3202 -S <number> sets the stack size, where <number> is in megabytes
3203
3204 The -S option isn't available for Windows.
3205
3206
3207 Version 6.6 06-Feb-06
3208 ---------------------
3209
3210 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
3211 in pcreposix.h. I have copied the definition from pcre.h.
3212
3213 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
3214 because pcre.h is no longer a built file.
3215
3216 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
3217 not normally included in the compiled code.
3218
3219
3220 Version 6.5 01-Feb-06
3221 ---------------------
3222
3223 1. When using the partial match feature with pcre_dfa_exec(), it was not
3224 anchoring the second and subsequent partial matches at the new starting
3225 point. This could lead to incorrect results. For example, with the pattern
3226 /1234/, partially matching against "123" and then "a4" gave a match.
3227
3228 2. Changes to pcregrep:
3229
3230 (a) All non-match returns from pcre_exec() were being treated as failures
3231 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
3232 error message is output. Some extra information is given for the
3233 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
3234 probably the only errors that are likely to be caused by users (by
3235 specifying a regex that has nested indefinite repeats, for instance).
3236 If there are more than 20 of these errors, pcregrep is abandoned.
3237
3238 (b) A binary zero was treated as data while matching, but terminated the
3239 output line if it was written out. This has been fixed: binary zeroes
3240 are now no different to any other data bytes.
3241
3242 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3243 used to set a locale for matching. The --locale=xxxx long option has
3244 been added (no short equivalent) to specify a locale explicitly on the
3245 pcregrep command, overriding the environment variables.
3246
3247 (d) When -B was used with -n, some line numbers in the output were one less
3248 than they should have been.
3249
3250 (e) Added the -o (--only-matching) option.
3251
3252 (f) If -A or -C was used with -c (count only), some lines of context were
3253 accidentally printed for the final match.
3254
3255 (g) Added the -H (--with-filename) option.
3256
3257 (h) The combination of options -rh failed to suppress file names for files
3258 that were found from directory arguments.
3259
3260 (i) Added the -D (--devices) and -d (--directories) options.
3261
3262 (j) Added the -F (--fixed-strings) option.
3263
3264 (k) Allow "-" to be used as a file name for -f as well as for a data file.
3265
3266 (l) Added the --colo(u)r option.
3267
3268 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3269 is not present by default.
3270
3271 3. A nasty bug was discovered in the handling of recursive patterns, that is,
3272 items such as (?R) or (?1), when the recursion could match a number of
3273 alternatives. If it matched one of the alternatives, but subsequently,
3274 outside the recursion, there was a failure, the code tried to back up into
3275 the recursion. However, because of the way PCRE is implemented, this is not
3276 possible, and the result was an incorrect result from the match.
3277
3278 In order to prevent this happening, the specification of recursion has
3279 been changed so that all such subpatterns are automatically treated as
3280 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3281
3282 4. I had overlooked the fact that, in some locales, there are characters for
3283 which isalpha() is true but neither isupper() nor islower() are true. In
3284 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3285 and ordfeminine) are like this. This affected the treatment of \w and \W
3286 when they appeared in character classes, but not when they appeared outside
3287 a character class. The bit map for "word" characters is now created
3288 separately from the results of isalnum() instead of just taking it from the
3289 upper, lower, and digit maps. (Plus the underscore character, of course.)
3290
3291 5. The above bug also affected the handling of POSIX character classes such as
3292 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3293 permanent tables. Instead, the bit maps for such a class were previously
3294 created as the appropriate unions of the upper, lower, and digit bitmaps.
3295 Now they are created by subtraction from the [[:word:]] class, which has
3296 its own bitmap.
3297
3298 6. The [[:blank:]] character class matches horizontal, but not vertical space.
3299 It is created by subtracting the vertical space characters (\x09, \x0a,
3300 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3301 subtraction was done in the overall bitmap for a character class, meaning
3302 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3303 be recognized. This bug has been fixed.
3304
3305 7. Patches from the folks at Google:
3306
3307 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3308 real life, but is still worth protecting against".
3309
3310 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3311 regular expressions".
3312
3313 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3314 have it.
3315
3316 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3317 "configure" and the latter not, in order to fix a problem somebody had
3318 with compiling the Arg class on HP-UX.
3319
3320 (e) Improve the error-handling of the C++ wrapper a little bit.
3321
3322 (f) New tests for checking recursion limiting.
3323
3324 8. The pcre_memmove() function, which is used only if the environment does not
3325 have a standard memmove() function (and is therefore rarely compiled),
3326 contained two bugs: (a) use of int instead of size_t, and (b) it was not
3327 returning a result (though PCRE never actually uses the result).
3328
3329 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3330 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3331 returned instead of calling malloc() with an overflowing number that would
3332 most likely cause subsequent chaos.
3333
3334 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3335
3336 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3337 with this option is matched, the nmatch and pmatch options of regexec() are
3338 ignored.
3339
3340 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3341 provided in case anyone wants to the the POSIX interface with UTF-8
3342 strings.
3343
3344 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3345 C++ linking (needed for some HP-UX environments).
3346
3347 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3348 (unused parameter) and in the pcre_printint() function (omitted "default"
3349 switch label when the default is to do nothing).
3350
3351 15. Added some code to make it possible, when PCRE is compiled as a C++
3352 library, to replace subject pointers for pcre_exec() with a smart pointer
3353 class, thus making it possible to process discontinuous strings.
3354
3355 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3356 much the same function. They were added by different people who were trying
3357 to make PCRE easy to compile on non-Unix systems. It has been suggested
3358 that PCRE_EXPORT be abolished now that there is more automatic apparatus
3359 for compiling on Windows systems. I have therefore replaced it with
3360 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3361 defaults to "extern" for C or "extern C" for C++, which works fine on
3362 Unix-like systems. It is now possible to override the value of PCRE_DATA_
3363 SCOPE with something explicit in config.h. In addition:
3364
3365 (a) pcreposix.h still had just "extern" instead of either of these macros;
3366 I have replaced it with PCRE_DATA_SCOPE.
3367
3368 (b) Functions such as _pcre_xclass(), which are internal to the library,
3369 but external in the C sense, all had PCRE_EXPORT in their definitions.
3370 This is apparently wrong for the Windows case, so I have removed it.
3371 (It makes no difference on Unix-like systems.)
3372
3373 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3374 of recursive calls to match(). This is different to MATCH_LIMIT because
3375 that limits the total number of calls to match(), not all of which increase
3376 the depth of recursion. Limiting the recursion depth limits the amount of
3377 stack (or heap if NO_RECURSE is set) that is used. The default can be set
3378 when PCRE is compiled, and changed at run time. A patch from Google adds
3379 this functionality to the C++ interface.
3380
3381 18. Changes to the handling of Unicode character properties:
3382
3383 (a) Updated the table to Unicode 4.1.0.
3384
3385 (b) Recognize characters that are not in the table as "Cn" (undefined).
3386
3387 (c) I revised the way the table is implemented to a much improved format
3388 which includes recognition of ranges. It now supports the ranges that
3389 are defined in UnicodeData.txt, and it also amalgamates other
3390 characters into ranges. This has reduced the number of entries in the
3391 table from around 16,000 to around 3,000, thus reducing its size
3392 considerably. I realized I did not need to use a tree structure after
3393 all - a binary chop search is just as efficient. Having reduced the
3394 number of entries, I extended their size from 6 bytes to 8 bytes to
3395 allow for more data.
3396
3397 (d) Added support for Unicode script names via properties such as \p{Han}.
3398
3399 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3400 matching that character.
3401
3402 20. When matching a repeated Unicode property with a minimum greater than zero,
3403 (for example \pL{2,}), PCRE could look past the end of the subject if it
3404 reached it while seeking the minimum number of characters. This could
3405 happen only if some of the characters were more than one byte long, because
3406 there is a check for at least the minimum number of bytes.
3407
3408 21. Refactored the implementation of \p and \P so as to be more general, to
3409 allow for more different types of property in future. This has changed the
3410 compiled form incompatibly. Anybody with saved compiled patterns that use
3411 \p or \P will have to recompile them.
3412
3413 22. Added "Any" and "L&" to the supported property types.
3414
3415 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3416 but give a compile time error if the value is greater than 0xff.
3417
3418 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3419 accidentally not being installed or uninstalled.
3420
3421 25. The pcre.h file was built from pcre.h.in, but the only changes that were
3422 made were to insert the current release number. This seemed silly, because
3423 it made things harder for people building PCRE on systems that don't run
3424 "configure". I have turned pcre.h into a distributed file, no longer built
3425 by "configure", with the version identification directly included. There is
3426 no longer a pcre.h.in file.
3427
3428 However, this change necessitated a change to the pcre-config script as
3429 well. It is built from pcre-config.in, and one of the substitutions was the
3430 release number. I have updated configure.ac so that ./configure now finds
3431 the release number by grepping pcre.h.
3432
3433 26. Added the ability to run the tests under valgrind.
3434
3435
3436 Version 6.4 05-Sep-05
3437 ---------------------
3438
3439 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3440 "--" to be printed when multiple files were scanned, even when none of the
3441 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3442 consider it to be a bug, and have restored the previous behaviour.
3443
3444 2. A couple of code tidies to get rid of compiler warnings.
3445
3446 3. The pcretest program used to cheat by referring to symbols in the library
3447 whose names begin with _pcre_. These are internal symbols that are not
3448 really supposed to be visible externally, and in some environments it is
3449 possible to suppress them. The cheating is now confined to including
3450 certain files from the library's source, which is a bit cleaner.
3451
3452 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3453 file's purpose clearer.
3454
3455 5. Reorganized pcre_ucp_findchar().
3456
3457
3458 Version 6.3 15-Aug-05
3459 ---------------------
3460
3461 1. The file libpcre.pc.in did not have general read permission in the tarball.
3462
3463 2. There were some problems when building without C++ support:
3464
3465 (a) If C++ support was not built, "make install" and "make test" still
3466 tried to test it.
3467
3468 (b) There were problems when the value of CXX was explicitly set. Some
3469 changes have been made to try to fix these, and ...
3470
3471 (c) --disable-cpp can now be used to explicitly disable C++ support.
3472
3473 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3474 backslash in a target when C++ was disabled. This confuses some
3475 versions of "make", apparently. Using an intermediate variable solves
3476 this. (Same for CPP_LOBJ.)
3477
3478 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3479 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3480 necessary on certain architectures.
3481
3482 4. Added a setting of -export-symbols-regex to the link command to remove
3483 those symbols that are exported in the C sense, but actually are local
3484 within the library, and not documented. Their names all begin with
3485 "_pcre_". This is not a perfect job, because (a) we have to except some
3486 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3487 available (and never for static libraries). I have made a note to try to
3488 find a way round (a) in the future.
3489
3490
3491 Version 6.2 01-Aug-05
3492 ---------------------
3493
3494 1. There was no test for integer overflow of quantifier values. A construction
3495 such as {1111111111111111} would give undefined results. What is worse, if
3496 a minimum quantifier for a parenthesized subpattern overflowed and became
3497 negative, the calculation of the memory size went wrong. This could have
3498 led to memory overwriting.
3499
3500 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3501
3502 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3503 operating environments where this matters.
3504
3505 4. Applied Giuseppe Maxia's patch to add additional features for controlling
3506 PCRE options from within the C++ wrapper.
3507
3508 5. Named capturing subpatterns were not being correctly counted when a pattern
3509 was compiled. This caused two problems: (a) If there were more than 100
3510 such subpatterns, the calculation of the memory needed for the whole
3511 compiled pattern went wrong, leading to an overflow error. (b) Numerical
3512 back references of the form \12, where the number was greater than 9, were
3513 not recognized as back references, even though there were sufficient
3514 previous subpatterns.
3515
3516 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3517 versions of gcc, e.g. 2.95.4.
3518
3519
3520 Version 6.1 21-Jun-05
3521 ---------------------
3522
3523 1. There was one reference to the variable "posix" in pcretest.c that was not
3524 surrounded by "#if !defined NOPOSIX".
3525
3526 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3527 the cross-check on the old pcre_info() function, for the benefit of the
3528 cut-down version of PCRE that is currently imported into Exim.
3529
3530 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3531 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3532 patterns that start (?i) but I don't think that matters. The use of (?i) is
3533 just an example; this all applies to the other options as well.
3534
3535 4. Since libtool seems to echo the compile commands it is issuing, the output
3536 from "make" can be reduced a bit by putting "@" in front of each libtool
3537 compile command.
3538
3539 5. Patch from the folks at Google for configure.in to be a bit more thorough
3540 in checking for a suitable C++ installation before trying to compile the
3541 C++ stuff. This should fix a reported problem when a compiler was present,
3542 but no suitable headers.
3543
3544 6. The man pages all had just "PCRE" as their title. I have changed them to
3545 be the relevant file name. I have also arranged that these names are
3546 retained in the file doc/pcre.txt, which is a concatenation in text format
3547 of all the man pages except the little individual ones for each function.
3548
3549 7. The NON-UNIX-USE file had not been updated for the different set of source
3550 files that come with release 6. I also added a few comments about the C++
3551 wrapper.
3552
3553
3554 Version 6.0 07-Jun-05
3555 ---------------------
3556
3557 1. Some minor internal re-organization to help with my DFA experiments.
3558
3559 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3560 didn't matter for the library itself when fully configured, but did matter
3561 when compiling without UCP support, or within Exim, where the ucp files are
3562 not imported.
3563
3564 3. Refactoring of the library code to split up the various functions into
3565 different source modules. The addition of the new DFA matching code (see
3566 below) to a single monolithic source would have made it really too
3567 unwieldy, quite apart from causing all the code to be include in a
3568 statically linked application, when only some functions are used. This is
3569 relevant even without the DFA addition now that patterns can be compiled in
3570 one application and matched in another.
3571
3572 The downside of splitting up is that there have to be some external
3573 functions and data tables that are used internally in different modules of
3574 the library but which are not part of the API. These have all had their
3575 names changed to start with "_pcre_" so that they are unlikely to clash
3576 with other external names.
3577
3578 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3579 a different (DFA) algorithm. Although it is slower than the original
3580 function, it does have some advantages for certain types of matching
3581 problem.
3582
3583 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3584 including restarting after a partial match.
3585
3586 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3587 defined when compiling for Windows was sent to me. I have put it into the
3588 code, though I have no means of testing or verifying it.
3589
3590 7. Added the pcre_refcount() auxiliary function.
3591
3592 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3593 match before or at the first newline in the subject string. In pcretest,
3594 the /f option on a pattern can be used to set this.
3595
3596 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3597 would behave wrongly. This has been present in PCRE since release 4.0.
3598
3599 10. A number of changes to the pcregrep command:
3600
3601 (a) Refactored how -x works; insert ^(...)$ instead of setting
3602 PCRE_ANCHORED and checking the length, in preparation for adding
3603 something similar for -w.
3604
3605 (b) Added the -w (match as a word) option.
3606
3607 (c) Refactored the way lines are read and buffered so as to have more
3608 than one at a time available.
3609
3610 (d) Implemented a pcregrep test script.
3611
3612 (e) Added the -M (multiline match) option. This allows patterns to match
3613 over several lines of the subject. The buffering ensures that at least
3614 8K, or the rest of the document (whichever is the shorter) is available
3615 for matching (and similarly the previous 8K for lookbehind assertions).
3616
3617 (f) Changed the --help output so that it now says
3618
3619 -w, --word-regex(p)
3620
3621 instead of two lines, one with "regex" and the other with "regexp"
3622 because that confused at least one person since the short forms are the
3623 same. (This required a bit of code, as the output is generated
3624 automatically from a table. It wasn't just a text change.)
3625
3626 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3627 option but starts with a hyphen. Could be a pattern or a path name
3628 starting with a hyphen, for instance.
3629
3630 (h) "-" can be given as a file name to represent stdin.
3631
3632 (i) When file names are being printed, "(standard input)" is used for
3633 the standard input, for compatibility with GNU grep. Previously
3634 "<stdin>" was used.
3635
3636 (j) The option --label=xxx can be used to supply a name to be used for
3637 stdin when file names are being printed. There is no short form.
3638
3639 (k) Re-factored the options decoding logic because we are going to add
3640 two more options that take data. Such options can now be given in four
3641 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3642
3643 (l) Added the -A, -B, and -C options for requesting that lines of context
3644 around matches be printed.
3645
3646 (m) Added the -L option to print the names of files that do not contain
3647 any matching lines, that is, the complement of -l.
3648
3649 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3650 continue to scan other files.
3651
3652 (o) The -s option was incorrectly implemented. For compatibility with other
3653 greps, it now suppresses the error message for a non-existent or non-
3654 accessible file (but not the return code). There is a new option called
3655 -q that suppresses the output of matching lines, which was what -s was
3656 previously doing.
3657
3658 (p) Added --include and --exclude options to specify files for inclusion
3659 and exclusion when recursing.
3660
3661 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3662 Hopefully, it now does.
3663
3664 12. Missing cast in pcre_study().
3665
3666 13. Added an "uninstall" target to the makefile.
3667
3668 14. Replaced "extern" in the function prototypes in Makefile.in with
3669 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3670 world, but is set differently for Windows.
3671
3672 15. Added a second compiling function called pcre_compile2(). The only
3673 difference is that it has an extra argument, which is a pointer to an
3674 integer error code. When there is a compile-time failure, this is set
3675 non-zero, in addition to the error test pointer being set to point to an
3676 error message. The new argument may be NULL if no error number is required
3677 (but then you may as well call pcre_compile(), which is now just a
3678 wrapper). This facility is provided because some applications need a
3679 numeric error indication, but it has also enabled me to tidy up the way
3680 compile-time errors are handled in the POSIX wrapper.
3681
3682 16. Added VPATH=.libs to the makefile; this should help when building with one
3683 prefix path and installing with another. (Or so I'm told by someone who
3684 knows more about this stuff than I do.)
3685
3686 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3687 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3688 match everything, including newlines. This is not POSIX-compatible, but
3689 somebody wanted the feature. From pcretest it can be activated by using
3690 both the P and the s flags.
3691
3692 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3693
3694 19. libpcre.pc was being incorrectly installed as executable.
3695
3696 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3697 it now also looks for '\r' so that it will work unmodified on Windows.
3698
3699 21. Added Google's contributed C++ wrapper to the distribution.
3700
3701 22. Added some untidy missing memory free() calls in pcretest, to keep
3702 Electric Fence happy when testing.
3703
3704
3705
3706 Version 5.0 13-Sep-04
3707 ---------------------
3708
3709 1. Internal change: literal characters are no longer packed up into items
3710 containing multiple characters in a single byte-string. Each character
3711 is now matched using a separate opcode. However, there may be more than one
3712 byte in the character in UTF-8 mode.
3713
3714 2. The pcre_callout_block structure has two new fields: pattern_position and
3715 next_item_length. These contain the offset in the pattern to the next match
3716 item, and its length, respectively.
3717
3718 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3719 insertion of callouts before each pattern item. Added the /C option to
3720 pcretest to make use of this.
3721
3722 4. On the advice of a Windows user, the lines
3723
3724 #if defined(_WIN32) || defined(WIN32)
3725 _setmode( _fileno( stdout ), 0x8000 );
3726 #endif /* defined(_WIN32) || defined(WIN32) */
3727
3728 have been added to the source of pcretest. This apparently does useful
3729 magic in relation to line terminators.
3730
3731 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3732 for the benefit of those environments where the "b" makes a difference.
3733
3734 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3735 to know about it. I have put a hack into configure.in that adds in code
3736 to set GCC=yes if CC=icc. This seems to end up at a point in the
3737 generated configure script that is early enough to affect the setting of
3738 compiler options, which is what is needed, but I have no means of testing
3739 whether it really works. (The user who reported this had patched the
3740 generated configure script, which of course I cannot do.)
3741
3742 LATER: After change 22 below (new libtool files), the configure script
3743 seems to know about icc (and also ecc). Therefore, I have commented out
3744 this hack in configure.in.
3745
3746 7. Added support for pkg-config (2 patches were sent in).
3747
3748 8. Negated POSIX character classes that used a combination of internal tables
3749 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3750 [[:^ascii]]. Typically, they would match almost any characters. The other
3751 POSIX classes were not broken in this way.
3752
3753 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3754 to find the match, as PCRE was deluded into thinking that the match had to
3755 start at the start point or following a newline. The same bug applied to
3756 patterns with negative forward assertions or any backward assertions
3757 preceding ".*" at the start, unless the pattern required a fixed first
3758 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3759
3760 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3761 starting at the last subject character, bytes beyond the end of the subject
3762 string were read.
3763
3764 11. Renamed the variable "class" as "classbits" to make life easier for C++
3765 users. (Previously there was a macro definition, but it apparently wasn't
3766 enough.)
3767
3768 12. Added the new field "tables" to the extra data so that tables can be passed
3769 in at exec time, or the internal tables can be re-selected. This allows
3770 a compiled regex to be saved and re-used at a later time by a different
3771 program that might have everything at different addresses.
3772
3773 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3774 -R library as well as a -L library.
3775
3776 14. The debugging options of pcretest (-d on the command line or D on a
3777 pattern) showed incorrect output for anything following an extended class
3778 that contained multibyte characters and which was followed by a quantifier.
3779
3780 15. Added optional support for general category Unicode character properties
3781 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3782 support. It adds about 90K to the size of the library. The meanings of the
3783 inbuilt class escapes such as \d and \s have NOT been changed.
3784
3785 16. Updated pcredemo.c to include calls to free() to release the memory for the
3786 compiled pattern.
3787
3788 17. The generated file chartables.c was being created in the source directory
3789 instead of in the building directory. This caused the build to fail if the
3790 source directory was different from the building directory, and was
3791 read-only.
3792
3793 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3794 file. No doubt somebody will tell me if they don't make sense... Also added
3795 Dan Mooney's comments about building on OpenVMS.
3796
3797 19. Added support for partial matching via the PCRE_PARTIAL option for
3798 pcre_exec() and the \P data escape in pcretest.
3799
3800 20. Extended pcretest with 3 new pattern features:
3801
3802 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3803 write the compiled pattern to the file whose name is "rest-of-line".
3804 This is a straight binary dump of the data, with the saved pointer to
3805 the character tables forced to be NULL. The study data, if any, is
3806 written too. After writing, pcretest reads a new pattern.
3807
3808 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3809 compiled pattern from the given file. There must not be any
3810 occurrences of "<" in the file name (pretty unlikely); if there are,
3811 pcretest will instead treat the initial "<" as a pattern delimiter.
3812 After reading in the pattern, pcretest goes on to read data lines as
3813 usual.
3814
3815 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3816 and 16-bit fields in a compiled pattern, to simulate a pattern that
3817 was compiled on a host of opposite endianness.
3818
3819 21. The pcre-exec() function can now cope with patterns that were compiled on
3820 hosts of opposite endianness, with this restriction:
3821
3822 As for any compiled expression that is saved and used later, the tables
3823 pointer field cannot be preserved; the extra_data field in the arguments
3824 to pcre_exec() should be used to pass in a tables address if a value
3825 other than the default internal tables were used at compile time.
3826
3827 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3828 now diagnosed as an error. Previously, most of the time, a negative number
3829 would have been treated as zero, but if in addition "ovector" was passed as
3830 NULL, a crash could occur.
3831
3832 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3833 new versions from the libtool 1.5 distribution (the last one is a copy of
3834 a file called libtool.m4). This seems to have fixed the need to patch
3835 "configure" to support Darwin 1.3 (which I used to do). However, I still
3836 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3837 workstation).
3838
3839 24. Changed the PCRE licence to be the more standard "BSD" licence.
3840
3841
3842 Version 4.5 01-Dec-03
3843 ---------------------
3844
3845 1. There has been some re-arrangement of the code for the match() function so
3846 that it can be compiled in a version that does not call itself recursively.
3847 Instead, it keeps those local variables that need separate instances for
3848 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3849 needs to "recurse". Keeping track of where control must go is done by means
3850 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3851 hide most of the details from the main code, and operates only if
3852 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3853 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3854 operating.
3855
3856 To make it easier for callers to provide specially tailored get/free
3857 functions for this usage, two new functions, pcre_stack_malloc, and
3858 pcre_stack_free, are used. They are always called in strict stacking order,
3859 and the size of block requested is always the same.
3860
3861 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3862 PCRE has been compiled to use the stack or the heap for recursion. The
3863 -C option of pcretest uses this to show which version is compiled.
3864
3865 A new data escape \S, is added to pcretest; it causes the amounts of store
3866 obtained and freed by both kinds of malloc/free at match time to be added
3867 to the output.
3868
3869 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3870 what's available on my current Linux desktop machine.
3871
3872 3. When matching a UTF-8 string, the test for a valid string at the start has
3873 been extended. If start_offset is not zero, PCRE now checks that it points
3874 to a byte that is the start of a UTF-8 character. If not, it returns
3875 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3876 this is necessary because there may be backward assertions in the pattern.
3877 When matching the same subject several times, it may save resources to use
3878 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3879
3880 4. The code for checking the validity of UTF-8 strings has been tightened so
3881 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3882 containing "overlong sequences".
3883
3884 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3885 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3886 should have been "&", but it just so happened that all the cases this let
3887 through by mistake were picked up later in the function.
3888
3889 6. I had used a variable called "isblank" - this is a C99 function, causing
3890 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3891
3892 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3893 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3894 so the version is shown; (c) stop "make test" echoing "./RunTest".
3895
3896 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3897
3898 9. The prototype for memmove() for systems that don't have it was using
3899 size_t, but the inclusion of the header that defines size_t was later. I've
3900 moved the #includes for the C headers earlier to avoid this.
3901
3902 10. Added some adjustments to the code to make it easier to compiler on certain
3903 special systems:
3904
3905 (a) Some "const" qualifiers were missing.
3906 (b) Added the macro EXPORT before all exported functions; by default this
3907 is defined to be empty.
3908 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3909 that it reads its output file name as an argument instead of writing
3910 to the standard output and assuming this can be redirected.
3911
3912 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3913 class containing characters with values greater than 255, PCRE compilation
3914 went into a loop.
3915
3916 12. A recursive reference to a subpattern that was within another subpattern
3917 that had a minimum quantifier of zero caused PCRE to crash. For example,
3918 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3919 recursion. If the recursively-called subpattern itself had a zero repeat,
3920 that was OK.
3921
3922 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3923 buffer into which it was copied (for escape processing) was still set at
3924 1024, so long lines caused crashes.
3925
3926 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3927 "internal error: code overflow...". This applied to any character class
3928 that was followed by a possessive quantifier.
3929
3930 15. Modified the Makefile to add libpcre.la as a prerequisite for
3931 libpcreposix.la because I was told this is needed for a parallel build to
3932 work.
3933
3934 16. If a pattern that contained .* following optional items at the start was
3935 studied, the wrong optimizing data was generated, leading to matching
3936 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3937 matching string must start with a or b or c. The correct conclusion for
3938 this pattern is that a match can start with any character.
3939
3940
3941 Version 4.4 13-Aug-03
3942 ---------------------
3943
3944 1. In UTF-8 mode, a character class containing characters with values between
3945 127 and 255 was not handled correctly if the compiled pattern was studied.
3946 In fixing this, I have also improved the studying algorithm for such
3947 classes (slightly).
3948
3949 2. Three internal functions had redundant arguments passed to them. Removal
3950 might give a very teeny performance improvement.
3951
3952 3. Documentation bug: the value of the capture_top field in a callout is *one
3953 more than* the number of the hightest numbered captured substring.
3954
3955 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3956 in incorrectly linking with a previously installed version. They now link
3957 explicitly with libpcre.la.
3958
3959 5. configure.in no longer needs to recognize Cygwin specially.
3960
3961 6. A problem in pcre.in for Windows platforms is fixed.
3962
3963 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3964 pcretest, it used to include the size of the study block as part of its
3965 output. Unfortunately, the structure contains a field that has a different
3966 size on different hardware architectures. This meant that the tests that
3967 showed this size failed. As the block is currently always of a fixed size,
3968 this information isn't actually particularly useful in pcretest output, so
3969 I have just removed it.
3970
3971 8. Three pre-processor statements accidentally did not start in column 1.
3972 Sadly, there are *still* compilers around that complain, even though
3973 standard C has not required this for well over a decade. Sigh.
3974
3975 9. In pcretest, the code for checking callouts passed small integers in the
3976 callout_data field, which is a void * field. However, some picky compilers
3977 complained about the casts involved for this on 64-bit systems. Now
3978 pcretest passes the address of the small integer instead, which should get
3979 rid of the warnings.
3980
3981 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3982 both compile and run time, and gives an error if an invalid UTF-8 sequence
3983 is found. There is a option for disabling this check in cases where the
3984 string is known to be correct and/or the maximum performance is wanted.
3985
3986 11. In response to a bug report, I changed one line in Makefile.in from
3987
3988 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3989 to
3990 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3991
3992 to look similar to other lines, but I have no way of telling whether this
3993 is the right thing to do, as I do not use Windows. No doubt I'll get told
3994 if it's wrong...
3995
3996
3997 Version 4.3 21-May-03
3998 ---------------------
3999
4000 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
4001 Makefile.
4002
4003 2. Some refactoring to improve the quality of the code:
4004
4005 (i) The utf8_table... variables are now declared "const".
4006
4007 (ii) The code for \cx, which used the "case flipping" table to upper case
4008 lower case letters, now just substracts 32. This is ASCII-specific,
4009 but the whole concept of \cx is ASCII-specific, so it seems
4010 reasonable.
4011
4012 (iii) PCRE was using its character types table to recognize decimal and
4013 hexadecimal digits in the pattern. This is silly, because it handles
4014 only 0-9, a-f, and A-F, but the character types table is locale-
4015 specific, which means strange things might happen. A private
4016 table is now used for this - though it costs 256 bytes, a table is
4017 much faster than multiple explicit tests. Of course, the standard
4018 character types table is still used for matching digits in subject
4019 strings against \d.
4020
4021 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
4022 ending in _t are). So I've renamed it as ESC_tee.
4023
4024 3. The first argument for regexec() in the POSIX wrapper should have been
4025 defined as "const".
4026
4027 4. Changed pcretest to use malloc() for its buffers so that they can be
4028 Electric Fenced for debugging.
4029
4030 5. There were several places in the code where, in UTF-8 mode, PCRE would try
4031 to read one or more bytes before the start of the subject string. Often this
4032 had no effect on PCRE's behaviour, but in some circumstances it could
4033 provoke a segmentation fault.
4034
4035 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
4036 to try to read one or more bytes before the start of the subject string.
4037
4038 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
4039 UTF-8 support could misbehave in various ways if the subject string
4040 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
4041 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
4042 back over UTF-8 characters.)
4043
4044
4045 Version 4.2 14-Apr-03
4046 ---------------------
4047
4048 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
4049
4050 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
4051 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
4052 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
4053 [WIN_PREFIX]: new variable, "cyg" for Cygwin
4054 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
4055 and BUILD_EXEEXT
4056 Note: automatic setting of the BUILD variables is not yet working
4057 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
4058 compile-time but not at link-time
4059 [LINK]: use for linking executables only
4060 make different versions for Windows and non-Windows
4061 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
4062 libraries
4063 [LINK_FOR_BUILD]: new variable
4064 [OBJEXT]: use throughout
4065 [EXEEXT]: use throughout
4066 <winshared>: new target
4067 <wininstall>: new target
4068 <dftables.o>: use native compiler
4069 <dftables>: use native linker
4070 <install>: handle Windows platform correctly
4071 <clean>: ditto
4072 <check>: ditto
4073 copy DLL to top builddir before testing
4074
4075 As part of these changes, -no-undefined was removed again. This was reported
4076 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
4077 in any case.
4078
4079 3. Some tidies to get rid of compiler warnings:
4080
4081 . In the match_data structure, match_limit was an unsigned long int, whereas
4082 match_call_count was an int. I've made them both unsigned long ints.
4083
4084 . In pcretest the fact that a const uschar * doesn't automatically cast to
4085 a void * provoked a warning.
4086
4087 . Turning on some more compiler warnings threw up some "shadow" variables
4088 and a few more missing casts.
4089
4090 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
4091 option, a class that contained a single character with a value between 128
4092 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
4093
4094 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
4095 option, a class that contained several characters, but with at least one
4096 whose value was between 128 and 255 caused PCRE to crash.
4097
4098
4099 Version 4.1 12-Mar-03
4100 ---------------------
4101
4102 1. Compiling with gcc -pedantic found a couple of places where casts were
4103 needed, and a string in dftables.c that was longer than standard compilers are
4104 required to support.
4105
4106 2. Compiling with Sun's compiler found a few more places where the code could
4107 be tidied up in order to avoid warnings.
4108
4109 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
4110 first of these names is deprecated in the latest Autoconf in favour of the name
4111 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
4112 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
4113 analogy I have changed it to CFLAGS_FOR_BUILD.
4114
4115 4. Added -no-undefined to the linking command in the Makefile, because this is
4116 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
4117 linking step for the pcreposix library.
4118
4119 5. PCRE was failing to diagnose the case of two named groups with the same
4120 name.
4121
4122 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
4123 literal character that is needed in the subject for a match, and scans along to
4124 ensure that it is present before embarking on the full matching process. This
4125 saves time in cases of nested unlimited repeats that are never going to match.
4126 Problem: the scan can take a lot of time if the subject is very long (e.g.
4127 megabytes), thus penalizing straightforward matches. It is now done only if the
4128 amount of subject to be scanned is less than 1000 bytes.
4129
4130 7. A lesser problem with the same optimization is that it was recording the
4131 first character of an anchored pattern as "needed", thus provoking a search
4132 right along the subject, even when the first match of the pattern was going to
4133 fail. The "needed" character is now not set for anchored patterns, unless it
4134 follows something in the pattern that is of non-fixed length. Thus, it still
4135 fulfils its original purpose of finding quick non-matches in cases of nested
4136 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
4137
4138
4139 Version 4.0 17-Feb-03
4140 ---------------------
4141
4142 1. If a comment in an extended regex that started immediately after a meta-item
4143 extended to the end of string, PCRE compiled incorrect data. This could lead to
4144 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
4145
4146 2. Moved to autoconf 2.53 and libtool 1.4.2.
4147
4148 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
4149 the special perltest8 script is no longer needed - all the tests can be run
4150 from a single perltest script.
4151
4152 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
4153 by \s. It has now been removed in PCRE. This means it isn't recognized as
4154 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
4155 class [:space:] *does* include VT, thereby creating a mess.
4156
4157 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
4158 space and tab.
4159
4160 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
4161 its new features into the main test script, reducing the number of scripts.
4162
4163 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
4164 were backward compatible, and made the (?i) apply to the whole pattern, as if
4165 /i were given. Now it behaves more logically, and applies the option setting
4166 only to what follows. PCRE has been changed to follow suit. However, if it
4167 finds options settings right at the start of the pattern, it extracts them into
4168 the global options, as before. Thus, they show up in the info data.
4169
4170 8. Added support for the \Q...\E escape sequence. Characters in between are
4171 treated as literals. This is slightly different from Perl in that $ and @ are
4172 also handled as literals inside the quotes. In Perl, they will cause variable
4173 interpolation. Note the following examples:
4174
4175 Pattern PCRE matches Perl matches
4176
4177 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
4178 \Qabc\$xyz\E abc\$xyz abc\$xyz
4179 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
4180
4181 For compatibility with Perl, \Q...\E sequences are recognized inside character
4182 classes as well as outside them.
4183
4184 9. Re-organized 3 code statements in pcretest to avoid "overflow in
4185 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
4186 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
4187 signed/unsigned warnings.
4188
4189 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
4190 option for pcretest, so I've replaced it by a simple function that does just
4191 that job.
4192
4193 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
4194 "pcregrep -".
4195
4196 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
4197 Java package. This provides some syntactic sugar for simple cases of what my
4198 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
4199 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
4200 item, you can use this simplified notation. Note that only makes sense with
4201 greedy quantifiers. Consequently, the use of the possessive quantifier forces
4202 greediness, whatever the setting of the PCRE_UNGREEDY option.
4203
4204 13. A change of greediness default within a pattern was not taking effect at
4205 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
4206 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
4207 was abstracted outside.
4208
4209 14. PCRE now supports the \G assertion. It is true when the current matching
4210 position is at the start point of the match. This differs from \A when the
4211 starting offset is non-zero. Used with the /g option of pcretest (or similar
4212 code), it works in the same way as it does for Perl's /g option. If all
4213 alternatives of a regex begin with \G, the expression is anchored to the start
4214 match position, and the "anchored" flag is set in the compiled expression.
4215
4216 15. Some bugs concerning the handling of certain option changes within patterns
4217 have been fixed. These applied to options other than (?ims). For example,
4218 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
4219 been the other way round. Some of this was related to change 7 above.
4220
4221 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
4222 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
4223 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
4224 POSIX classes only within a class (e.g. /[[:alpha:]]/).
4225
4226 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
4227 mode. Unlike ".", it always matches newline, whatever the setting of
4228 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
4229 assertions. Perl allows it, but it doesn't (in general) work because it can't
4230 calculate the length of the lookbehind. At least, that's the case for Perl
4231 5.8.0 - I've been told they are going to document that it doesn't work in
4232 future.
4233
4234 18. Added an error diagnosis for escapes that PCRE does not support: these are
4235 \L, \l, \N, \P, \p, \U, \u, and \X.
4236
4237 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
4238 reading past the end of the pattern in cases such as /[abcd/.
4239
4240 20. PCRE was getting more memory than necessary for patterns with classes that
4241 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
4242
4243 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
4244 compiling PCRE for use with Virtual Pascal.
4245
4246 22. Small fix to the Makefile to make it work properly if the build is done
4247 outside the source tree.
4248
4249 23. Added a new extension: a condition to go with recursion. If a conditional
4250 subpattern starts with (?(R) the "true" branch is used if recursion has
4251 happened, whereas the "false" branch is used only at the top level.
4252
4253 24. When there was a very long string of literal characters (over 255 bytes
4254 without UTF support, over 250 bytes with UTF support), the computation of how
4255 much memory was required could be incorrect, leading to segfaults or other
4256 strange effects.
4257
4258 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
4259 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
4260 there was a subsequent back reference to those brackets. This meant that, for
4261 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
4262 possible to check for precisely this case. All we can do is abandon the
4263 optimization if .* occurs inside capturing brackets when there are any back
4264 references whatsoever. (See below for a better fix that came later.)
4265
4266 26. The handling of the optimization for finding the first character of a
4267 non-anchored pattern, and for finding a character that is required later in the
4268 match were failing in some cases. This didn't break the matching; it just
4269 failed to optimize when it could. The way this is done has been re-implemented.
4270
4271 27. Fixed typo in error message for invalid (?R item (it said "(?p").
4272
4273 28. Added a new feature that provides some of the functionality that Perl
4274 provides with (?{...}). The facility is termed a "callout". The way it is done
4275 in PCRE is for the caller to provide an optional function, by setting
4276 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
4277 global variable. By default it is unset, which disables all calling out. To get
4278 the function called, the regex must include (?C) at appropriate points. This
4279 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
4280 This provides a means of identifying different callout points. When PCRE
4281 reaches such a point in the regex, if pcre_callout has been set, the external
4282 function is called. It is provided with data in a structure called
4283 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
4284 matching continues; if it returns a non-zero value, the match at the current
4285 point fails. However, backtracking will occur if possible. [This was changed
4286 later and other features added - see item 49 below.]
4287
4288 29. pcretest is upgraded to test the callout functionality. It provides a
4289 callout function that displays information. By default, it shows the start of
4290 the match and the current position in the text. There are some new data escapes
4291 to vary what happens:
4292
4293 \C+ in addition, show current contents of captured substrings
4294 \C- do not supply a callout function
4295 \C!n return 1 when callout number n is reached
4296 \C!n!m return 1 when callout number n is reached for the mth time
4297
4298 30. If pcregrep was called with the -l option and just a single file name, it
4299 output "<stdin>" if a match was found, instead of the file name.
4300
4301 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
4302 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
4303 pcre_exec(). This saves a malloc/free per call. The default value of
4304 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
4305 when configuring.
4306
4307 32. The default maximum size of a compiled pattern is 64K. There have been a
4308 few cases of people hitting this limit. The code now uses macros to handle the
4309 storing of links as offsets within the compiled pattern. It defaults to 2-byte
4310 links, but this can be changed to 3 or 4 bytes by --with-link-size when
4311 configuring. Tests 2 and 5 work only with 2-byte links because they output
4312 debugging information about compiled patterns.
4313
4314 33. Internal code re-arrangements:
4315
4316 (a) Moved the debugging function for printing out a compiled regex into
4317 its own source file (printint.c) and used #include to pull it into
4318 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
4319 separate copies.
4320
4321 (b) Defined the list of op-code names for debugging as a macro in
4322 internal.h so that it is next to the definition of the opcodes.
4323
4324 (c) Defined a table of op-code lengths for simpler skipping along compiled
4325 code. This is again a macro in internal.h so that it is next to the
4326 definition of the opcodes.
4327
4328 34. Added support for recursive calls to individual subpatterns, along the
4329 lines of Robin Houston's patch (but implemented somewhat differently).
4330
4331 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
4332 allow it to read and process whole directories in Win32. This code was
4333 contributed by Lionel Fourquaux; it has not been tested by me.
4334
4335 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
4336 used to name a group. Names consist of alphanumerics and underscores, and must
4337 be unique. Back references use the syntax (?P=name) and recursive calls use
4338 (?P>name) which is a PCRE extension to the Python extension. Groups still have
4339 numbers. The function pcre_fullinfo() can be used after compilation to extract
4340 a name/number map. There are three relevant calls:
4341
4342 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
4343 PCRE_INFO_NAMECOUNT yields the number of entries
4344 PCRE_INFO_NAMETABLE yields a pointer to the map.
4345
4346 The map is a vector of fixed-size entries. The size of each entry depends on
4347 the length of the longest name used. The first two bytes of each entry are the
4348 group number, most significant byte first. There follows the corresponding
4349 name, zero terminated. The names are in alphabetical order.
4350
4351 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
4352 case instead of 255. Making it the same both with and without UTF-8 support
4353 means that the same test output works with both.
4354
4355 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
4356 calling malloc() with a zero argument.
4357
4358 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
4359 optimization. I've improved things by keeping a bitmap of backreferences with
4360 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
4361 fact referenced, the optimization can be applied. It is unlikely that a
4362 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
4363 the match to follow \n) will appear inside brackets with a number greater than
4364 31, but if it does, any back reference > 31 suppresses the optimization.
4365
4366 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
4367 of disabling numbered capturing parentheses. Any opening parenthesis that is
4368 not followed by ? behaves as if it were followed by ?: but named parentheses
4369 can still be used for capturing (and they will acquire numbers in the usual
4370 way).
4371
4372 41. Redesigned the return codes from the match() function into yes/no/error so
4373 that errors can be passed back from deep inside the nested calls. A malloc
4374 failure while inside a recursive subpattern call now causes the
4375 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
4376
4377 42. It is now possible to set a limit on the number of times the match()
4378 function is called in a call to pcre_exec(). This facility makes it possible to
4379 limit the amount of recursion and backtracking, though not in a directly
4380 obvious way, because the match() function is used in a number of different
4381 circumstances. The count starts from zero for each position in the subject
4382 string (for non-anchored patterns). The default limit is, for compatibility, a
4383 large number, namely 10 000 000. You can change this in two ways:
4384
4385 (a) When configuring PCRE before making, you can use --with-match-limit=n
4386 to set a default value for the compiled library.
4387
4388 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
4389 a different value is set. See 45 below.
4390
4391 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
4392
4393 43. Added a new function pcre_config(int, void *) to enable run-time extraction
4394 of things that can be changed at compile time. The first argument specifies
4395 what is wanted and the second points to where the information is to be placed.
4396 The current list of available information is:
4397
4398 PCRE_CONFIG_UTF8
4399
4400 The output is an integer that is set to one if UTF-8 support is available;
4401 otherwise it is set to zero.
4402
4403 PCRE_CONFIG_NEWLINE
4404
4405 The output is an integer that it set to the value of the code that is used for
4406 newline. It is either LF (10) or CR (13).
4407
4408 PCRE_CONFIG_LINK_SIZE
4409
4410 The output is an integer that contains the number of bytes used for internal
4411 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
4412
4413 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
4414
4415 The output is an integer that contains the threshold above which the POSIX
4416 interface uses malloc() for output vectors. See item 31 above.
4417
4418 PCRE_CONFIG_MATCH_LIMIT
4419
4420 The output is an unsigned integer that contains the default limit of the number
4421 of match() calls in a pcre_exec() execution. See 42 above.
4422
4423 44. pcretest has been upgraded by the addition of the -C option. This causes it
4424 to extract all the available output from the new pcre_config() function, and to
4425 output it. The program then exits immediately.
4426
4427 45. A need has arisen to pass over additional data with calls to pcre_exec() in
4428 order to support additional features. One way would have been to define
4429 pcre_exec2() (for example) with extra arguments, but this would not have been
4430 extensible, and would also have required all calls to the original function to
4431 be mapped to the new one. Instead, I have chosen to extend the mechanism that
4432 is used for passing in "extra" data from pcre_study().
4433
4434 The pcre_extra structure is now exposed and defined in pcre.h. It currently
4435 contains the following fields:
4436
4437 flags a bitmap indicating which of the following fields are set
4438 study_data opaque data from pcre_study()
4439 match_limit a way of specifying a limit on match() calls for a specific
4440 call to pcre_exec()
4441 callout_data data for callouts (see 49 below)
4442
4443 The flag bits are also defined in pcre.h, and are
4444
4445 PCRE_EXTRA_STUDY_DATA
4446 PCRE_EXTRA_MATCH_LIMIT
4447 PCRE_EXTRA_CALLOUT_DATA
4448
4449 The pcre_study() function now returns one of these new pcre_extra blocks, with
4450 the actual study data pointed to by the study_data field, and the
4451 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
4452 before. That is, this change is entirely upwards-compatible and requires no
4453 change to existing code.
4454
4455 If you want to pass in additional data to pcre_exec(), you can either place it
4456 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
4457 block.
4458
4459 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
4460 data string contains the escape sequence \M, pcretest calls pcre_exec() several
4461 times with different match limits, until it finds the minimum value needed for
4462 pcre_exec() to complete. The value is then output. This can be instructive; for
4463 most simple matches the number is quite small, but for pathological cases it
4464 gets very large very quickly.
4465
4466 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
4467 returns the size of the data block pointed to by the study_data field in a
4468 pcre_extra block, that is, the value that was passed as the argument to
4469 pcre_malloc() when PCRE was getting memory in which to place the information
4470 created by pcre_study(). The fourth argument should point to a size_t variable.
4471 pcretest has been extended so that this information is shown after a successful
4472 pcre_study() call when information about the compiled regex is being displayed.
4473
4474 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
4475 because what follows is always an absolute path. (Later: it turns out that this
4476 is more than cosmetic for MinGW, because it doesn't like empty path
4477 components.)
4478
4479 49. Some changes have been made to the callout feature (see 28 above):
4480
4481 (i) A callout function now has three choices for what it returns:
4482
4483 0 => success, carry on matching
4484 > 0 => failure at this point, but backtrack if possible
4485 < 0 => serious error, return this value from pcre_exec()
4486
4487 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
4488 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
4489 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
4490 use by callout functions. It will never be used by PCRE itself.
4491
4492 (ii) The pcre_extra structure (see 45 above) has a void * field called
4493 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
4494 pcre_callout_block structure has a field of the same name. The contents of
4495 the field passed in the pcre_extra structure are passed to the callout
4496 function in the corresponding field in the callout block. This makes it
4497 easier to use the same callout-containing regex from multiple threads. For
4498 testing, the pcretest program has a new data escape
4499
4500 \C*n pass the number n (may be negative) as callout_data
4501
4502 If the callout function in pcretest receives a non-zero value as
4503 callout_data, it returns that value.
4504
4505 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
4506 there were some redundant $(CFLAGS) in commands that are now specified as
4507 $(LINK), which already includes $(CFLAGS).
4508
4509 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
4510 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
4511 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
4512 one-byte characters throughout. Note that case-insensitive matching applies
4513 only to characters whose values are less than 256. PCRE doesn't support the
4514 notion of cases for higher-valued characters.
4515
4516 (i) A character class whose characters are all within 0-255 is handled as
4517 a bit map, and the map is inverted for negative classes. Previously, a
4518 character > 255 always failed to match such a class; however it should
4519 match if the class was a negative one (e.g. [^ab]). This has been fixed.
4520
4521 (ii) A negated character class with a single character < 255 is coded as
4522 "not this character" (OP_NOT). This wasn't working properly when the test
4523 character was multibyte, either singly or repeated.
4524
4525 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
4526 mode, for example: \x{100}{2,3}.
4527
4528 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
4529 singly or repeated) now correctly test multibyte characters. However,
4530 PCRE doesn't recognize any characters with values greater than 255 as
4531 digits, spaces, or word characters. Such characters always match \D, \S,
4532 and \W, and never match \d, \s, or \w.
4533
4534 (v) Classes may now contain characters and character ranges with values
4535 greater than 255. For example: [ab\x{100}-\x{400}].
4536
4537 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
4538 PCRE in UTF-8 mode.
4539
4540 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
4541 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
4542 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
4543 value.)
4544
4545 53. The single man page has become too large. I have therefore split it up into
4546 a number of separate man pages. These also give rise to individual HTML pages;
4547 these are now put in a separate directory, and there is an index.html page that
4548 lists them all. Some hyperlinking between the pages has been installed.
4549
4550 54. Added convenience functions for handling named capturing parentheses.
4551
4552 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4553 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4554 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4555 are faulted.
4556
4557 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4558 calling configure. These values are used when compiling the dftables.c program
4559 which is run to generate the source of the default character tables. They
4560 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4561 you will need to set these values.
4562
4563 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4564
4565
4566 Version 3.9 02-Jan-02
4567 ---------------------
4568
4569 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4570
4571 2. If --disable-static was given, the building process failed when trying to
4572 build pcretest and pcregrep. (For some reason it was using libtool to compile
4573 them, which is not right, as they aren't part of the library.)
4574
4575
4576 Version 3.8 18-Dec-01
4577 ---------------------
4578
4579 1. The experimental UTF-8 code was completely screwed up. It was packing the
4580 bytes in the wrong order. How dumb can you get?
4581
4582
4583 Version 3.7 29-Oct-01
4584 ---------------------
4585
4586 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4587 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4588 this didn't happen under Solaris 8, where I normally test things.
4589
4590 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4591 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4592 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4593 this fix an hour or so after the initial 3.7 release.)
4594
4595
4596 Version 3.6 23-Oct-01
4597 ---------------------
4598
4599 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4600 offsets passed as NULL with zero offset count.
4601
4602 2. The config.guess and config.sub files had not been updated when I moved to
4603 the latest autoconf.
4604
4605
4606 Version 3.5 15-Aug-01
4607 ---------------------
4608
4609 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4610 had been forgotten.
4611
4612 2. By using declared but undefined structures, we can avoid using "void"
4613 definitions in pcre.h while keeping the internal definitions of the structures
4614 private.
4615
4616 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4617 user point of view, this means that both static and shared libraries are built
4618 by default, but this can be individually controlled. More of the work of
4619 handling this static/shared cases is now inside libtool instead of PCRE's make
4620 file.
4621
4622 4. The pcretest utility is now installed along with pcregrep because it is
4623 useful for users (to test regexs) and by doing this, it automatically gets
4624 relinked by libtool. The documentation has been turned into a man page, so
4625 there are now .1, .txt, and .html versions in /doc.
4626
4627 5. Upgrades to pcregrep:
4628 (i) Added long-form option names like gnu grep.
4629 (ii) Added --help to list all options with an explanatory phrase.
4630 (iii) Added -r, --recursive to recurse into sub-directories.
4631 (iv) Added -f, --file to read patterns from a file.
4632
4633 6. pcre_exec() was referring to its "code" argument before testing that
4634 argument for NULL (and giving an error if it was NULL).
4635
4636 7. Upgraded Makefile.in to allow for compiling in a different directory from
4637 the source directory.
4638
4639 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4640 options bits, the pointer it was passed was to an int instead of to an unsigned
4641 long int. This mattered only on 64-bit systems.
4642
4643 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4644 generated) instead of pcre.in, which it its source. Also made the same change
4645 in several of the .c files.
4646
4647 10. A new release of gcc defines printf() as a macro, which broke pcretest
4648 because it had an ifdef in the middle of a string argument for printf(). Fixed
4649 by using separate calls to printf().
4650
4651 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4652 script, to force use of CR or LF instead of \n in the source. On non-Unix
4653 systems, the value can be set in config.h.
4654
4655 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4656 absolute limit. Changed the text of the error message to make this clear, and
4657 likewise updated the man page.
4658
4659 13. The limit of 99 on the number of capturing subpatterns has been removed.
4660 The new limit is 65535, which I hope will not be a "real" limit.
4661
4662
4663 Version 3.4 22-Aug-00
4664 ---------------------
4665
4666 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4667
4668 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4669
4670
4671 Version 3.3 01-Aug-00
4672 ---------------------
4673
4674 1. If an octal character was given, but the value was greater than \377, it
4675 was not getting masked to the least significant bits, as documented. This could
4676 lead to crashes in some systems.
4677
4678 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4679 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4680
4681 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4682 These just pass their arguments on to (pcre_free)(), but they are provided
4683 because some uses of PCRE bind it to non-C systems that can call its functions,
4684 but cannot call free() or pcre_free() directly.
4685
4686 4. Add "make test" as a synonym for "make check". Corrected some comments in
4687 the Makefile.
4688
4689 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4690 Makefile.
4691
4692 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4693 command called pgrep for grepping around the active processes.
4694
4695 7. Added the beginnings of support for UTF-8 character strings.
4696
4697 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4698 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4699 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4700 out for the ar command.)
4701
4702
4703 Version 3.2 12-May-00
4704 ---------------------
4705
4706 This is purely a bug fixing release.
4707
4708 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4709 of ZA. This was just one example of several cases that could provoke this bug,
4710 which was introduced by change 9 of version 2.00. The code for breaking
4711 infinite loops after an iteration that matches an empty string was't working
4712 correctly.
4713
4714 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4715 when matched against abbab (for example). After matching an empty string, it
4716 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4717 caused it to match further down the string than it should.
4718
4719 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4720 was there because it doesn't seem to be needed, and it causes trouble on some
4721 systems, as it is not a Standard C header. It has been removed.
4722
4723 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4724 were reported on the Macintosh. The changes were from
4725
4726 while ((c = *(++ptr)) != 0 && c != '\n');
4727 to
4728 while ((c = *(++ptr)) != 0 && c != '\n') ;
4729
4730 Totally extraordinary, but if that's what it takes...
4731
4732 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4733 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4734 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4735 assumes the way PCRE uses memmove() (always moving upwards).
4736
4737 6. PCRE is being used in one environment where strchr() is not available. There
4738 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4739 faster code anyway.
4740
4741
4742 Version 3.1 09-Feb-00
4743 ---------------------
4744
4745 The only change in this release is the fixing of some bugs in Makefile.in for
4746 the "install" target:
4747
4748 (1) It was failing to install pcreposix.h.
4749
4750 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4751
4752
4753 Version 3.0 01-Feb-00
4754 ---------------------
4755
4756 1. Add support for the /+ modifier to perltest (to output $` like it does in
4757 pcretest).
4758
4759 2. Add support for the /g modifier to perltest.
4760
4761 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4762 matches null strings.
4763
4764 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4765 pattern. Perl treats empty patterns specially - it reuses the most recent
4766 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4767 effect.
4768
4769 5. The POSIX interface was broken in that it was just handing over the POSIX
4770 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4771 required a bigger vector, with some working space on the end. This means that
4772 the POSIX wrapper now has to get and free some memory, and copy the results.
4773
4774 6. Added some simple autoconf support, placing the test data and the
4775 documentation in separate directories, re-organizing some of the
4776 information files, and making it build pcre-config (a GNU standard). Also added
4777 libtool support for building PCRE as a shared library, which is now the
4778 default.
4779
4780 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4781 09 are not valid octal constants. Single digits will be used for minor values
4782 less than 10.
4783
4784 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4785 existing programs that set these in the POSIX interface can use PCRE without
4786 modification.
4787
4788 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4789 return all that pcre_info() returns, plus additional data. The pcre_info()
4790 function is retained for compatibility, but is considered to be obsolete.
4791
4792 10. Added experimental recursion feature (?R) to handle one common case that
4793 Perl 5.6 will be able to do with (?p{...}).
4794
4795 11. Added support for POSIX character classes like [:alpha:], which Perl is
4796 adopting.
4797
4798
4799 Version 2.08 31-Aug-99
4800 ----------------------
4801
4802 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4803 trying to match at the startoffset position, but instead was moving forward to
4804 the next newline as if a previous match had failed.
4805
4806 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4807 and could get into a loop if a null string was matched other than at the start
4808 of the subject.
4809
4810 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4811 be distinguished at compile time, and for completeness also added PCRE_DATE.
4812
4813 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4814 in GnuWin32 environments.
4815
4816
4817 Version 2.07 29-Jul-99
4818 ----------------------
4819
4820 1. The documentation is now supplied in plain text form and HTML as well as in
4821 the form of man page sources.
4822
4823 2. C++ compilers don't like assigning (void *) values to other pointer types.
4824 In particular this affects malloc(). Although there is no problem in Standard
4825 C, I've put in casts to keep C++ compilers happy.
4826
4827 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4828 should be (const char *).
4829
4830 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4831 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4832 However, I haven't made this a standard facility. The documentation doesn't
4833 mention it, and the Makefile doesn't support it.
4834
4835 5. The Makefile now contains an "install" target, with editable destinations at
4836 the top of the file. The pcretest program is not installed.
4837
4838 6. pgrep -V now gives the PCRE version number and date.
4839
4840 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4841 causing the entire string to be ignored, instead of just the last character.
4842
4843 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4844 non-matching string, it can take a very, very long time, even for strings of
4845 quite modest length, because of the nested recursion. PCRE now does better in
4846 some of these cases. It does this by remembering the last required literal
4847 character in the pattern, and pre-searching the subject to ensure it is present
4848 before running the real match. In other words, it applies a heuristic to detect
4849 some types of certain failure quickly, and in the above example, if presented
4850 with a string that has no trailing " it gives "no match" very quickly.
4851
4852 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4853 other alternatives are tried instead.
4854
4855
4856 Version 2.06 09-Jun-99
4857 ----------------------
4858
4859 1. Change pcretest's output for amount of store used to show just the code
4860 space, because the remainder (the data block) varies in size between 32-bit and
4861 64-bit systems.
4862
4863 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4864 start matching at. This allows lookbehinds to work when searching for multiple
4865 occurrences in a string.
4866
4867 3. Added additional options to pcretest for testing multiple occurrences:
4868
4869 /+ outputs the rest of the string that follows a match
4870 /g loops for multiple occurrences, using the new startoffset argument
4871 /G loops for multiple occurrences by passing an incremented pointer
4872
4873 4. PCRE wasn't doing the "first character" optimization for patterns starting
4874 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4875 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4876 the letter 'x'. On long subject strings, this gives a significant speed-up.
4877
4878
4879 Version 2.05 21-Apr-99
4880 ----------------------
4881
4882 1. Changed the type of magic_number from int to long int so that it works
4883 properly on 16-bit systems.
4884
4885 2. Fixed a bug which caused patterns starting with .* not to work correctly
4886 when the subject string contained newline characters. PCRE was assuming
4887 anchoring for such patterns in all cases, which is not correct because .* will
4888 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4889 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4890 must be retried after every newline in the subject.
4891
4892
4893 Version 2.04 18-Feb-99
4894 ----------------------
4895
4896 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4897 computation of the store needed to hold the pattern was incorrect (too large).
4898 If such patterns were nested a few deep, this could multiply and become a real
4899 problem.
4900
4901 2. Added /M option to pcretest to show the memory requirement of a specific
4902 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4903
4904 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4905 compiled in such a way that the backtracking after subsequent failure was
4906 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4907 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4908
4909
4910 Version 2.03 02-Feb-99
4911 ----------------------
4912
4913 1. Fixed typo and small mistake in man page.
4914
4915 2. Added 4th condition (GPL supersedes if conflict) and created separate
4916 LICENCE file containing the conditions.
4917
4918 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4919 Perl, that is the internal \ allows the delimiter to be included in the
4920 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4921 the final delimiter, add \ to the end of the pattern (to test the error).
4922
4923 4. Added the convenience functions for extracting substrings after a successful
4924 match. Updated pcretest to make it able to test these functions.
4925
4926
4927 Version 2.02 14-Jan-99
4928 ----------------------
4929
4930 1. Initialized the working variables associated with each extraction so that
4931 their saving and restoring doesn't refer to uninitialized store.
4932
4933 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4934 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4935 fix the problem.
4936
4937 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4938 calls, and wasn't printing the correct value for compiling calls. Increased the
4939 default value of LOOPREPEAT, and the number of significant figures in the
4940 times.
4941
4942 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4943
4944 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4945 a building problem on Windows NT with a FAT file system.
4946
4947
4948 Version 2.01 21-Oct-98
4949 ----------------------
4950
4951 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4952 to character tables built by pcre_maketables() in the current locale. If NULL
4953 is passed, the default tables are used.
4954
4955
4956 Version 2.00 24-Sep-98
4957 ----------------------
4958
4959 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4960 it any more.
4961
4962 2. Allow quantification of (?>) groups, and make it work correctly.
4963
4964 3. The first character computation wasn't working for (?>) groups.
4965
4966 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4967 end of the subject) and add 5.005's \z, which really does match only at the
4968 very end of the subject.
4969
4970 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4971
4972 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4973 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4974 localized options. All options to pcre_study() were also removed.
4975
4976 7. Add other new features from 5.005:
4977
4978 $(?<= positive lookbehind
4979 $(?<! negative lookbehind
4980 (?imsx-imsx) added the unsetting capability
4981 such a setting is global if at outer level; local otherwise
4982 (?imsx-imsx:) non-capturing groups with option setting
4983 (?(cond)re|re) conditional pattern matching
4984
4985 A backreference to itself in a repeated group matches the previous
4986 captured string.
4987
4988 8. General tidying up of studying (both automatic and via "study")
4989 consequential on the addition of new assertions.
4990
4991 9. As in 5.005, unlimited repeated groups that could match an empty substring
4992 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4993 runtime if any iteration does actually match an empty substring.
4994
4995 10. Include the RunTest script in the distribution.
4996
4997 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4998 discrepancies, some of which were old and were also with respect to 5.004. They
4999 have now been fixed.
5000
5001
5002 Version 1.09 28-Apr-98
5003 ----------------------
5004
5005 1. A negated single character class followed by a quantifier with a minimum
5006 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
5007 program crashes, or just wrong answers. This did not apply to negated classes
5008 containing more than one character, or to minima other than one.
5009
5010
5011 Version 1.08 27-Mar-98
5012 ----------------------
5013
5014 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
5015
5016 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
5017 latter must appear before anything that relies on it in the pattern.
5018
5019
5020 Version 1.07 16-Feb-98
5021 ----------------------
5022
5023 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
5024 repeat of a potentially empty string).
5025
5026
5027 Version 1.06 23-Jan-98
5028 ----------------------
5029
5030 1. Added Markus Oberhumer's little patches for C++.
5031
5032 2. Literal strings longer than 255 characters were broken.
5033
5034
5035 Version 1.05 23-Dec-97
5036 ----------------------
5037
5038 1. Negated character classes containing more than one character were failing if
5039 PCRE_CASELESS was set at run time.
5040
5041
5042 Version 1.04 19-Dec-97
5043 ----------------------
5044
5045 1. Corrected the man page, where some "const" qualifiers had been omitted.
5046
5047 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
5048 input syntax.
5049
5050 3. Fixed memory leak which occurred when a regex with back references was
5051 matched with an offsets vector that wasn't big enough. The temporary memory
5052 that is used in this case wasn't being freed if the match failed.
5053
5054 4. Tidied pcretest to ensure it frees memory that it gets.
5055
5056 5. Temporary memory was being obtained in the case where the passed offsets
5057 vector was exactly big enough.
5058
5059 6. Corrected definition of offsetof() from change 5 below.
5060
5061 7. I had screwed up change 6 below and broken the rules for the use of
5062 setjmp(). Now fixed.
5063
5064
5065 Version 1.03 18-Dec-97
5066 ----------------------
5067
5068 1. A erroneous regex with a missing opening parenthesis was correctly
5069 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
5070 on some systems.
5071
5072 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
5073 it was reported that one broken compiler failed on the former because "code" is
5074 also an independent variable.
5075
5076 3. The erroneous regex a[]b caused an array overrun reference.
5077
5078 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
5079 fail on data ending with that character. (It was going on too far, and checking
5080 the next character, typically a binary zero.) This was specific to the
5081 optimized code for single-character negative classes.
5082
5083 5. Added a contributed patch from the TIN world which does the following:
5084
5085 + Add an undef for memmove, in case the the system defines a macro for it.
5086
5087 + Add a definition of offsetof(), in case there isn't one. (I don't know
5088 the reason behind this - offsetof() is part of the ANSI standard - but
5089 it does no harm).
5090
5091 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
5092 most of the places where whitespace preceded '#'. I have given up and
5093 allowed the remaining 2 cases to be at the margin.
5094
5095 + Rename some variables in pcre to eliminate shadowing. This seems very
5096 pedantic, but does no harm, of course.
5097
5098 6. Moved the call to setjmp() into its own function, to get rid of warnings
5099 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
5100
5101 7. Constructs such as \d{8,} were compiling into the equivalent of
5102 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
5103 outcome, but in this particular case used more store than had been allocated,
5104 which caused the bug to be discovered because it threw up an internal error.
5105
5106 8. The debugging code in both pcre and pcretest for outputting the compiled
5107 form of a regex was going wrong in the case of back references followed by
5108 curly-bracketed repeats.
5109
5110
5111 Version 1.02 12-Dec-97
5112 ----------------------
5113
5114 1. Typos in pcre.3 and comments in the source fixed.
5115
5116 2. Applied a contributed patch to get rid of places where it used to remove
5117 'const' from variables, and fixed some signed/unsigned and uninitialized
5118 variable warnings.
5119
5120 3. Added the "runtest" target to Makefile.
5121
5122 4. Set default compiler flag to -O2 rather than just -O.
5123
5124
5125 Version 1.01 19-Nov-97
5126 ----------------------
5127
5128 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
5129 like /([ab]*)*/, that is, for classes with more than one character in them.
5130
5131 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
5132 as /((?>a*))*/ (a PCRE_EXTRA facility).
5133
5134
5135 Version 1.00 18-Nov-97
5136 ----------------------
5137
5138 1. Added compile-time macros to support systems such as SunOS4 which don't have
5139 memmove() or strerror() but have other things that can be used instead.
5140
5141 2. Arranged that "make clean" removes the executables.
5142
5143
5144 Version 0.99 27-Oct-97
5145 ----------------------
5146
5147 1. Fixed bug in code for optimizing classes with only one character. It was
5148 initializing a 32-byte map regardless, which could cause it to run off the end
5149 of the memory it had got.
5150
5151 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
5152
5153
5154 Version 0.98 22-Oct-97
5155 ----------------------
5156
5157 1. Fixed bug in code for handling temporary memory usage when there are more
5158 back references than supplied space in the ovector. This could cause segfaults.
5159
5160
5161 Version 0.97 21-Oct-97
5162 ----------------------
5163
5164 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
5165
5166 2. Optimized negated single characters not to use a bit map.
5167
5168 3. Brought error texts together as macro definitions; clarified some of them;
5169 fixed one that was wrong - it said "range out of order" when it meant "invalid
5170 escape sequence".
5171
5172 4. Changed some char * arguments to const char *.
5173
5174 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
5175
5176 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
5177 pcretest.
5178
5179
5180 Version 0.96 16-Oct-97
5181 ----------------------
5182
5183 1. Added a simple "pgrep" utility to the distribution.
5184
5185 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
5186 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
5187 where "ddd" means "one or more decimal digits".
5188
5189 3. Fixed serious bug. If a pattern had a back reference, but the call to
5190 pcre_exec() didn't supply a large enough ovector to record the related
5191 identifying subpattern, the match always failed. PCRE now remembers the number
5192 of the largest back reference, and gets some temporary memory in which to save
5193 the offsets during matching if necessary, in order to ensure that
5194 backreferences always work.
5195
5196 4. Increased the compatibility with Perl in a number of ways:
5197
5198 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
5199 to request this handling. The option can be set at compile or exec time.
5200
5201 (b) $ matches before a terminating newline by default; an option
5202 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
5203 mode). The option can be set at compile or exec time.
5204
5205 (c) The handling of \ followed by a digit other than 0 is now supposed to be
5206 the same as Perl's. If the decimal number it represents is less than 10
5207 or there aren't that many previous left capturing parentheses, an octal
5208 escape is read. Inside a character class, it's always an octal escape,
5209 even if it is a single digit.
5210
5211 (d) An escaped but undefined alphabetic character is taken as a literal,
5212 unless PCRE_EXTRA is set. Currently this just reserves the remaining
5213 escapes.
5214
5215 (e) {0} is now permitted. (The previous item is removed from the compiled
5216 pattern).
5217
5218 5. Changed all the names of code files so that the basic parts are no longer
5219 than 10 characters, and abolished the teeny "globals.c" file.
5220
5221 6. Changed the handling of character classes; they are now done with a 32-byte
5222 bit map always.
5223
5224 7. Added the -d and /D options to pcretest to make it possible to look at the
5225 internals of compilation without having to recompile pcre.
5226
5227
5228 Version 0.95 23-Sep-97
5229 ----------------------
5230
5231 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
5232 \x20 at the start of a run of normal characters. These were being treated as
5233 real characters, instead of the source characters being re-checked.
5234
5235
5236 Version 0.94 18-Sep-97
5237 ----------------------
5238
5239 1. The functions are now thread-safe, with the caveat that the global variables
5240 containing pointers to malloc() and free() or alternative functions are the
5241 same for all threads.
5242
5243 2. Get pcre_study() to generate a bitmap of initial characters for non-
5244 anchored patterns when this is possible, and use it if passed to pcre_exec().
5245
5246
5247 Version 0.93 15-Sep-97
5248 ----------------------
5249
5250 1. /(b)|(:+)/ was computing an incorrect first character.
5251
5252 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
5253 but not actually doing anything yet.
5254
5255 3. Treat "-" characters in classes that cannot be part of ranges as literals,
5256 as Perl does (e.g. [-az] or [az-]).
5257
5258 4. Set the anchored flag if a branch starts with .* or .*? because that tests
5259 all possible positions.
5260
5261 5. Split up into different modules to avoid including unneeded functions in a
5262 compiled binary. However, compile and exec are still in one module. The "study"
5263 function is split off.
5264
5265 6. The character tables are now in a separate module whose source is generated
5266 by an auxiliary program - but can then be edited by hand if required. There are
5267 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
5268 toupper() in the code.
5269
5270 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
5271 make them global. Abolish the function for setting them, as the caller can now
5272 set them directly.
5273
5274
5275 Version 0.92 11-Sep-97
5276 ----------------------
5277
5278 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
5279 (e.g. /a{1,3}/) was broken (I mis-optimized it).
5280
5281 2. Caseless matching was not working in character classes if the characters in
5282 the pattern were in upper case.
5283
5284 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
5285
5286 4. Make PCRE_ANCHORED public and accept as a compile option.
5287
5288 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
5289 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
5290 pass them.
5291
5292 6. Give an error if bad option bits passed at compile or run time.
5293
5294 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
5295 pcretest to cause it to pass that flag.
5296
5297 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
5298 options, and the first character, if set.
5299
5300 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
5301
5302
5303 Version 0.91 10-Sep-97
5304 ----------------------
5305
5306 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
5307 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
5308
5309 2. PCRE was looping on encountering an indefinitely repeated back reference to
5310 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
5311 Perl does - treats the match as successful.
5312
5313 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5