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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1397 - (show annotations)
Mon Nov 11 18:33:23 2013 UTC (6 years, 1 month ago) by ph10
File size: 256924 byte(s)
Error occurred while calculating annotation data.
Fix \K bug in possessively repeated groups.
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. It can also be set by
90 (*NO_AUTO_POSSESS) at the start of a pattern.
91
92 18. The character VT has been added to the set of characters that match \s and
93 are generally treated as white space, following this same change in Perl
94 5.18. There is now no difference between "Perl space" and "POSIX space".
95
96 19. The code for checking named groups as conditions, either for being set or
97 for being recursed, has been refactored (this is related to 14 and 15
98 above). Processing unduplicated named groups should now be as fast at
99 numerical groups, and processing duplicated groups should be faster than
100 before.
101
102 20. Two patches to the CMake build system, by Alexander Barkov:
103
104 (1) Replace the "source" command by "." in CMakeLists.txt because
105 "source" is a bash-ism.
106
107 (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
108 without these the CMake build does not work on Solaris.
109
110 21. Perl has changed its handling of \8 and \9. If there is no previously
111 encountered capturing group of those numbers, they are treated as the
112 literal characters 8 and 9 instead of a binary zero followed by the
113 literals. PCRE now does the same.
114
115 22. Following Perl, added \o{} to specify codepoints in octal, making it
116 possible to specify values greater than 0777 and also making them
117 unambiguous.
118
119 23. Perl now gives an error for missing closing braces after \x{... instead of
120 treating the string as literal. PCRE now does the same.
121
122 24. RunTest used to grumble if an inappropriate test was selected explicitly,
123 but just skip it when running all tests. This make it awkward to run ranges
124 of tests when one of them was inappropriate. Now it just skips any
125 inappropriate tests, as it always did when running all tests.
126
127 25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
128 character types such as \d or \w, too many callouts were inserted, and the
129 data that they returned was rubbish.
130
131 26. In UCP mode, \s was not matching two of the characters that Perl matches,
132 namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
133 were matched by \h. The code has now been refactored so that the lists of
134 the horizontal and vertical whitespace characters used for \h and \v (which
135 are defined only in one place) are now also used for \s.
136
137 27. Add JIT support for the 64 bit TileGX architecture.
138 Patch by Jiong Wang (Tilera Corporation).
139
140 28. Possessive quantifiers for classes (both explicit and automatically
141 generated) now use special opcodes instead of wrapping in ONCE brackets.
142
143 29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
144 (because it's meaningless), this was not happening when PCRE_CASELESS was
145 set. Not wrong, but inefficient.
146
147 30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
148 properties for \w, \d, etc) is present in a test regex. Otherwise if the
149 test contains no characters greater than 255, Perl doesn't realise it
150 should be using Unicode semantics.
151
152 31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
153 [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
154 does in Unicode mode.
155
156 32. Added the "forbid" facility to pcretest so that putting tests into the
157 wrong test files can sometimes be quickly detected.
158
159 33. There is now a limit (default 250) on the depth of nesting of parentheses.
160 This limit is imposed to control the amount of system stack used at compile
161 time. It can be changed at build time by --with-parens-nest-limit=xxx or
162 the equivalent in CMake.
163
164 34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
165 errors. Perl warns for these when in warning mode, but PCRE has no facility
166 for giving warnings.
167
168 35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
169 However, this was not working for (?!) because it is optimized to (*FAIL),
170 for which PCRE does not allow quantifiers. The optimization is now disabled
171 when a quantifier follows (?!). I can't see any use for this, but it makes
172 things uniform.
173
174 36. Perl no longer allows group names to start with digits, so I have made this
175 change also in PCRE. It simplifies the code a bit.
176
177 37. In extended mode, Perl ignores spaces before a + that indicates a
178 possessive quantifier. PCRE allowed a space before the quantifier, but not
179 before the possessive +. It now does.
180
181 38. The use of \K (reset reported match start) within a repeated possessive
182 group such as (a\Kb)*+ was not working.
183
184
185 Version 8.33 28-May-2013
186 ------------------------
187
188 1. Added 'U' to some constants that are compared to unsigned integers, to
189 avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
190 variables that are added to signed variables, to ensure the result is
191 signed and can be negated.
192
193 2. Applied patch by Daniel Richard G for quashing MSVC warnings to the
194 CMake config files.
195
196 3. Revise the creation of config.h.generic so that all boolean macros are
197 #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
198 overriding via -D on the command line possible.
199
200 4. Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
201 to unsigned int is reported to make a quite noticeable speed difference in
202 a specific Windows environment. Testing on Linux did also appear to show
203 some benefit (and it is clearly not harmful). Also fixed the definition of
204 Xop which should be unsigned.
205
206 5. Related to (4), changing the definition of the intermediate variable cc
207 in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
208 improvements.
209
210 6. Fix forward search in JIT when link size is 3 or greater. Also removed some
211 unnecessary spaces.
212
213 7. Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
214 and later.
215
216 8. Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
217
218 9. Optimizing fast_forward_start_bits in JIT.
219
220 10. Adding support for callouts in JIT, and fixing some issues revealed
221 during this work. Namely:
222
223 (a) Unoptimized capturing brackets incorrectly reset on backtrack.
224
225 (b) Minimum length was not checked before the matching is started.
226
227 11. The value of capture_last that is passed to callouts was incorrect in some
228 cases when there was a capture on one path that was subsequently abandoned
229 after a backtrack. Also, the capture_last value is now reset after a
230 recursion, since all captures are also reset in this case.
231
232 12. The interpreter no longer returns the "too many substrings" error in the
233 case when an overflowing capture is in a branch that is subsequently
234 abandoned after a backtrack.
235
236 13. In the pathological case when an offset vector of size 2 is used, pcretest
237 now prints out the matched string after a yield of 0 or 1.
238
239 14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
240 Only supported by the JIT compiler at the moment.
241
242 15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
243
244 16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
245 offset of the starting point of the matching process, provided the offsets
246 vector is large enough.
247
248 17. The \A escape now records a lookbehind value of 1, though its execution
249 does not actually inspect the previous character. This is to ensure that,
250 in partial multi-segment matching, at least one character from the old
251 segment is retained when a new segment is processed. Otherwise, if there
252 are no lookbehinds in the pattern, \A might match incorrectly at the start
253 of a new segment.
254
255 18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
256
257 19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
258 gives some modest performance improvement in 8-bit mode.
259
260 20. Added the PCRE-specific property \p{Xuc} for matching characters that can
261 be expressed in certain programming languages using Universal Character
262 Names.
263
264 21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
265 which points out that "non characters" are not "characters that may not
266 appear in Unicode strings" but rather "characters that are reserved for
267 internal use and have only local meaning".
268
269 22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
270 there was a conditional group that depended on an assertion, if the
271 assertion was false, the callout that immediately followed the alternation
272 in the condition was skipped when pcre_exec() was used for matching.
273
274 23. Allow an explicit callout to be inserted before an assertion that is the
275 condition for a conditional group, for compatibility with automatic
276 callouts, which always insert a callout at this point.
277
278 24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
279 confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
280
281 25. (*PRUNE) is now supported by the JIT compiler.
282
283 26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
284
285 27. Fix the case where there are two or more SKIPs with arguments that may be
286 ignored.
287
288 28. (*SKIP) is now supported by the JIT compiler.
289
290 29. (*THEN) is now supported by the JIT compiler.
291
292 30. Update RunTest with additional test selector options.
293
294 31. The way PCRE handles backtracking verbs has been changed in two ways.
295
296 (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
297 SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
298 backtracking. In some cases this makes it more Perl-compatible, but Perl's
299 rather obscure rules do not always do the same thing.
300
301 (2) Previously, backtracking verbs were confined within assertions. This is
302 no longer the case for positive assertions, except for (*ACCEPT). Again,
303 this sometimes improves Perl compatibility, and sometimes does not.
304
305 32. A number of tests that were in test 2 because Perl did things differently
306 have been moved to test 1, because either Perl or PCRE has changed, and
307 these tests are now compatible.
308
309 32. Backtracking control verbs are now handled in the same way in JIT and
310 interpreter.
311
312 33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
313 contained a forward subroutine reference caused a compile error.
314
315 34. Auto-detect and optimize limited repetitions in JIT.
316
317 35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
318 blocking (*UTF) etc.
319
320 36. In the interpreter, maximizing pattern repetitions for characters and
321 character types now use tail recursion, which reduces stack usage.
322
323 37. The value of the max lookbehind was not correctly preserved if a compiled
324 and saved regex was reloaded on a host of different endianness.
325
326 38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
327 of the compiled pattern block, expand the flags field from 16 to 32 bits
328 because it was almost full.
329
330 39. Try madvise first before posix_madvise.
331
332 40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
333 with a pattern such as ^$. It has taken 4 years for anybody to notice! The
334 original change locked out all matches of empty strings. This has been
335 changed so that one match of an empty string per line is recognized.
336 Subsequent searches on the same line (for colouring or for --only-matching,
337 for example) do not recognize empty strings.
338
339 41. Applied a user patch to fix a number of spelling mistakes in comments.
340
341 42. Data lines longer than 65536 caused pcretest to crash.
342
343 43. Clarified the data type for length and startoffset arguments for pcre_exec
344 and pcre_dfa_exec in the function-specific man pages, where they were
345 explicitly stated to be in bytes, never having been updated. I also added
346 some clarification to the pcreapi man page.
347
348 44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
349 a segmentation fault.
350
351
352 Version 8.32 30-November-2012
353 -----------------------------
354
355 1. Improved JIT compiler optimizations for first character search and single
356 character iterators.
357
358 2. Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
359 Patch by Daniel Richard G.
360
361 3. Single character iterator optimizations in the JIT compiler.
362
363 4. Improved JIT compiler optimizations for character ranges.
364
365 5. Rename the "leave" variable names to "quit" to improve WinCE compatibility.
366 Reported by Giuseppe D'Angelo.
367
368 6. The PCRE_STARTLINE bit, indicating that a match can occur only at the start
369 of a line, was being set incorrectly in cases where .* appeared inside
370 atomic brackets at the start of a pattern, or where there was a subsequent
371 *PRUNE or *SKIP.
372
373 7. Improved instruction cache flush for POWER/PowerPC.
374 Patch by Daniel Richard G.
375
376 8. Fixed a number of issues in pcregrep, making it more compatible with GNU
377 grep:
378
379 (a) There is now no limit to the number of patterns to be matched.
380
381 (b) An error is given if a pattern is too long.
382
383 (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
384 are now supported.
385
386 (d) --exclude-from and --include-from (multiple use) have been added.
387
388 (e) Exclusions and inclusions now apply to all files and directories, not
389 just to those obtained from scanning a directory recursively.
390
391 (f) Multiple uses of -f and --file-list are now supported.
392
393 (g) In a Windows environment, the default for -d has been changed from
394 "read" (the GNU grep default) to "skip", because otherwise the presence
395 of a directory in the file list provokes an error.
396
397 (h) The documentation has been revised and clarified in places.
398
399 9. Improve the matching speed of capturing brackets.
400
401 10. Changed the meaning of \X so that it now matches a Unicode extended
402 grapheme cluster.
403
404 11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
405 out POSIX threads when JIT support is configured.
406
407 12. Added support for PCRE_STUDY_EXTRA_NEEDED.
408
409 13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
410 structure could go wrong in environments where size_t is not the same size
411 as int.
412
413 14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
414 set.
415
416 15. The EBCDIC support had decayed; later updates to the code had included
417 explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
418 general tidy up of EBCDIC-related issues, and the documentation was also
419 not quite right. There is now a test that can be run on ASCII systems to
420 check some of the EBCDIC-related things (but is it not a full test).
421
422 16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
423 in a small tidy to the code.
424
425 17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
426
427 18. If the --only-matching (-o) option in pcregrep is specified multiple
428 times, each one causes appropriate output. For example, -o1 -o2 outputs the
429 substrings matched by the 1st and 2nd capturing parentheses. A separating
430 string can be specified by --om-separator (default empty).
431
432 19. Improving the first n character searches.
433
434 20. Turn case lists for horizontal and vertical white space into macros so that
435 they are defined only once.
436
437 21. This set of changes together give more compatible Unicode case-folding
438 behaviour for characters that have more than one other case when UCP
439 support is available.
440
441 (a) The Unicode property table now has offsets into a new table of sets of
442 three or more characters that are case-equivalent. The MultiStage2.py
443 script that generates these tables (the pcre_ucd.c file) now scans
444 CaseFolding.txt instead of UnicodeData.txt for character case
445 information.
446
447 (b) The code for adding characters or ranges of characters to a character
448 class has been abstracted into a generalized function that also handles
449 case-independence. In UTF-mode with UCP support, this uses the new data
450 to handle characters with more than one other case.
451
452 (c) A bug that is fixed as a result of (b) is that codepoints less than 256
453 whose other case is greater than 256 are now correctly matched
454 caselessly. Previously, the high codepoint matched the low one, but not
455 vice versa.
456
457 (d) The processing of \h, \H, \v, and \ in character classes now makes use
458 of the new class addition function, using character lists defined as
459 macros alongside the case definitions of 20 above.
460
461 (e) Caseless back references now work with characters that have more than
462 one other case.
463
464 (f) General caseless matching of characters with more than one other case
465 is supported.
466
467 22. Unicode character properties were updated from Unicode 6.2.0
468
469 23. Improved CMake support under Windows. Patch by Daniel Richard G.
470
471 24. Add support for 32-bit character strings, and UTF-32
472
473 25. Major JIT compiler update (code refactoring and bugfixing).
474 Experimental Sparc 32 support is added.
475
476 26. Applied a modified version of Daniel Richard G's patch to create
477 pcre.h.generic and config.h.generic by "make" instead of in the
478 PrepareRelease script.
479
480 27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
481 pcre_compile.c when checking for a zero character.
482
483 28. Introducing a native interface for JIT. Through this interface, the compiled
484 machine code can be directly executed. The purpose of this interface is to
485 provide fast pattern matching, so several sanity checks are not performed.
486 However, feature tests are still performed. The new interface provides
487 1.4x speedup compared to the old one.
488
489 29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
490 the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
491 was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
492
493 30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
494 greater than 0x7fffffff (which cannot be represented in UTF-8, even under
495 the "old" RFC 2279). Instead, it ended up passing a negative length to
496 pcre_exec().
497
498 31. Add support for GCC's visibility feature to hide internal functions.
499
500 32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
501 "unknown -C option" after outputting 0 or 1.
502
503 33. There is now support for generating a code coverage report for the test
504 suite in environments where gcc is the compiler and lcov is installed. This
505 is mainly for the benefit of the developers.
506
507 34. If PCRE is built with --enable-valgrind, certain memory regions are marked
508 unaddressable using valgrind annotations, allowing valgrind to detect
509 invalid memory accesses. This is mainly for the benefit of the developers.
510
511 25. (*UTF) can now be used to start a pattern in any of the three libraries.
512
513 26. Give configure error if --enable-cpp but no C++ compiler found.
514
515
516 Version 8.31 06-July-2012
517 -------------------------
518
519 1. Fixing a wrong JIT test case and some compiler warnings.
520
521 2. Removed a bashism from the RunTest script.
522
523 3. Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
524 to unsigned type, result still unsigned" that was given by an MS compiler
525 on encountering the code "-sizeof(xxx)".
526
527 4. Partial matching support is added to the JIT compiler.
528
529 5. Fixed several bugs concerned with partial matching of items that consist
530 of more than one character:
531
532 (a) /^(..)\1/ did not partially match "aba" because checking references was
533 done on an "all or nothing" basis. This also applied to repeated
534 references.
535
536 (b) \R did not give a hard partial match if \r was found at the end of the
537 subject.
538
539 (c) \X did not give a hard partial match after matching one or more
540 characters at the end of the subject.
541
542 (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
543 a partial match for the string "\r".
544
545 (e) When newline was set to CRLF, the metacharacter "." did not recognize
546 a partial match for a CR character at the end of the subject string.
547
548 6. If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
549 running pcretest, the text "(JIT)" added to the output whenever JIT is
550 actually used to run the match.
551
552 7. Individual JIT compile options can be set in pcretest by following -s+[+]
553 or /S+[+] with a digit between 1 and 7.
554
555 8. OP_NOT now supports any UTF character not just single-byte ones.
556
557 9. (*MARK) control verb is now supported by the JIT compiler.
558
559 10. The command "./RunTest list" lists the available tests without actually
560 running any of them. (Because I keep forgetting what they all are.)
561
562 11. Add PCRE_INFO_MAXLOOKBEHIND.
563
564 12. Applied a (slightly modified) user-supplied patch that improves performance
565 when the heap is used for recursion (compiled with --disable-stack-for-
566 recursion). Instead of malloc and free for each heap frame each time a
567 logical recursion happens, frames are retained on a chain and re-used where
568 possible. This sometimes gives as much as 30% improvement.
569
570 13. As documented, (*COMMIT) is now confined to within a recursive subpattern
571 call.
572
573 14. As documented, (*COMMIT) is now confined to within a positive assertion.
574
575 15. It is now possible to link pcretest with libedit as an alternative to
576 libreadline.
577
578 16. (*COMMIT) control verb is now supported by the JIT compiler.
579
580 17. The Unicode data tables have been updated to Unicode 6.1.0.
581
582 18. Added --file-list option to pcregrep.
583
584 19. Added binary file support to pcregrep, including the -a, --binary-files,
585 -I, and --text options.
586
587 20. The madvise function is renamed for posix_madvise for QNX compatibility
588 reasons. Fixed by Giuseppe D'Angelo.
589
590 21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
591
592 22. Changed the option for creating symbolic links for 16-bit man pages from
593 -s to -sf so that re-installing does not cause issues.
594
595 23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
596
597 24. Fixed a very old bug in pcretest that caused errors with restarted DFA
598 matches in certain environments (the workspace was not being correctly
599 retained). Also added to pcre_dfa_exec() a simple plausibility check on
600 some of the workspace data at the beginning of a restart.
601
602 25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
603 was not doing so when it should - probably a typo introduced by SVN 528
604 (change 8.10/14).
605
606 26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
607 \w+ when the character tables indicated that \x{c4} was a word character.
608 There were several related cases, all because the tests for doing a table
609 lookup were testing for characters less than 127 instead of 255.
610
611 27. If a pattern contains capturing parentheses that are not used in a match,
612 their slots in the ovector are set to -1. For those that are higher than
613 any matched groups, this happens at the end of processing. In the case when
614 there were back references that the ovector was too small to contain
615 (causing temporary malloc'd memory to be used during matching), and the
616 highest capturing number was not used, memory off the end of the ovector
617 was incorrectly being set to -1. (It was using the size of the temporary
618 memory instead of the true size.)
619
620 28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
621 ovector size, it uses memory at the end of the block that it has got.
622
623 29. Check for an overlong MARK name and give an error at compile time. The
624 limit is 255 for the 8-bit library and 65535 for the 16-bit library.
625
626 30. JIT compiler update.
627
628 31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
629 Rill for the patch.
630
631 32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
632
633 33. Variable renamings in the PCRE-JIT compiler. No functionality change.
634
635 34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
636 SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
637 was enabled.
638
639 35. Improve JIT code generation for greedy plus quantifier.
640
641 36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
642 1 to "aa" instead of to an empty string. The bug affected repeated groups
643 that could potentially match an empty string.
644
645 37. Optimizing single character iterators in JIT.
646
647 38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
648 the same checks as \x{...} characters in non-JavaScript mode. Specifically,
649 codepoints that are too big for the mode are faulted, and in a UTF mode,
650 disallowed codepoints are also faulted.
651
652 39. If PCRE was compiled with UTF support, in three places in the DFA
653 matcher there was code that should only have been obeyed in UTF mode, but
654 was being obeyed unconditionally. In 8-bit mode this could cause incorrect
655 processing when bytes with values greater than 127 were present. In 16-bit
656 mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
657 both cases the values are those that cannot be the first data item in a UTF
658 character. The three items that might have provoked this were recursions,
659 possessively repeated groups, and atomic groups.
660
661 40. Ensure that libpcre is explicitly listed in the link commands for pcretest
662 and pcregrep, because some OS require shared objects to be explicitly
663 passed to ld, causing the link step to fail if they are not.
664
665 41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
666 mode, patterns that started with \h* or \R* might be incorrectly matched.
667
668
669 Version 8.30 04-February-2012
670 -----------------------------
671
672 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
673 name is defined in ctype.h.
674
675 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
676 only in quite long subpatterns.
677
678 3. Removed the function pcre_info(), which has been obsolete and deprecated
679 since it was replaced by pcre_fullinfo() in February 2000.
680
681 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
682 match a (*MARK), and the match failed at the start of the subject, a
683 reference to memory before the start of the subject could occur. This bug
684 was introduced by fix 17 of release 8.21.
685
686 5. A reference to an unset group with zero minimum repetition was giving
687 totally wrong answers (in non-JavaScript-compatibility mode). For example,
688 /(another)?(\1?)test/ matched against "hello world test". This bug was
689 introduced in release 8.13.
690
691 6. Add support for 16-bit character strings (a large amount of work involving
692 many changes and refactorings).
693
694 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
695 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
696 from a file.
697
698 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
699 rounding is not applied in this particular case).
700
701 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
702 if they appear, or are escaped, in patterns.
703
704 10. Get rid of a number of -Wunused-but-set-variable warnings.
705
706 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
707 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
708 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
709 also returns the mark "x". This bug applied to capturing parentheses,
710 non-capturing parentheses, and atomic parentheses. It also applied to some
711 assertions.
712
713 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
714 information out of configure.ac instead of relying on pcre.h.generic, which
715 is not stored in the repository.
716
717 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
718 -lreadline.
719
720 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
721
722 15. Applied Graycode's patch to put the top-level frame on the stack rather
723 than the heap when not using the stack for recursion. This gives a
724 performance improvement in many cases when recursion is not deep.
725
726 16. Experimental code added to "pcretest -C" to output the stack frame size.
727
728
729 Version 8.21 12-Dec-2011
730 ------------------------
731
732 1. Updating the JIT compiler.
733
734 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
735 are added as well.
736
737 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
738 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
739 calling _pcre_jit_exec. Some extra comments are added.
740
741 4. (*MARK) settings inside atomic groups that do not contain any capturing
742 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
743 was introduced by change 18 for 8.20.
744
745 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
746 ECMA-262 standard.
747
748 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
749 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
750 This bug was probably introduced by change 9 of 8.13.
751
752 7. While fixing 6 above, I noticed that a number of other items were being
753 incorrectly rejected as "not fixed length". This arose partly because newer
754 opcodes had not been added to the fixed-length checking code. I have (a)
755 corrected the bug and added tests for these items, and (b) arranged for an
756 error to occur if an unknown opcode is encountered while checking for fixed
757 length instead of just assuming "not fixed length". The items that were
758 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
759 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
760 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
761
762 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
763 being incorrectly compiled and would have given unpredicatble results.
764
765 9. A possessively repeated subpattern with minimum repeat count greater than
766 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
767 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
768 the first (A) could occur when it should not.
769
770 10. Add a cast and remove a redundant test from the code.
771
772 11. JIT should use pcre_malloc/pcre_free for allocation.
773
774 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
775 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
776 is anything other than /usr, -L is still shown).
777
778 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
779
780 14. Perl does not support \N without a following name in a [] class; PCRE now
781 also gives an error.
782
783 15. If a forward reference was repeated with an upper limit of around 2000,
784 it caused the error "internal error: overran compiling workspace". The
785 maximum number of forward references (including repeats) was limited by the
786 internal workspace, and dependent on the LINK_SIZE. The code has been
787 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
788 the default depends on LINK_SIZE. There is a new upper limit (for safety)
789 of around 200,000 forward references. While doing this, I also speeded up
790 the filling in of repeated forward references.
791
792 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
793 incorrectly expecting the subject to contain another "a" after the start.
794
795 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
796 in the match, the SKIP should be ignored. This was not happening; instead
797 the SKIP was being treated as NOMATCH. For patterns such as
798 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
799 tested.
800
801 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
802 now much more compatible with Perl, in particular in cases where the result
803 is a non-match for a non-anchored pattern. For example, if
804 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
805 "m", where previously it did not return a name. A side effect of this
806 change is that for partial matches, the last encountered mark name is
807 returned, as for non matches. A number of tests that were previously not
808 Perl-compatible have been moved into the Perl-compatible test files. The
809 refactoring has had the pleasing side effect of removing one argument from
810 the match() function, thus reducing its stack requirements.
811
812 19. If the /S+ option was used in pcretest to study a pattern using JIT,
813 subsequent uses of /S (without +) incorrectly behaved like /S+.
814
815 21. Retrieve executable code size support for the JIT compiler and fixing
816 some warnings.
817
818 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
819 not work when the shorter character appeared right at the end of the
820 subject string.
821
822 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
823 systems.
824
825 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
826 output it when the /M option is used in pcretest.
827
828 25. The CheckMan script was not being included in the distribution. Also, added
829 an explicit "perl" to run Perl scripts from the PrepareRelease script
830 because this is reportedly needed in Windows.
831
832 26. If study data was being save in a file and studying had not found a set of
833 "starts with" bytes for the pattern, the data written to the file (though
834 never used) was taken from uninitialized memory and so caused valgrind to
835 complain.
836
837 27. Updated RunTest.bat as provided by Sheri Pierce.
838
839 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
840
841 29. Computation of memory usage for the table of capturing group names was
842 giving an unnecessarily large value.
843
844
845 Version 8.20 21-Oct-2011
846 ------------------------
847
848 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
849 a POSIX class. After further experiments with Perl, which convinced me that
850 Perl has bugs and confusions, a closing square bracket is no longer allowed
851 in a POSIX name. This bug also affected patterns with classes that started
852 with full stops.
853
854 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
855 captured substring, but while checking the failing first alternative,
856 substring 1 is temporarily captured. If the output vector supplied to
857 pcre_exec() was not big enough for this capture, the yield of the function
858 was still zero ("insufficient space for captured substrings"). This cannot
859 be totally fixed without adding another stack variable, which seems a lot
860 of expense for a edge case. However, I have improved the situation in cases
861 such as /(a)(b)x|abc/ matched against "abc", where the return code
862 indicates that fewer than the maximum number of slots in the ovector have
863 been set.
864
865 3. Related to (2) above: when there are more back references in a pattern than
866 slots in the output vector, pcre_exec() uses temporary memory during
867 matching, and copies in the captures as far as possible afterwards. It was
868 using the entire output vector, but this conflicts with the specification
869 that only 2/3 is used for passing back captured substrings. Now it uses
870 only the first 2/3, for compatibility. This is, of course, another edge
871 case.
872
873 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
874 main code base, and can be used by building with --enable-jit. When this is
875 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
876 runtime --no-jit option is given.
877
878 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
879 ovector, the return from the function was zero, implying that there were
880 other matches that did not fit. The correct "exactly full" value is now
881 returned.
882
883 6. If a subpattern that was called recursively or as a subroutine contained
884 (*PRUNE) or any other control that caused it to give a non-standard return,
885 invalid errors such as "Error -26 (nested recursion at the same subject
886 position)" or even infinite loops could occur.
887
888 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
889 computing the minimum length on reaching *ACCEPT, and so ended up with the
890 wrong value of 1 rather than 0. Further investigation indicates that
891 computing a minimum subject length in the presence of *ACCEPT is difficult
892 (think back references, subroutine calls), and so I have changed the code
893 so that no minimum is registered for a pattern that contains *ACCEPT.
894
895 8. If (*THEN) was present in the first (true) branch of a conditional group,
896 it was not handled as intended. [But see 16 below.]
897
898 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
899 Sheri Pierce.
900
901 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
902 the first byte in a match must be "a".
903
904 11. Change 17 for 8.13 increased the recursion depth for patterns like
905 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
906 pattern contains any instances of (*THEN). If it does not, the old
907 optimizations are restored. It would be nice to do this on a per-group
908 basis, but at the moment that is not feasible.
909
910 12. In some environments, the output of pcretest -C is CRLF terminated. This
911 broke RunTest's code that checks for the link size. A single white space
912 character after the value is now allowed for.
913
914 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
915 For "fr", it uses the Windows-specific input and output files.
916
917 14. If (*THEN) appeared in a group that was called recursively or as a
918 subroutine, it did not work as intended. [But see next item.]
919
920 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
921 pattern fragments (but not containing any | characters). If A and B are
922 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
923 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
924 D. In other words, Perl considers parentheses that do not contain any |
925 characters to be part of a surrounding alternative, whereas PCRE was
926 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
927 differently. PCRE now behaves in the same way as Perl, except in the case
928 of subroutine/recursion calls such as (?1) which have in any case always
929 been different (but PCRE had them first :-).
930
931 16. Related to 15 above: Perl does not treat the | in a conditional group as
932 creating alternatives. Such a group is treated in the same way as an
933 ordinary group without any | characters when processing (*THEN). PCRE has
934 been changed to match Perl's behaviour.
935
936 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
937 RunGrepTest script failed.
938
939 18. Change 22 for version 13 caused atomic groups to use more stack. This is
940 inevitable for groups that contain captures, but it can lead to a lot of
941 stack use in large patterns. The old behaviour has been restored for atomic
942 groups that do not contain any capturing parentheses.
943
944 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
945 suppress the check for a minimum subject length at run time. (If it was
946 given to pcre_exec() or pcre_dfa_exec() it did work.)
947
948 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
949 fail to work when decoding hex characters in data strings in EBCDIC
950 environments.
951
952 21. It appears that in at least one Mac OS environment, the isxdigit() function
953 is implemented as a macro that evaluates to its argument more than once,
954 contravening the C 90 Standard (I haven't checked a later standard). There
955 was an instance in pcretest which caused it to go wrong when processing
956 \x{...} escapes in subject strings. The has been rewritten to avoid using
957 things like p++ in the argument of isxdigit().
958
959
960 Version 8.13 16-Aug-2011
961 ------------------------
962
963 1. The Unicode data tables have been updated to Unicode 6.0.0.
964
965 2. Two minor typos in pcre_internal.h have been fixed.
966
967 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
968 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
969 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
970
971 4. There were a number of related bugs in the code for matching backrefences
972 caselessly in UTF-8 mode when codes for the characters concerned were
973 different numbers of bytes. For example, U+023A and U+2C65 are an upper
974 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
975 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
976 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
977 2-byte code at the end of the subject (it thought there wasn't enough data
978 left).
979
980 5. Comprehensive information about what went wrong is now returned by
981 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
982 as the output vector has at least 2 elements. The offset of the start of
983 the failing character and a reason code are placed in the vector.
984
985 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
986 now returned is for the first byte of the failing character, instead of the
987 last byte inspected. This is an incompatible change, but I hope it is small
988 enough not to be a problem. It makes the returned offset consistent with
989 pcre_exec() and pcre_dfa_exec().
990
991 7. pcretest now gives a text phrase as well as the error number when
992 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
993 failure, the offset and reason code are output.
994
995 8. When \R was used with a maximizing quantifier it failed to skip backwards
996 over a \r\n pair if the subsequent match failed. Instead, it just skipped
997 back over a single character (\n). This seems wrong (because it treated the
998 two characters as a single entity when going forwards), conflicts with the
999 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1000 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1001 has been changed.
1002
1003 9. Some internal refactoring has changed the processing so that the handling
1004 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1005 time (the PCRE_DOTALL option was changed this way some time ago: version
1006 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1007 which was always a bit of a fudge. It also means that there is one less
1008 argument for the match() function, which reduces its stack requirements
1009 slightly. This change also fixes an incompatibility with Perl: the pattern
1010 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1011
1012 10. More internal refactoring has drastically reduced the number of recursive
1013 calls to match() for possessively repeated groups such as (abc)++ when
1014 using pcre_exec().
1015
1016 11. While implementing 10, a number of bugs in the handling of groups were
1017 discovered and fixed:
1018
1019 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1020 (a|)*(?1) gave a compile-time internal error.
1021 ((a|)+)+ did not notice that the outer group could match an empty string.
1022 (^a|^)+ was not marked as anchored.
1023 (.*a|.*)+ was not marked as matching at start or after a newline.
1024
1025 12. Yet more internal refactoring has removed another argument from the match()
1026 function. Special calls to this function are now indicated by setting a
1027 value in a variable in the "match data" data block.
1028
1029 13. Be more explicit in pcre_study() instead of relying on "default" for
1030 opcodes that mean there is no starting character; this means that when new
1031 ones are added and accidentally left out of pcre_study(), testing should
1032 pick them up.
1033
1034 14. The -s option of pcretest has been documented for ages as being an old
1035 synonym of -m (show memory usage). I have changed it to mean "force study
1036 for every regex", that is, assume /S for every regex. This is similar to -i
1037 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1038 using it. It makes it easier to run collections of tests with and without
1039 study enabled, and thereby test pcre_study() more easily. All the standard
1040 tests are now run with and without -s (but some patterns can be marked as
1041 "never study" - see 20 below).
1042
1043 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1044 restoration of the capturing data to the outer values was not happening
1045 correctly.
1046
1047 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1048 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1049 pattern had matched an empty string, and so incorrectly returned a no
1050 match.
1051
1052 17. There was optimizing code for the last branch of non-capturing parentheses,
1053 and also for the obeyed branch of a conditional subexpression, which used
1054 tail recursion to cut down on stack usage. Unfortunately, now that there is
1055 the possibility of (*THEN) occurring in these branches, tail recursion is
1056 no longer possible because the return has to be checked for (*THEN). These
1057 two optimizations have therefore been removed. [But see 8.20/11 above.]
1058
1059 18. If a pattern containing \R was studied, it was assumed that \R always
1060 matched two bytes, thus causing the minimum subject length to be
1061 incorrectly computed because \R can also match just one byte.
1062
1063 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1064 was incorrectly computed.
1065
1066 20. If /S is present twice on a test pattern in pcretest input, it now
1067 *disables* studying, thereby overriding the use of -s on the command line
1068 (see 14 above). This is necessary for one or two tests to keep the output
1069 identical in both cases.
1070
1071 21. When (*ACCEPT) was used in an assertion that matched an empty string and
1072 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1073
1074 22. When an atomic group that contained a capturing parenthesis was
1075 successfully matched, but the branch in which it appeared failed, the
1076 capturing was not being forgotten if a higher numbered group was later
1077 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1078 group 1 to "a", when in fact it should be unset. This applied to multi-
1079 branched capturing and non-capturing groups, repeated or not, and also to
1080 positive assertions (capturing in negative assertions does not happen
1081 in PCRE) and also to nested atomic groups.
1082
1083 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1084 subject after a captured substring, to make it easier to tell which of a
1085 number of identical substrings has been captured.
1086
1087 24. The way atomic groups are processed by pcre_exec() has been changed so that
1088 if they are repeated, backtracking one repetition now resets captured
1089 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1090 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1091 "aaa". Previously, it would have been "a". As part of this code
1092 refactoring, the way recursive calls are handled has also been changed.
1093
1094 25. If an assertion condition captured any substrings, they were not passed
1095 back unless some other capturing happened later. For example, if
1096 (?(?=(a))a) was matched against "a", no capturing was returned.
1097
1098 26. When studying a pattern that contained subroutine calls or assertions,
1099 the code for finding the minimum length of a possible match was handling
1100 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1101 group 1 called group 2 while simultaneously a separate group 2 called group
1102 1). A stack overflow occurred in this case. I have fixed this by limiting
1103 the recursion depth to 10.
1104
1105 27. Updated RunTest.bat in the distribution to the version supplied by Tom
1106 Fortmann. This supports explicit test numbers on the command line, and has
1107 argument validation and error reporting.
1108
1109 28. An instance of \X with an unlimited repeat could fail if at any point the
1110 first character it looked at was a mark character.
1111
1112 29. Some minor code refactoring concerning Unicode properties and scripts
1113 should reduce the stack requirement of match() slightly.
1114
1115 30. Added the '=' option to pcretest to check the setting of unused capturing
1116 slots at the end of the pattern, which are documented as being -1, but are
1117 not included in the return count.
1118
1119 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1120 compiled something random. Now it gives a compile-time error (as does
1121 Perl).
1122
1123 32. A *MARK encountered during the processing of a positive assertion is now
1124 recorded and passed back (compatible with Perl).
1125
1126 33. If --only-matching or --colour was set on a pcregrep call whose pattern
1127 had alternative anchored branches, the search for a second match in a line
1128 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1129 matched the line "0102" twice. The same bug affected patterns that started
1130 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1131 twice.
1132
1133 34. Previously, PCRE did not allow quantification of assertions. However, Perl
1134 does, and because of capturing effects, quantifying parenthesized
1135 assertions may at times be useful. Quantifiers are now allowed for
1136 parenthesized assertions.
1137
1138 35. A minor code tidy in pcre_compile() when checking options for \R usage.
1139
1140 36. \g was being checked for fancy things in a character class, when it should
1141 just be a literal "g".
1142
1143 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1144 appearance of a nested POSIX class supersedes an apparent external class.
1145 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1146 unescaped square brackets may also appear as part of class names. For
1147 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1148 more like Perl. (But see 8.20/1 above.)
1149
1150 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1151 was because it thought it was \N{name}, which is not supported).
1152
1153 39. Add minix to OS list not supporting the -S option in pcretest.
1154
1155 40. PCRE tries to detect cases of infinite recursion at compile time, but it
1156 cannot analyze patterns in sufficient detail to catch mutual recursions
1157 such as ((?1))((?2)). There is now a runtime test that gives an error if a
1158 subgroup is called recursively as a subpattern for a second time at the
1159 same position in the subject string. In previous releases this might have
1160 been caught by the recursion limit, or it might have run out of stack.
1161
1162 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1163 happen only once. PCRE was, however incorrectly giving a compile time error
1164 "recursive call could loop indefinitely" because it cannot analyze the
1165 pattern in sufficient detail. The compile time test no longer happens when
1166 PCRE is compiling a conditional subpattern, but actual runaway loops are
1167 now caught at runtime (see 40 above).
1168
1169 42. It seems that Perl allows any characters other than a closing parenthesis
1170 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1171 has been changed to be the same.
1172
1173 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1174 as not to get warnings when autogen.sh is called. Also changed
1175 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1176
1177 44. To help people who use pcregrep to scan files containing exceedingly long
1178 lines, the following changes have been made:
1179
1180 (a) The default value of the buffer size parameter has been increased from
1181 8K to 20K. (The actual buffer used is three times this size.)
1182
1183 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1184 PCRE is built.
1185
1186 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1187 to be set at run time.
1188
1189 (d) Numerical values in pcregrep options can be followed by K or M, for
1190 example --buffer-size=50K.
1191
1192 (e) If a line being scanned overflows pcregrep's buffer, an error is now
1193 given and the return code is set to 2.
1194
1195 45. Add a pointer to the latest mark to the callout data block.
1196
1197 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1198 partial match of an empty string instead of no match. This was specific to
1199 the use of ".".
1200
1201 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1202 complete match instead of a partial match. This bug was dependent on both
1203 the PCRE_UTF8 and PCRE_DOTALL options being set.
1204
1205 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1206 starting byte set, because \b was not being ignored.
1207
1208
1209 Version 8.12 15-Jan-2011
1210 ------------------------
1211
1212 1. Fixed some typos in the markup of the man pages, and wrote a script that
1213 checks for such things as part of the documentation building process.
1214
1215 2. On a big-endian 64-bit system, pcregrep did not correctly process the
1216 --match-limit and --recursion-limit options (added for 8.11). In
1217 particular, this made one of the standard tests fail. (The integer value
1218 went into the wrong half of a long int.)
1219
1220 3. If the --colour option was given to pcregrep with -v (invert match), it
1221 did strange things, either producing crazy output, or crashing. It should,
1222 of course, ignore a request for colour when reporting lines that do not
1223 match.
1224
1225 4. Another pcregrep bug caused similar problems if --colour was specified with
1226 -M (multiline) and the pattern match finished with a line ending.
1227
1228 5. In pcregrep, when a pattern that ended with a literal newline sequence was
1229 matched in multiline mode, the following line was shown as part of the
1230 match. This seems wrong, so I have changed it.
1231
1232 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
1233 the check for further matches in the same line (so they could be coloured)
1234 to overrun the end of the current line. If another match was found, it was
1235 incorrectly shown (and then shown again when found in the next line).
1236
1237 7. If pcregrep was compiled under Windows, there was a reference to the
1238 function pcregrep_exit() before it was defined. I am assuming this was
1239 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1240 reported by a user. I've moved the definition above the reference.
1241
1242
1243 Version 8.11 10-Dec-2010
1244 ------------------------
1245
1246 1. (*THEN) was not working properly if there were untried alternatives prior
1247 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1248 backtracked to try for "b" instead of moving to the next alternative branch
1249 at the same level (in this case, to look for "c"). The Perl documentation
1250 is clear that when (*THEN) is backtracked onto, it goes to the "next
1251 alternative in the innermost enclosing group".
1252
1253 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1254 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
1255 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1256 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1257 (*THEN).
1258
1259 3. If \s appeared in a character class, it removed the VT character from
1260 the class, even if it had been included by some previous item, for example
1261 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1262 of \s, but is part of the POSIX "space" class.)
1263
1264 4. A partial match never returns an empty string (because you can always
1265 match an empty string at the end of the subject); however the checking for
1266 an empty string was starting at the "start of match" point. This has been
1267 changed to the "earliest inspected character" point, because the returned
1268 data for a partial match starts at this character. This means that, for
1269 example, /(?<=abc)def/ gives a partial match for the subject "abc"
1270 (previously it gave "no match").
1271
1272 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1273 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1274 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1275 has an implication that the given string is incomplete (because a partial
1276 match is preferred over a full match). For this reason, these items now
1277 give a partial match in this situation. [Aside: previously, the one case
1278 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1279 match rather than a full match, which was wrong by the old rules, but is
1280 now correct.]
1281
1282 6. There was a bug in the handling of #-introduced comments, recognized when
1283 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1284 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1285 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1286 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1287 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1288 places in pcre_compile().
1289
1290 7. Related to (6) above, when pcre_compile() was skipping #-introduced
1291 comments when looking ahead for named forward references to subpatterns,
1292 the only newline sequence it recognized was NL. It now handles newlines
1293 according to the set newline convention.
1294
1295 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1296 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1297 cater for a lack of strerror(). These oversights have been fixed.
1298
1299 9. Added --match-limit and --recursion-limit to pcregrep.
1300
1301 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1302
1303 11. When the -o option was used, pcregrep was setting a return code of 1, even
1304 when matches were found, and --line-buffered was not being honoured.
1305
1306 12. Added an optional parentheses number to the -o and --only-matching options
1307 of pcregrep.
1308
1309 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1310 can match an empty string. The code to do it in pcretest and pcredemo
1311 needed fixing:
1312
1313 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1314 only one byte after an empty string match just before CRLF (this case
1315 just got forgotten; "any" and "anycrlf" were OK).
1316
1317 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1318 mode when an empty string match preceded an ASCII character followed by
1319 a non-ASCII character. (The code for advancing by one character rather
1320 than one byte was nonsense.)
1321
1322 (c) The pcredemo.c sample program did not have any code at all to handle
1323 the cases when CRLF is a valid newline sequence.
1324
1325 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1326 as a starting offset was within the subject string. There is now a new
1327 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1328 negative or greater than the length of the string. In order to test this,
1329 pcretest is extended to allow the setting of negative starting offsets.
1330
1331 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1332 starting offset points to the beginning of a UTF-8 character was
1333 unnecessarily clumsy. I tidied it up.
1334
1335 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1336 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1337
1338 17. Nobody had reported that the --include_dir option, which was added in
1339 release 7.7 should have been called --include-dir (hyphen, not underscore)
1340 for compatibility with GNU grep. I have changed it to --include-dir, but
1341 left --include_dir as an undocumented synonym, and the same for
1342 --exclude-dir, though that is not available in GNU grep, at least as of
1343 release 2.5.4.
1344
1345 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1346 characters from a string of bytes) have been redefined so as not to use
1347 loops, in order to improve performance in some environments. At the same
1348 time, I abstracted some of the common code into auxiliary macros to save
1349 repetition (this should not affect the compiled code).
1350
1351 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1352 compile-time error is now given if \c is not followed by an ASCII
1353 character, that is, a byte less than 128. (In EBCDIC mode, the code is
1354 different, and any byte value is allowed.)
1355
1356 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1357 START_OPTIMIZE option, which is now allowed at compile time - but just
1358 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1359 to pcregrep and other applications that have no direct access to PCRE
1360 options. The new /Y option in pcretest sets this option when calling
1361 pcre_compile().
1362
1363 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1364 back references. Groups containing recursive back references were forced to
1365 be atomic by that change, but in the case of named groups, the amount of
1366 memory required was incorrectly computed, leading to "Failed: internal
1367 error: code overflow". This has been fixed.
1368
1369 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1370 pcretest.c, to avoid build problems in some Borland environments.
1371
1372
1373 Version 8.10 25-Jun-2010
1374 ------------------------
1375
1376 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1377 THEN.
1378
1379 2. (*ACCEPT) was not working when inside an atomic group.
1380
1381 3. Inside a character class, \B is treated as a literal by default, but
1382 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1383 causes the error). The code is unchanged, but I tidied the documentation.
1384
1385 4. Inside a character class, PCRE always treated \R and \X as literals,
1386 whereas Perl faults them if its -w option is set. I have changed PCRE so
1387 that it faults them when PCRE_EXTRA is set.
1388
1389 5. Added support for \N, which always matches any character other than
1390 newline. (It is the same as "." when PCRE_DOTALL is not set.)
1391
1392 6. When compiling pcregrep with newer versions of gcc which may have
1393 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1394 declared with attribute warn_unused_result" were given. Just casting the
1395 result to (void) does not stop the warnings; a more elaborate fudge is
1396 needed. I've used a macro to implement this.
1397
1398 7. Minor change to pcretest.c to avoid a compiler warning.
1399
1400 8. Added four artifical Unicode properties to help with an option to make
1401 \s etc use properties (see next item). The new properties are: Xan
1402 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1403
1404 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1405 use Unicode properties. (*UCP) at the start of a pattern can be used to set
1406 this option. Modified pcretest to add /W to test this facility. Added
1407 REG_UCP to make it available via the POSIX interface.
1408
1409 10. Added --line-buffered to pcregrep.
1410
1411 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1412 studied, and the match started with a letter with a code point greater than
1413 127 whose first byte was different to the first byte of the other case of
1414 the letter, the other case of this starting letter was not recognized
1415 (#976).
1416
1417 12. If a pattern that was studied started with a repeated Unicode property
1418 test, for example, \p{Nd}+, there was the theoretical possibility of
1419 setting up an incorrect bitmap of starting bytes, but fortunately it could
1420 not have actually happened in practice until change 8 above was made (it
1421 added property types that matched character-matching opcodes).
1422
1423 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1424 possible starting bytes for non-anchored patterns.
1425
1426 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1427 \R, and also a number of cases that involve Unicode properties, both
1428 explicit and implicit when PCRE_UCP is set.
1429
1430 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1431 input, it could crash or give wrong results if characters with values
1432 greater than 0xc0 were present in the subject string. (Detail: it assumed
1433 UTF-8 input when processing these items.)
1434
1435 16. Added a lot of (int) casts to avoid compiler warnings in systems where
1436 size_t is 64-bit (#991).
1437
1438 17. Added a check for running out of memory when PCRE is compiled with
1439 --disable-stack-for-recursion (#990).
1440
1441 18. If the last data line in a file for pcretest does not have a newline on
1442 the end, a newline was missing in the output.
1443
1444 19. The default pcre_chartables.c file recognizes only ASCII characters (values
1445 less than 128) in its various bitmaps. However, there is a facility for
1446 generating tables according to the current locale when PCRE is compiled. It
1447 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1448 characters, are recognized by isspace() and therefore were getting set in
1449 these tables, and indeed these tables seem to approximate to ISO 8859. This
1450 caused a problem in UTF-8 mode when pcre_study() was used to create a list
1451 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1452 which of course cannot start UTF-8 characters. I have changed the code so
1453 that only real ASCII characters (less than 128) and the correct starting
1454 bytes for UTF-8 encodings are set for characters greater than 127 when in
1455 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1456 altogether.)
1457
1458 20. Added the /T option to pcretest so as to be able to run tests with non-
1459 standard character tables, thus making it possible to include the tests
1460 used for 19 above in the standard set of tests.
1461
1462 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1463 reference to a subpattern the other side of a comment that contains an
1464 opening parenthesis caused either an internal compiling error, or a
1465 reference to the wrong subpattern.
1466
1467
1468 Version 8.02 19-Mar-2010
1469 ------------------------
1470
1471 1. The Unicode data tables have been updated to Unicode 5.2.0.
1472
1473 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
1474 configured.
1475
1476 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
1477 original author of that file, following a query about its status.
1478
1479 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
1480 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1481
1482 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1483 quantifier applied to a forward-referencing subroutine call, could compile
1484 incorrect code or give the error "internal error: previously-checked
1485 referenced subpattern not found".
1486
1487 6. Both MS Visual Studio and Symbian OS have problems with initializing
1488 variables to point to external functions. For these systems, therefore,
1489 pcre_malloc etc. are now initialized to local functions that call the
1490 relevant global functions.
1491
1492 7. There were two entries missing in the vectors called coptable and poptable
1493 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1494 I've fixed the data, and added a kludgy way of testing at compile time that
1495 the lengths are correct (equal to the number of opcodes).
1496
1497 8. Following on from 7, I added a similar kludge to check the length of the
1498 eint vector in pcreposix.c.
1499
1500 9. Error texts for pcre_compile() are held as one long string to avoid too
1501 much relocation at load time. To find a text, the string is searched,
1502 counting zeros. There was no check for running off the end of the string,
1503 which could happen if a new error number was added without updating the
1504 string.
1505
1506 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1507
1508 11. \K was not working if it appeared in an atomic group or in a group that
1509 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1510 \K is "not well defined" if used in an assertion. PCRE now accepts it if
1511 the assertion is positive, but not if it is negative.
1512
1513 12. Change 11 fortuitously reduced the size of the stack frame used in the
1514 "match()" function of pcre_exec.c by one pointer. Forthcoming
1515 implementation of support for (*MARK) will need an extra pointer on the
1516 stack; I have reserved it now, so that the stack frame size does not
1517 decrease.
1518
1519 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1520 item in branch that calls a recursion is a subroutine call - as in the
1521 second branch in the above example - was incorrectly given the compile-
1522 time error "recursive call could loop indefinitely" because pcre_compile()
1523 was not correctly checking the subroutine for matching a non-empty string.
1524
1525 14. The checks for overrunning compiling workspace could trigger after an
1526 overrun had occurred. This is a "should never occur" error, but it can be
1527 triggered by pathological patterns such as hundreds of nested parentheses.
1528 The checks now trigger 100 bytes before the end of the workspace.
1529
1530 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1531
1532
1533 Version 8.01 19-Jan-2010
1534 ------------------------
1535
1536 1. If a pattern contained a conditional subpattern with only one branch (in
1537 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1538 computed the wrong minimum data length (which is of course zero for such
1539 subpatterns). This could cause incorrect "no match" results.
1540
1541 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1542 the pattern is reset in the first branch, pcre_compile() failed with
1543 "internal error: code overflow at offset...". This happened only when
1544 the reset was to the original external option setting. (An optimization
1545 abstracts leading options settings into an external setting, which was the
1546 cause of this.)
1547
1548 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1549 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1550 assertion pattern did not match (meaning that the assertion was true), it
1551 was incorrectly treated as false if the SKIP had been reached during the
1552 matching. This also applied to assertions used as conditions.
1553
1554 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
1555 assertion subpattern, including such a pattern used as a condition,
1556 unpredictable results occurred, instead of the error return
1557 PCRE_ERROR_DFA_UITEM.
1558
1559 5. The C++ GlobalReplace function was not working like Perl for the special
1560 situation when an empty string is matched. It now does the fancy magic
1561 stuff that is necessary.
1562
1563 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1564 removed. (These were left over from very, very early versions of PCRE.)
1565
1566 7. Some cosmetic changes to the code to make life easier when compiling it
1567 as part of something else:
1568
1569 (a) Change DEBUG to PCRE_DEBUG.
1570
1571 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1572 called "current" as "current_branch", to prevent a collision with the
1573 Linux macro when compiled as a kernel module.
1574
1575 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1576 prevent a collision with the Linux macro when compiled as a kernel
1577 module.
1578
1579 8. In pcre_compile() there are some checks for integer overflows that used to
1580 cast potentially large values to (double). This has been changed to that
1581 when building, a check for int64_t is made, and if it is found, it is used
1582 instead, thus avoiding the use of floating point arithmetic. (There is no
1583 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1584 double.
1585
1586 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1587 2005 (difference between two addresses compared to an unsigned value).
1588
1589 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1590 custom one, because of the following reported problem in Windows:
1591
1592 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1593 under Win32.
1594 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1595 therefore missing the function definition.
1596 - The compiler thus generates a "C" signature for the test function.
1597 - The linker fails to find the "C" function.
1598 - PCRE fails to configure if asked to do so against libbz2.
1599
1600 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1601 messages were output:
1602
1603 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1604 rerunning libtoolize, to keep the correct libtool macros in-tree.
1605 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1606
1607 I have done both of these things.
1608
1609 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1610 most of the time, it *can* run out if it is given a pattern that contains a
1611 runaway infinite recursion. I updated the discussion in the pcrestack man
1612 page.
1613
1614 13. Now that we have gone to the x.xx style of version numbers, the minor
1615 version may start with zero. Using 08 or 09 is a bad idea because users
1616 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1617 interpreted as invalid octal numbers. I've updated the previous comment in
1618 configure.ac, and also added a check that gives an error if 08 or 09 are
1619 used.
1620
1621 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1622 causing partial matching to fail when the end of the subject matched \W
1623 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1624
1625 15. There were some discrepancies between the declarations in pcre_internal.h
1626 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1627 their definitions. The declarations used "const uschar *" and the
1628 definitions used USPTR. Even though USPTR is normally defined as "const
1629 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1630 reported that: "This difference in casting confuses some C++ compilers, for
1631 example, SunCC recognizes above declarations as different functions and
1632 generates broken code for hbpcre." I have changed the declarations to use
1633 USPTR.
1634
1635 16. GNU libtool is named differently on some systems. The autogen.sh script now
1636 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1637 (FreeBSD).
1638
1639 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1640 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1641 comment: "Figure out how to create a longlong from a string: strtoll and
1642 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1643 instance, but it only takes 2 args instead of 3!"
1644
1645 18. A subtle bug concerned with back references has been fixed by a change of
1646 specification, with a corresponding code fix. A pattern such as
1647 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1648 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1649 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1650 same bug. Such groups have to be quantified to be useful, or contained
1651 inside another quantified group. (If there's no repetition, the reference
1652 can never match.) The problem arises because, having left the group and
1653 moved on to the rest of the pattern, a later failure that backtracks into
1654 the group uses the captured value from the final iteration of the group
1655 rather than the correct earlier one. I have fixed this in PCRE by forcing
1656 any group that contains a reference to itself to be an atomic group; that
1657 is, there cannot be any backtracking into it once it has completed. This is
1658 similar to recursive and subroutine calls.
1659
1660
1661 Version 8.00 19-Oct-09
1662 ----------------------
1663
1664 1. The table for translating pcre_compile() error codes into POSIX error codes
1665 was out-of-date, and there was no check on the pcre_compile() error code
1666 being within the table. This could lead to an OK return being given in
1667 error.
1668
1669 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1670 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1671 in a Windows environment.
1672
1673 3. The pcregrep --count option prints the count for each file even when it is
1674 zero, as does GNU grep. However, pcregrep was also printing all files when
1675 --files-with-matches was added. Now, when both options are given, it prints
1676 counts only for those files that have at least one match. (GNU grep just
1677 prints the file name in this circumstance, but including the count seems
1678 more useful - otherwise, why use --count?) Also ensured that the
1679 combination -clh just lists non-zero counts, with no names.
1680
1681 4. The long form of the pcregrep -F option was incorrectly implemented as
1682 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1683 but it seems right to fix it, and I didn't think it was worth preserving
1684 the old behaviour.
1685
1686 5. The command line items --regex=pattern and --regexp=pattern were not
1687 recognized by pcregrep, which required --regex pattern or --regexp pattern
1688 (with a space rather than an '='). The man page documented the '=' forms,
1689 which are compatible with GNU grep; these now work.
1690
1691 6. No libpcreposix.pc file was created for pkg-config; there was just
1692 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1693
1694 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1695 when UCP support is not needed, by modifying the Python script that
1696 generates it from Unicode data files. This should not matter if the module
1697 is correctly used as a library, but I received one complaint about 50K of
1698 unwanted data. My guess is that the person linked everything into his
1699 program rather than using a library. Anyway, it does no harm.
1700
1701 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1702 was a minimum greater than 1 for a wide character in a possessive
1703 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1704 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1705 character. Chaos in the form of incorrect output or a compiling loop could
1706 result.
1707
1708 9. The restrictions on what a pattern can contain when partial matching is
1709 requested for pcre_exec() have been removed. All patterns can now be
1710 partially matched by this function. In addition, if there are at least two
1711 slots in the offset vector, the offset of the earliest inspected character
1712 for the match and the offset of the end of the subject are set in them when
1713 PCRE_ERROR_PARTIAL is returned.
1714
1715 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1716 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1717 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1718 and may be more useful for multi-segment matching.
1719
1720 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1721 used to be given if ever the end of the subject was reached; now it is
1722 given only if matching could not proceed because another character was
1723 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1724 string "Z", which now yields "no match" instead of "partial match". In the
1725 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1726 final character ended with (*FAIL).
1727
1728 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1729 if the pattern had a "must contain" character that was already found in the
1730 earlier partial match, unless partial matching was again requested. For
1731 example, with the pattern /dog.(body)?/, the "must contain" character is
1732 "g". If the first part-match was for the string "dog", restarting with
1733 "sbody" failed. This bug has been fixed.
1734
1735 13. The string returned by pcre_dfa_exec() after a partial match has been
1736 changed so that it starts at the first inspected character rather than the
1737 first character of the match. This makes a difference only if the pattern
1738 starts with a lookbehind assertion or \b or \B (\K is not supported by
1739 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1740 matching functions compatible, and I think it's the right thing to do.
1741
1742 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1743 so that the demonstration program is easily available in environments where
1744 PCRE has not been installed from source.
1745
1746 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1747 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1748 library.
1749
1750 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1751 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1752 is not the first non-POSIX option to be added. Clearly some people find
1753 these options useful.
1754
1755 17. If a caller to the POSIX matching function regexec() passes a non-zero
1756 value for nmatch with a NULL value for pmatch, the value of
1757 nmatch is forced to zero.
1758
1759 18. RunGrepTest did not have a test for the availability of the -u option of
1760 the diff command, as RunTest does. It now checks in the same way as
1761 RunTest, and also checks for the -b option.
1762
1763 19. If an odd number of negated classes containing just a single character
1764 interposed, within parentheses, between a forward reference to a named
1765 subpattern and the definition of the subpattern, compilation crashed with
1766 an internal error, complaining that it could not find the referenced
1767 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1768 [The bug was that it was starting one character too far in when skipping
1769 over the character class, thus treating the ] as data rather than
1770 terminating the class. This meant it could skip too much.]
1771
1772 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1773 /g option in pcretest when the pattern contains \K, which makes it possible
1774 to have an empty string match not at the start, even when the pattern is
1775 anchored. Updated pcretest and pcredemo to use this option.
1776
1777 21. If the maximum number of capturing subpatterns in a recursion was greater
1778 than the maximum at the outer level, the higher number was returned, but
1779 with unset values at the outer level. The correct (outer level) value is
1780 now given.
1781
1782 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1783 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1784 make it do so. The string so far is captured, making this feature
1785 compatible with Perl.
1786
1787 23. The tests have been re-organized, adding tests 11 and 12, to make it
1788 possible to check the Perl 5.10 features against Perl 5.10.
1789
1790 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1791 pattern matches a fixed length string. PCRE did not allow this; now it
1792 does. Neither allows recursion.
1793
1794 25. I finally figured out how to implement a request to provide the minimum
1795 length of subject string that was needed in order to match a given pattern.
1796 (It was back references and recursion that I had previously got hung up
1797 on.) This code has now been added to pcre_study(); it finds a lower bound
1798 to the length of subject needed. It is not necessarily the greatest lower
1799 bound, but using it to avoid searching strings that are too short does give
1800 some useful speed-ups. The value is available to calling programs via
1801 pcre_fullinfo().
1802
1803 26. While implementing 25, I discovered to my embarrassment that pcretest had
1804 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1805 study optimizations had never been tested with that matching function.
1806 Oops. What is worse, even when it was passed study data, there was a bug in
1807 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1808 were also very few tests of studied patterns with pcre_dfa_exec().
1809
1810 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1811 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1812 on the other side of the coin, they are no longer allowed to have different
1813 names, because these cannot be distinguished in PCRE, and this has caused
1814 confusion. (This is a difference from Perl.)
1815
1816 28. When duplicate subpattern names are present (necessarily with different
1817 numbers, as required by 27 above), and a test is made by name in a
1818 conditional pattern, either for a subpattern having been matched, or for
1819 recursion in such a pattern, all the associated numbered subpatterns are
1820 tested, and the overall condition is true if the condition is true for any
1821 one of them. This is the way Perl works, and is also more like the way
1822 testing by number works.
1823
1824
1825 Version 7.9 11-Apr-09
1826 ---------------------
1827
1828 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1829 (pcretest), all targets were linked against these libraries. This included
1830 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1831 libraries. This caused unwanted dependencies to be created. This problem
1832 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1833 pcretest is linked with readline.
1834
1835 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1836 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1837 moved outside it again, because FALSE and TRUE are already defined in AIX,
1838 but BOOL is not.
1839
1840 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1841 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1842
1843 4. The pcregrep documentation said spaces were inserted as well as colons (or
1844 hyphens) following file names and line numbers when outputting matching
1845 lines. This is not true; no spaces are inserted. I have also clarified the
1846 wording for the --colour (or --color) option.
1847
1848 5. In pcregrep, when --colour was used with -o, the list of matching strings
1849 was not coloured; this is different to GNU grep, so I have changed it to be
1850 the same.
1851
1852 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1853 each matching line was coloured. Now it goes on to look for further matches
1854 of any of the test patterns, which is the same behaviour as GNU grep.
1855
1856 7. A pattern that could match an empty string could cause pcregrep to loop; it
1857 doesn't make sense to accept an empty string match in pcregrep, so I have
1858 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1859 seems to be how GNU grep behaves. [But see later change 40 for release
1860 8.33.]
1861
1862 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1863 start or after a newline", because the conditional assertion was not being
1864 correctly handled. The rule now is that both the assertion and what follows
1865 in the first alternative must satisfy the test.
1866
1867 9. If auto-callout was enabled in a pattern with a conditional group whose
1868 condition was an assertion, PCRE could crash during matching, both with
1869 pcre_exec() and pcre_dfa_exec().
1870
1871 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1872 used for matching.
1873
1874 11. Unicode property support in character classes was not working for
1875 characters (bytes) greater than 127 when not in UTF-8 mode.
1876
1877 12. Added the -M command line option to pcretest.
1878
1879 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1880
1881 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1882
1883 16. Added comments and documentation about mis-use of no_arg in the C++
1884 wrapper.
1885
1886 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1887 from Martin Jerabek that uses macro names for all relevant character and
1888 string constants.
1889
1890 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1891 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1892 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1893 these, but not everybody uses configure.
1894
1895 19. A conditional group that had only one branch was not being correctly
1896 recognized as an item that could match an empty string. This meant that an
1897 enclosing group might also not be so recognized, causing infinite looping
1898 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1899 with the subject "ab", where knowledge that the repeated group can match
1900 nothing is needed in order to break the loop.
1901
1902 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1903 exec(), but without supplying a callout function, matching went wrong.
1904
1905 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1906 leak if the size of the offset vector was greater than 30. When the vector
1907 is smaller, the saved offsets during recursion go onto a local stack
1908 vector, but for larger vectors malloc() is used. It was failing to free
1909 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1910 error, in fact).
1911
1912 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1913 heapframe that is used only when UTF-8 support is enabled. This caused no
1914 problem, but was untidy.
1915
1916 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1917 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1918 included within another project.
1919
1920 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1921 slightly modified by me:
1922
1923 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1924 not building pcregrep.
1925
1926 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1927 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1928
1929 25. Forward references, both numeric and by name, in patterns that made use of
1930 duplicate group numbers, could behave incorrectly or give incorrect errors,
1931 because when scanning forward to find the reference group, PCRE was not
1932 taking into account the duplicate group numbers. A pattern such as
1933 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1934
1935 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1936 the feature of a custom pointer more persuasive (as requested by a user).
1937
1938 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1939 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1940 pre-defined.
1941
1942 28. Added support for (*UTF8) at the start of a pattern.
1943
1944 29. Arrange for flags added by the "release type" setting in CMake to be shown
1945 in the configuration summary.
1946
1947
1948 Version 7.8 05-Sep-08
1949 ---------------------
1950
1951 1. Replaced UCP searching code with optimized version as implemented for Ad
1952 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1953 stage table and inline lookup instead of a function, giving speed ups of 2
1954 to 5 times on some simple patterns that I tested. Permission was given to
1955 distribute the MultiStage2.py script that generates the tables (it's not in
1956 the tarball, but is in the Subversion repository).
1957
1958 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1959 scripts.
1960
1961 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1962 a group with a zero qualifier. The result of the study could be incorrect,
1963 or the function might crash, depending on the pattern.
1964
1965 4. Caseless matching was not working for non-ASCII characters in back
1966 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1967 It now works when Unicode Property Support is available.
1968
1969 5. In pcretest, an escape such as \x{de} in the data was always generating
1970 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1971 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1972 truncation.
1973
1974 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1975
1976 7. Added two (int) casts to pcregrep when printing the difference of two
1977 pointers, in case they are 64-bit values.
1978
1979 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1980 test 2 if it fails.
1981
1982 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1983 and a #define of that name to empty if it is not externally set. This is to
1984 allow users of MSVC to set it if necessary.
1985
1986 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1987 the convenience functions in the pcre_get.c source file.
1988
1989 11. An option change at the start of a pattern that had top-level alternatives
1990 could cause overwriting and/or a crash. This command provoked a crash in
1991 some environments:
1992
1993 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1994
1995 This potential security problem was recorded as CVE-2008-2371.
1996
1997 12. For a pattern where the match had to start at the beginning or immediately
1998 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1999 pcre_dfa_exec() could read past the end of the passed subject if there was
2000 no match. To help with detecting such bugs (e.g. with valgrind), I modified
2001 pcretest so that it places the subject at the end of its malloc-ed buffer.
2002
2003 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2004 exec() might read past the end of the data buffer in UTF-8 mode.
2005
2006 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2007 the data contained the byte 0x85 as part of a UTF-8 character within its
2008 first line. This applied both to normal and DFA matching.
2009
2010 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2011 /^[^d]*?$/8 failed to match "abc".
2012
2013 16. Added a missing copyright notice to pcrecpp_internal.h.
2014
2015 17. Make it more clear in the documentation that values returned from
2016 pcre_exec() in ovector are byte offsets, not character counts.
2017
2018 18. Tidied a few places to stop certain compilers from issuing warnings.
2019
2020 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2021 supplied by Stefan Weber. I made a further small update for 7.8 because
2022 there is a change of source arrangements: the pcre_searchfuncs.c module is
2023 replaced by pcre_ucd.c.
2024
2025
2026 Version 7.7 07-May-08
2027 ---------------------
2028
2029 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
2030 a string to a long long, pretend we don't even have a long long." This is
2031 done by checking for the strtoq, strtoll, and _strtoi64 functions.
2032
2033 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2034 pre-7.6 versions, which defined a global no_arg variable instead of putting
2035 it in the RE class. (See also #8 below.)
2036
2037 3. Remove a line of dead code, identified by coverity and reported by Nuno
2038 Lopes.
2039
2040 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
2041
2042 (1) The include/exclude patterns were being applied to the whole pathnames
2043 of files, instead of just to the final components.
2044
2045 (2) If there was more than one level of directory, the subdirectories were
2046 skipped unless they satisfied the include/exclude conditions. This is
2047 inconsistent with GNU grep (and could even be seen as contrary to the
2048 pcregrep specification - which I improved to make it absolutely clear).
2049 The action now is always to scan all levels of directory, and just
2050 apply the include/exclude patterns to regular files.
2051
2052 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2053 --exclude_dir in the tests to avoid scanning .svn directories.
2054
2055 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
2056 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2057 doesn't support NULs in patterns.
2058
2059 7. Added some missing "const"s to declarations of static tables in
2060 pcre_compile.c and pcre_dfa_exec.c.
2061
2062 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2063 caused by fix #2 above. (Subsequently also a second patch to fix the
2064 first patch. And a third patch - this was a messy problem.)
2065
2066 9. Applied Craig's patch to remove the use of push_back().
2067
2068 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2069 matching function regexec().
2070
2071 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2072 which, however, unlike Perl's \g{...}, are subroutine calls, not back
2073 references. PCRE supports relative numbers with this syntax (I don't think
2074 Oniguruma does).
2075
2076 12. Previously, a group with a zero repeat such as (...){0} was completely
2077 omitted from the compiled regex. However, this means that if the group
2078 was called as a subroutine from elsewhere in the pattern, things went wrong
2079 (an internal error was given). Such groups are now left in the compiled
2080 pattern, with a new opcode that causes them to be skipped at execution
2081 time.
2082
2083 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2084 to the way PCRE behaves:
2085
2086 (a) A lone ] character is dis-allowed (Perl treats it as data).
2087
2088 (b) A back reference to an unmatched subpattern matches an empty string
2089 (Perl fails the current match path).
2090
2091 (c) A data ] in a character class must be notated as \] because if the
2092 first data character in a class is ], it defines an empty class. (In
2093 Perl it is not possible to have an empty class.) The empty class []
2094 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2095 The negative empty class [^] matches any one character, independently
2096 of the DOTALL setting.
2097
2098 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2099 non-existent subpattern following a character class starting with ']' and
2100 containing () gave an internal compiling error instead of "reference to
2101 non-existent subpattern". Fortunately, when the pattern did exist, the
2102 compiled code was correct. (When scanning forwards to check for the
2103 existence of the subpattern, it was treating the data ']' as terminating
2104 the class, so got the count wrong. When actually compiling, the reference
2105 was subsequently set up correctly.)
2106
2107 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2108 it was being rejected as not supported by pcre_dfa_exec(), even though
2109 other assertions are supported. I have made pcre_dfa_exec() support
2110 (*FAIL).
2111
2112 16. The implementation of 13c above involved the invention of a new opcode,
2113 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2114 cannot be changed at match time, I realized I could make a small
2115 improvement to matching performance by compiling OP_ALLANY instead of
2116 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2117 on the OP_ANY path.
2118
2119 17. Compiling pcretest on Windows with readline support failed without the
2120 following two fixes: (1) Make the unistd.h include conditional on
2121 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2122
2123 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2124 ncurses library to be included for pcretest when ReadLine support is
2125 requested, but also to allow for it to be overridden. This patch came from
2126 Daniel Bergström.
2127
2128 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2129 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2130 any errors with the current Unicode tables. Thanks to Peter Kankowski for
2131 spotting this.
2132
2133
2134 Version 7.6 28-Jan-08
2135 ---------------------
2136
2137 1. A character class containing a very large number of characters with
2138 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2139 overflow.
2140
2141 2. Patch to cut out the "long long" test in pcrecpp_unittest when
2142 HAVE_LONG_LONG is not defined.
2143
2144 3. Applied Christian Ehrlicher's patch to update the CMake build files to
2145 bring them up to date and include new features. This patch includes:
2146
2147 - Fixed PH's badly added libz and libbz2 support.
2148 - Fixed a problem with static linking.
2149 - Added pcredemo. [But later removed - see 7 below.]
2150 - Fixed dftables problem and added an option.
2151 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2152 HAVE_LONG_LONG.
2153 - Added readline support for pcretest.
2154 - Added an listing of the option settings after cmake has run.
2155
2156 4. A user submitted a patch to Makefile that makes it easy to create
2157 "pcre.dll" under mingw when using Configure/Make. I added stuff to
2158 Makefile.am that cause it to include this special target, without
2159 affecting anything else. Note that the same mingw target plus all
2160 the other distribution libraries and programs are now supported
2161 when configuring with CMake (see 6 below) instead of with
2162 Configure/Make.
2163
2164 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2165 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2166 exported in the Windows port". It has not yet been confirmed that the patch
2167 solves the problem, but it does no harm.
2168
2169 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2170 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2171 with CMake, and also correct the comment about stack recursion.
2172
2173 7. Remove the automatic building of pcredemo from the ./configure system and
2174 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2175 of a program that users should build themselves after PCRE is installed, so
2176 building it automatically is not really right. What is more, it gave
2177 trouble in some build environments.
2178
2179 8. Further tidies to CMakeLists.txt from Sheri and Christian.
2180
2181
2182 Version 7.5 10-Jan-08
2183 ---------------------
2184
2185 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
2186 values in parens when parsing an RE using the C++ wrapper."
2187
2188 2. Negative specials like \S did not work in character classes in UTF-8 mode.
2189 Characters greater than 255 were excluded from the class instead of being
2190 included.
2191
2192 3. The same bug as (2) above applied to negated POSIX classes such as
2193 [:^space:].
2194
2195 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2196 defined or documented. It seems to have been a typo for PCRE_STATIC, so
2197 I have changed it.
2198
2199 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
2200 first named subpattern) and a construct such as (?&a) would reference the
2201 first named subpattern whose name started with "a" (in other words, the
2202 length check was missing). Both these problems are fixed. "Subpattern name
2203 expected" is now given for (?&) (a zero-length name), and this patch also
2204 makes it give the same error for \k'' (previously it complained that that
2205 was a reference to a non-existent subpattern).
2206
2207 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
2208 this is right because (?- can be followed by option settings as well as by
2209 digits. I have, however, made the messages clearer.
2210
2211 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2212 than the number used in the conditional) now cause a compile-time error.
2213 This is actually not compatible with Perl, which accepts such patterns, but
2214 treats the conditional as always being FALSE (as PCRE used to), but it
2215 seems to me that giving a diagnostic is better.
2216
2217 8. Change "alphameric" to the more common word "alphanumeric" in comments
2218 and messages.
2219
2220 9. Fix two occurrences of "backslash" in comments that should have been
2221 "backspace".
2222
2223 10. Remove two redundant lines of code that can never be obeyed (their function
2224 was moved elsewhere).
2225
2226 11. The program that makes PCRE's Unicode character property table had a bug
2227 which caused it to generate incorrect table entries for sequences of
2228 characters that have the same character type, but are in different scripts.
2229 It amalgamated them into a single range, with the script of the first of
2230 them. In other words, some characters were in the wrong script. There were
2231 thirteen such cases, affecting characters in the following ranges:
2232
2233 U+002b0 - U+002c1
2234 U+0060c - U+0060d
2235 U+0061e - U+00612
2236 U+0064b - U+0065e
2237 U+0074d - U+0076d
2238 U+01800 - U+01805
2239 U+01d00 - U+01d77
2240 U+01d9b - U+01dbf
2241 U+0200b - U+0200f
2242 U+030fc - U+030fe
2243 U+03260 - U+0327f
2244 U+0fb46 - U+0fbb1
2245 U+10450 - U+1049d
2246
2247 12. The -o option (show only the matching part of a line) for pcregrep was not
2248 compatible with GNU grep in that, if there was more than one match in a
2249 line, it showed only the first of them. It now behaves in the same way as
2250 GNU grep.
2251
2252 13. If the -o and -v options were combined for pcregrep, it printed a blank
2253 line for every non-matching line. GNU grep prints nothing, and pcregrep now
2254 does the same. The return code can be used to tell if there were any
2255 non-matching lines.
2256
2257 14. Added --file-offsets and --line-offsets to pcregrep.
2258
2259 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2260 infinitely looping recursion. The bug was that positive lookaheads were not
2261 being skipped when checking for a possible empty match (negative lookaheads
2262 and both kinds of lookbehind were skipped).
2263
2264 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2265 inclusion of <windows.h> to before rather than after the definition of
2266 INVALID_FILE_ATTRIBUTES (patch from David Byron).
2267
2268 17. Specifying a possessive quantifier with a specific limit for a Unicode
2269 character property caused pcre_compile() to compile bad code, which led at
2270 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2271 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2272 caused the error; without that there was no problem.
2273
2274 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2275
2276 19. Added --enable-pcretest-libreadline.
2277
2278 20. In pcrecpp.cc, the variable 'count' was incremented twice in
2279 RE::GlobalReplace(). As a result, the number of replacements returned was
2280 double what it should be. I removed one of the increments, but Craig sent a
2281 later patch that removed the other one (the right fix) and added unit tests
2282 that check the return values (which was not done before).
2283
2284 21. Several CMake things:
2285
2286 (1) Arranged that, when cmake is used on Unix, the libraries end up with
2287 the names libpcre and libpcreposix, not just pcre and pcreposix.
2288
2289 (2) The above change means that pcretest and pcregrep are now correctly
2290 linked with the newly-built libraries, not previously installed ones.
2291
2292 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2293
2294 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2295 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2296 UTF-8 newline character). The key issue is that the pattern starts .*;
2297 this means that the match must be either at the beginning, or after a
2298 newline. The bug was in the code for advancing after a failed match and
2299 checking that the new position followed a newline. It was not taking
2300 account of UTF-8 characters correctly.
2301
2302 23. PCRE was behaving differently from Perl in the way it recognized POSIX
2303 character classes. PCRE was not treating the sequence [:...:] as a
2304 character class unless the ... were all letters. Perl, however, seems to
2305 allow any characters between [: and :], though of course it rejects as
2306 unknown any "names" that contain non-letters, because all the known class
2307 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2308 for example, whereas PCRE did not - it did not recognize a POSIX character
2309 class. This seemed a bit dangerous, so the code has been changed to be
2310 closer to Perl. The behaviour is not identical to Perl, because PCRE will
2311 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2312 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2313 Perl does, and where it didn't before.
2314
2315 24. Rewrite so as to remove the single use of %n from pcregrep because in some
2316 Windows environments %n is disabled by default.
2317
2318
2319 Version 7.4 21-Sep-07
2320 ---------------------
2321
2322 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
2323 means that a class such as [\s] counted as "explicit reference to CR or
2324 LF". That isn't really right - the whole point of the change was to try to
2325 help when there was an actual mention of one of the two characters. So now
2326 the change happens only if \r or \n (or a literal CR or LF) character is
2327 encountered.
2328
2329 2. The 32-bit options word was also used for 6 internal flags, but the numbers
2330 of both had grown to the point where there were only 3 bits left.
2331 Fortunately, there was spare space in the data structure, and so I have
2332 moved the internal flags into a new 16-bit field to free up more option
2333 bits.
2334
2335 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2336 but did not set the internal JCHANGED flag - either of these is enough to
2337 control the way the "get" function works - but the PCRE_INFO_JCHANGED
2338 facility is supposed to tell if (?J) was ever used, so now (?J) at the
2339 start sets both bits.
2340
2341 4. Added options (at build time, compile time, exec time) to change \R from
2342 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2343
2344 5. doc/pcresyntax.html was missing from the distribution.
2345
2346 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2347 compatibility, even though it is no longer used.
2348
2349 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2350 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2351 windows.h file is present (where different names are used). [This was
2352 reversed later after testing - see 16 below.]
2353
2354 8. Changed all #include <config.h> to #include "config.h". There were also
2355 some further <pcre.h> cases that I changed to "pcre.h".
2356
2357 9. When pcregrep was used with the --colour option, it missed the line ending
2358 sequence off the lines that it output.
2359
2360 10. It was pointed out to me that arrays of string pointers cause lots of
2361 relocations when a shared library is dynamically loaded. A technique of
2362 using a single long string with a table of offsets can drastically reduce
2363 these. I have refactored PCRE in four places to do this. The result is
2364 dramatic:
2365
2366 Originally: 290
2367 After changing UCP table: 187
2368 After changing error message table: 43
2369 After changing table of "verbs" 36
2370 After changing table of Posix names 22
2371
2372 Thanks to the folks working on Gregex for glib for this insight.
2373
2374 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2375 unicode-properties was also set.
2376
2377 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2378
2379 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2380 checked only for CRLF.
2381
2382 14. Added casts to pcretest.c to avoid compiler warnings.
2383
2384 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2385
2386 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2387 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2388 entirely. This removes changes made in 7 above.
2389
2390 17. The CMake files have been updated, and there is now more information about
2391 building with CMake in the NON-UNIX-USE document.
2392
2393
2394 Version 7.3 28-Aug-07
2395 ---------------------
2396
2397 1. In the rejigging of the build system that eventually resulted in 7.1, the
2398 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2399 brackets there is not right, since it causes compilers to look for an
2400 installed pcre.h, not the version that is in the source that is being
2401 compiled (which of course may be different). I have changed it back to:
2402
2403 #include "pcre.h"
2404
2405 I have a vague recollection that the change was concerned with compiling in
2406 different directories, but in the new build system, that is taken care of
2407 by the VPATH setting the Makefile.
2408
2409 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2410 when the subject happened to end in the byte 0x85 (e.g. if the last
2411 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2412 characters but of course it shouldn't be taken as a newline when it is part
2413 of another character. The bug was that, for an unlimited repeat of . in
2414 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2415 characters when looking for a newline.
2416
2417 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2418
2419 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2420 in debug output.
2421
2422 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2423 long printing in the pcrecpp unittest when running under MinGW.
2424
2425 6. ESC_K was left out of the EBCDIC table.
2426
2427 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2428 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2429 limit also applies to "virtual nesting" when a pattern is recursive, and in
2430 this case 1000 isn't so big. I have been able to remove this limit at the
2431 expense of backing off one optimization in certain circumstances. Normally,
2432 when pcre_exec() would call its internal match() function recursively and
2433 immediately return the result unconditionally, it uses a "tail recursion"
2434 feature to save stack. However, when a subpattern that can match an empty
2435 string has an unlimited repetition quantifier, it no longer makes this
2436 optimization. That gives it a stack frame in which to save the data for
2437 checking that an empty string has been matched. Previously this was taken
2438 from the 1000-entry workspace that had been reserved. So now there is no
2439 explicit limit, but more stack is used.
2440
2441 8. Applied Daniel's patches to solve problems with the import/export magic
2442 syntax that is required for Windows, and which was going wrong for the
2443 pcreposix and pcrecpp parts of the library. These were overlooked when this
2444 problem was solved for the main library.
2445
2446 9. There were some crude static tests to avoid integer overflow when computing
2447 the size of patterns that contain repeated groups with explicit upper
2448 limits. As the maximum quantifier is 65535, the maximum group length was
2449 set at 30,000 so that the product of these two numbers did not overflow a
2450 32-bit integer. However, it turns out that people want to use groups that
2451 are longer than 30,000 bytes (though not repeat them that many times).
2452 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2453 made it possible to implement the integer overflow checks in a much more
2454 dynamic way, which I have now done. The artificial limitation on group
2455 length has been removed - we now have only the limit on the total length of
2456 the compiled pattern, which depends on the LINK_SIZE setting.
2457
2458 10. Fixed a bug in the documentation for get/copy named substring when
2459 duplicate names are permitted. If none of the named substrings are set, the
2460 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2461 empty string.
2462
2463 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2464 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2465 because the ] is interpreted as the first data character and the
2466 terminating ] is not found. PCRE has been made compatible with Perl in this
2467 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2468 cause memory overwriting.
2469
2470 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2471 string has been matched (to stop an infinite loop). It was not recognizing
2472 a conditional subpattern that could match an empty string if that
2473 subpattern was within another subpattern. For example, it looped when
2474 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
2475 condition was not nested. This bug has been fixed.
2476
2477 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2478 past the start of the subject in the presence of bytes with the top bit
2479 set, for example "\x8aBCD".
2480
2481 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2482 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2483
2484 14. Optimized (?!) to (*FAIL).
2485
2486 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2487 This restricts code points to be within the range 0 to 0x10FFFF, excluding
2488 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2489 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2490 does: it's just the validity check that is more restrictive.
2491
2492 16. Inserted checks for integer overflows during escape sequence (backslash)
2493 processing, and also fixed erroneous offset values for syntax errors during
2494 backslash processing.
2495
2496 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2497 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2498
2499 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2500 caused an overrun.
2501
2502 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2503 something other than just ASCII characters) inside a group that had an
2504 unlimited repeat caused a loop at compile time (while checking to see
2505 whether the group could match an empty string).
2506
2507 20. Debugging a pattern containing \p or \P could cause a crash. For example,
2508 [\P{Any}] did so. (Error in the code for printing property names.)
2509
2510 21. An orphan \E inside a character class could cause a crash.
2511
2512 22. A repeated capturing bracket such as (A)? could cause a wild memory
2513 reference during compilation.
2514
2515 23. There are several functions in pcre_compile() that scan along a compiled
2516 expression for various reasons (e.g. to see if it's fixed length for look
2517 behind). There were bugs in these functions when a repeated \p or \P was
2518 present in the pattern. These operators have additional parameters compared
2519 with \d, etc, and these were not being taken into account when moving along
2520 the compiled data. Specifically:
2521
2522 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2523 length.
2524
2525 (b) An item such as \pL+ within a repeated group could cause crashes or
2526 loops.
2527
2528 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2529 "reference to non-existent subpattern" error.
2530
2531 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2532
2533 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2534 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2535
2536 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2537
2538 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2539 character were causing crashes (broken optimization).
2540
2541 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2542 \p or \P) caused a compile-time loop.
2543
2544 28. More problems have arisen in unanchored patterns when CRLF is a valid line
2545 break. For example, the unstudied pattern [\r\n]A does not match the string
2546 "\r\nA" because change 7.0/46 below moves the current point on by two
2547 characters after failing to match at the start. However, the pattern \nA
2548 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2549 the same is true. There doesn't seem any very clean way out of this, but
2550 what I have chosen to do makes the common cases work: PCRE now takes note
2551 of whether there can be an explicit match for \r or \n anywhere in the
2552 pattern, and if so, 7.0/46 no longer applies. As part of this change,
2553 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2554 pattern has explicit CR or LF references.
2555
2556 29. Added (*CR) etc for changing newline setting at start of pattern.
2557
2558
2559 Version 7.2 19-Jun-07
2560 ---------------------
2561
2562 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2563 which is apparently normally available under Windows.
2564
2565 2. Re-jig the pcregrep tests with different newline settings in an attempt
2566 to make them independent of the local environment's newline setting.
2567
2568 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2569
2570 4. Some of the "internals" tests were previously cut out when the link size
2571 was not 2, because the output contained actual offsets. The recent new
2572 "Z" feature of pcretest means that these can be cut out, making the tests
2573 usable with all link sizes.
2574
2575 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2576 stack recursion. This gives a massive performance boost under BSD, but just
2577 a small improvement under Linux. However, it saves one field in the frame
2578 in all cases.
2579
2580 6. Added more features from the forthcoming Perl 5.10:
2581
2582 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2583 recursion call. It refers to the nth most recently opened parentheses.
2584
2585 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2586 to be opened parentheses.
2587
2588 (c) Conditions that refer to capturing parentheses can be specified
2589 relatively, for example, (?(-2)... or (?(+3)...
2590
2591 (d) \K resets the start of the current match so that everything before
2592 is not part of it.
2593
2594 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2595
2596 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2597 reference syntax.
2598
2599 (g) (?| introduces a group in which the numbering of parentheses in each
2600 alternative starts with the same number.
2601
2602 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2603
2604 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2605 PCRE_INFO_JCHANGED.
2606
2607 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2608 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2609 for detecting groups that can match an empty string.
2610
2611 9. A pattern with a very large number of alternatives (more than several
2612 hundred) was running out of internal workspace during the pre-compile
2613 phase, where pcre_compile() figures out how much memory will be needed. A
2614 bit of new cunning has reduced the workspace needed for groups with
2615 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2616 workspace instead of running out of the 4096 that are available.
2617
2618 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2619
2620 11. Applied patch from Google to remove an optimization that didn't quite work.
2621 The report of the bug said:
2622
2623 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2624 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2625 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2626
2627 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2628 it matched the wrong number of bytes.
2629
2630
2631 Version 7.1 24-Apr-07
2632 ---------------------
2633
2634 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2635 that is more "standard", making use of automake and other Autotools. There
2636 is some re-arrangement of the files and adjustment of comments consequent
2637 on this.
2638
2639 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2640 for recursive directory scanning broke on some systems because the files
2641 are not scanned in any specific order and on different systems the order
2642 was different. A call to "sort" has been inserted into RunGrepTest for the
2643 approprate test as a short-term fix. In the longer term there may be an
2644 alternative.
2645
2646 3. I had an email from Eric Raymond about problems translating some of PCRE's
2647 man pages to HTML (despite the fact that I distribute HTML pages, some
2648 people do their own conversions for various reasons). The problems
2649 concerned the use of low-level troff macros .br and .in. I have therefore
2650 removed all such uses from the man pages (some were redundant, some could
2651 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2652 HTML has been updated to handle .nf/.fi and to complain if it encounters
2653 .br or .in.
2654
2655 4. Updated comments in configure.ac that get placed in config.h.in and also
2656 arranged for config.h to be included in the distribution, with the name
2657 config.h.generic, for the benefit of those who have to compile without
2658 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2659
2660 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2661 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2662 makevp.bat for the current PCRE, using the additional files
2663 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2664
2665 6. A Windows user reported a minor discrepancy with test 2, which turned out
2666 to be caused by a trailing space on an input line that had got lost in his
2667 copy. The trailing space was an accident, so I've just removed it.
2668
2669 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2670 that is needed.
2671
2672 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2673 as "const" (a) because they are and (b) because it helps the PHP
2674 maintainers who have recently made a script to detect big data structures
2675 in the php code that should be moved to the .rodata section. I remembered
2676 to update Builducptable as well, so it won't revert if ucptable.h is ever
2677 re-created.
2678
2679 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2680 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2681 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2682 support is not required. This saves 1.5-2K of code, which is important in
2683 some applications.
2684
2685 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2686 so as not to refer to the tables, even though these functions will never be
2687 called when UTF-8 support is disabled. Otherwise there are problems with a
2688 shared library.
2689
2690 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2691
2692 (a) It was defining its arguments as char * instead of void *.
2693
2694 (b) It was assuming that all moves were upwards in memory; this was true
2695 a long time ago when I wrote it, but is no longer the case.
2696
2697 The emulated memove() is provided for those environments that have neither
2698 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2699 is clearly not the case, as these two bugs were recently reported.
2700
2701 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2702 and Detrail to create the HTML documentation, the .txt form of the man
2703 pages, and it removes trailing spaces from listed files. It also creates
2704 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2705 case, it wraps all the #defines with #ifndefs. This script should be run
2706 before "make dist".
2707
2708 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2709 with Unicode property support.
2710
2711 (a) For a maximizing quantifier, if the two different cases of the
2712 character were of different lengths in their UTF-8 codings (there are
2713 some cases like this - I found 11), and the matching function had to
2714 back up over a mixture of the two cases, it incorrectly assumed they
2715 were both the same length.
2716
2717 (b) When PCRE was configured to use the heap rather than the stack for
2718 recursion during matching, it was not correctly preserving the data for
2719 the other case of a UTF-8 character when checking ahead for a match
2720 while processing a minimizing repeat. If the check also involved
2721 matching a wide character, but failed, corruption could cause an
2722 erroneous result when trying to check for a repeat of the original
2723 character.
2724
2725 13. Some tidying changes to the testing mechanism:
2726
2727 (a) The RunTest script now detects the internal link size and whether there
2728 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2729 values substituted by "configure". (The RunGrepTest script already did
2730 this for UTF-8.) The configure.ac script no longer substitutes the
2731 relevant variables.
2732
2733 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2734 with length and offset values. This means that the output is different
2735 for different internal link sizes. Test 2 is skipped for link sizes
2736 other than 2 because of this, bypassing the problem. Unfortunately,
2737 there was also a test in test 3 (the locale tests) that used /B and
2738 failed for link sizes other than 2. Rather than cut the whole test out,
2739 I have added a new /Z option to pcretest that replaces the length and
2740 offset values with spaces. This is now used to make test 3 independent
2741 of link size. (Test 2 will be tidied up later.)
2742
2743 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2744 segmentation fault instead of returning the appropriate error message.
2745
2746 15. In multiline mode when the newline sequence was set to "any", the pattern
2747 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2748 This doesn't seem right; it now treats the CRLF combination as the line
2749 ending, and so does not match in that case. It's only a pattern such as ^$
2750 that would hit this one: something like ^ABC$ would have failed after \r
2751 and then tried again after \r\n.
2752
2753 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2754 in an attempt to make files that differ only in their line terminators
2755 compare equal. This works on Linux.
2756
2757 17. Under certain error circumstances pcregrep might try to free random memory
2758 as it exited. This is now fixed, thanks to valgrind.
2759
2760 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2761 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2762 was because its rules for how to advance for /g after matching an empty
2763 string at the end of a line did not allow for this case. They now check for
2764 it specially.
2765
2766 20. pcretest is supposed to handle patterns and data of any length, by
2767 extending its buffers when necessary. It was getting this wrong when the
2768 buffer for a data line had to be extended.
2769
2770 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2771 CRLF as a newline sequence.
2772
2773 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2774 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2775 I have nevertheless tidied it up.
2776
2777 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2778
2779 24. Added a man page for pcre-config.
2780
2781
2782 Version 7.0 19-Dec-06
2783 ---------------------
2784
2785 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2786 moving to gcc 4.1.1.
2787
2788 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2789 sys/time.h, which is documented as needed for this function. It doesn't
2790 seem to matter on Linux, but it showed up on some releases of OS X.
2791
2792 3. It seems that there are systems where bytes whose values are greater than
2793 127 match isprint() in the "C" locale. The "C" locale should be the
2794 default when a C program starts up. In most systems, only ASCII printing
2795 characters match isprint(). This difference caused the output from pcretest
2796 to vary, making some of the tests fail. I have changed pcretest so that:
2797
2798 (a) When it is outputting text in the compiled version of a pattern, bytes
2799 other than 32-126 are always shown as hex escapes.
2800
2801 (b) When it is outputting text that is a matched part of a subject string,
2802 it does the same, unless a different locale has been set for the match
2803 (using the /L modifier). In this case, it uses isprint() to decide.
2804
2805 4. Fixed a major bug that caused incorrect computation of the amount of memory
2806 required for a compiled pattern when options that changed within the
2807 pattern affected the logic of the preliminary scan that determines the
2808 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2809 that the computed length was too small. The symptoms of this bug were
2810 either the PCRE error "internal error: code overflow" from pcre_compile(),
2811 or a glibc crash with a message such as "pcretest: free(): invalid next
2812 size (fast)". Examples of patterns that provoked this bug (shown in
2813 pcretest format) are:
2814
2815 /(?-x: )/x
2816 /(?x)(?-x: \s*#\s*)/
2817 /((?i)[\x{c0}])/8
2818 /(?i:[\x{c0}])/8
2819
2820 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2821 is now done differently.
2822
2823 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2824 wrapper classes; (b) implement a new function in the C++ scanner that is
2825 more efficient than the old way of doing things because it avoids levels of
2826 recursion in the regex matching; (c) add a paragraph to the documentation
2827 for the FullMatch() function.
2828
2829 6. The escape sequence \n was being treated as whatever was defined as
2830 "newline". Not only was this contrary to the documentation, which states
2831 that \n is character 10 (hex 0A), but it also went horribly wrong when
2832 "newline" was defined as CRLF. This has been fixed.
2833
2834 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2835 was being set to -1 for the "end of line" case (supposedly a value that no
2836 character can have). Though this value is never used (the check for end of
2837 line is "zero bytes in current character"), it caused compiler complaints.
2838 I've changed it to 0xffffffff.
2839
2840 8. In pcre_version.c, the version string was being built by a sequence of
2841 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2842 string (as it is for production releases) called a macro with an empty
2843 argument. The C standard says the result of this is undefined. The gcc
2844 compiler treats it as an empty string (which was what was wanted) but it is
2845 reported that Visual C gives an error. The source has been hacked around to
2846 avoid this problem.
2847
2848 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2849 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2850 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2851 of them did).
2852
2853 10. Originally, pcretest opened its input and output without "b"; then I was
2854 told that "b" was needed in some environments, so it was added for release
2855 5.0 to both the input and output. (It makes no difference on Unix-like
2856 systems.) Later I was told that it is wrong for the input on Windows. I've
2857 now abstracted the modes into two macros, to make it easier to fiddle with
2858 them, and removed "b" from the input mode under Windows.
2859
2860 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2861
2862 12. Added -help and --help to pcretest as an official way of being reminded
2863 of the options.
2864
2865 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2866 and pcrecpp.cc because they annoy compilers at high warning levels.
2867
2868 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2869
2870 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2871 scripts are not C programs :-) and which was not noticed because it works
2872 on Linux.
2873
2874 16. pcretest is supposed to handle any length of pattern and data line (as one
2875 line or as a continued sequence of lines) by extending its input buffer if
2876 necessary. This feature was broken for very long pattern lines, leading to
2877 a string of junk being passed to pcre_compile() if the pattern was longer
2878 than about 50K.
2879
2880 17. I have done a major re-factoring of the way pcre_compile() computes the
2881 amount of memory needed for a compiled pattern. Previously, there was code
2882 that made a preliminary scan of the pattern in order to do this. That was
2883 OK when PCRE was new, but as the facilities have expanded, it has become
2884 harder and harder to keep it in step with the real compile phase, and there
2885 have been a number of bugs (see for example, 4 above). I have now found a
2886 cunning way of running the real compile function in a "fake" mode that
2887 enables it to compute how much memory it would need, while actually only
2888 ever using a few hundred bytes of working memory and without too many
2889 tests of the mode. This should make future maintenance and development
2890 easier. A side effect of this work is that the limit of 200 on the nesting
2891 depth of parentheses has been removed (though this was never a serious
2892 limitation, I suspect). However, there is a downside: pcre_compile() now
2893 runs more slowly than before (30% or more, depending on the pattern). I
2894 hope this isn't a big issue. There is no effect on runtime performance.
2895
2896 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2897 newline (only possible for the last line of a file) and it was a
2898 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2899
2900 19. Added additional timing features to pcretest. (1) The -tm option now times
2901 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2902 separate command line item, by a number that specifies the number of
2903 repeats to use when timing. The default is 50000; this gives better
2904 precision, but takes uncomfortably long for very large patterns.
2905
2906 20. Extended pcre_study() to be more clever in cases where a branch of a
2907 subpattern has no definite first character. For example, (a*|b*)[cd] would
2908 previously give no result from pcre_study(). Now it recognizes that the
2909 first character must be a, b, c, or d.
2910
2911 21. There was an incorrect error "recursive call could loop indefinitely" if
2912 a subpattern (or the entire pattern) that was being tested for matching an
2913 empty string contained only one non-empty item after a nested subpattern.
2914 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2915 incorrectly, because the \d was being skipped in the check.
2916
2917 22. The pcretest program now has a new pattern option /B and a command line
2918 option -b, which is equivalent to adding /B to every pattern. This causes
2919 it to show the compiled bytecode, without the additional information that
2920 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2921 is the same as /B/I).
2922
2923 23. A new optimization is now able automatically to treat some sequences such
2924 as a*b as a*+b. More specifically, if something simple (such as a character
2925 or a simple class like \d) has an unlimited quantifier, and is followed by
2926 something that cannot possibly match the quantified thing, the quantifier
2927 is automatically "possessified".
2928
2929 24. A recursive reference to a subpattern whose number was greater than 39
2930 went wrong under certain circumstances in UTF-8 mode. This bug could also
2931 have affected the operation of pcre_study().
2932
2933 25. Realized that a little bit of performance could be had by replacing
2934 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2935
2936 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2937
2938 27. Possessive quantifiers such as a++ were previously implemented by turning
2939 them into atomic groups such as ($>a+). Now they have their own opcodes,
2940 which improves performance. This includes the automatically created ones
2941 from 23 above.
2942
2943 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2944 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2945 the first matched character to be a colon. This applied both to named and
2946 numbered groups.
2947
2948 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2949
2950 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2951 building PCRE on Windows easier, so I have included it in the distribution.
2952
2953 31. There is now a check in pcretest against a ridiculously large number being
2954 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2955 loop, the loop is abandoned.
2956
2957 32. Forward references to subpatterns in conditions such as (?(2)...) where
2958 subpattern 2 is defined later cause pcre_compile() to search forwards in
2959 the pattern for the relevant set of parentheses. This search went wrong
2960 when there were unescaped parentheses in a character class, parentheses
2961 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2962
2963 33. "Subroutine" calls and backreferences were previously restricted to
2964 referencing subpatterns earlier in the regex. This restriction has now
2965 been removed.
2966
2967 34. Added a number of extra features that are going to be in Perl 5.10. On the
2968 whole, these are just syntactic alternatives for features that PCRE had
2969 previously implemented using the Python syntax or my own invention. The
2970 other formats are all retained for compatibility.
2971
2972 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2973 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2974 also .NET compatible.
2975
2976 (b) A recursion or subroutine call to a named group can now be defined as
2977 (?&name) as well as (?P>name).
2978
2979 (c) A backreference to a named group can now be defined as \k<name> or
2980 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2981 5.10, are also .NET compatible.
2982
2983 (d) A conditional reference to a named group can now use the syntax
2984 (?(<name>) or (?('name') as well as (?(name).
2985
2986 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2987 groups (named and numbered) that are never evaluated inline, but can be
2988 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2989 is always false. There may be only one alternative in such a group.
2990
2991 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2992 as the simple (?(R). The condition is true only if the most recent
2993 recursion is that of the given number or name. It does not search out
2994 through the entire recursion stack.
2995
2996 (g) The escape \gN or \g{N} has been added, where N is a positive or
2997 negative number, specifying an absolute or relative reference.
2998
2999 35. Tidied to get rid of some further signed/unsigned compiler warnings and
3000 some "unreachable code" warnings.
3001
3002 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
3003 things, this adds five new scripts.
3004
3005 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
3006 There were also incompatibilities regarding the handling of \Q..\E inside
3007 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
3008 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
3009
3010 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
3011 matches an empty string, and forcibly breaks the loop. There were bugs in
3012 this code in non-simple cases. For a pattern such as ^(a()*)* matched
3013 against aaaa the result was just "a" rather than "aaaa", for example. Two
3014 separate and independent bugs (that affected different cases) have been
3015 fixed.
3016
3017 39. Refactored the code to abolish the use of different opcodes for small
3018 capturing bracket numbers. This is a tidy that I avoided doing when I
3019 removed the limit on the number of capturing brackets for 3.5 back in 2001.
3020 The new approach is not only tidier, it makes it possible to reduce the
3021 memory needed to fix the previous bug (38).
3022
3023 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
3024 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
3025 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
3026 mode.
3027
3028 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
3029 report.
3030
3031 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
3032 copy construction and assignment in the C++ wrapper.
3033
3034 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
3035 couple of bugs that could have given wrong results in the "--newline=crlf"
3036 case.
3037
3038 44. Added a number of casts and did some reorganization of signed/unsigned int
3039 variables following suggestions from Dair Grant. Also renamed the variable
3040 "this" as "item" because it is a C++ keyword.
3041
3042 45. Arranged for dftables to add
3043
3044 #include "pcre_internal.h"
3045
3046 to pcre_chartables.c because without it, gcc 4.x may remove the array
3047 definition from the final binary if PCRE is built into a static library and
3048 dead code stripping is activated.
3049
3050 46. For an unanchored pattern, if a match attempt fails at the start of a
3051 newline sequence, and the newline setting is CRLF or ANY, and the next two
3052 characters are CRLF, advance by two characters instead of one.
3053
3054
3055 Version 6.7 04-Jul-06
3056 ---------------------
3057
3058 1. In order to handle tests when input lines are enormously long, pcretest has
3059 been re-factored so that it automatically extends its buffers when
3060 necessary. The code is crude, but this _is_ just a test program. The
3061 default size has been increased from 32K to 50K.
3062
3063 2. The code in pcre_study() was using the value of the re argument before
3064 testing it for NULL. (Of course, in any sensible call of the function, it
3065 won't be NULL.)
3066
3067 3. The memmove() emulation function in pcre_internal.h, which is used on
3068 systems that lack both memmove() and bcopy() - that is, hardly ever -
3069 was missing a "static" storage class specifier.
3070
3071 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
3072 containing an extended class (one that cannot be represented by a bitmap
3073 because it contains high-valued characters or Unicode property items, e.g.
3074 [\pZ]). Almost always one would set UTF-8 mode when processing such a
3075 pattern, but PCRE should not loop if you do not (it no longer does).
3076 [Detail: two cases were found: (a) a repeated subpattern containing an
3077 extended class; (b) a recursive reference to a subpattern that followed a
3078 previous extended class. It wasn't skipping over the extended class
3079 correctly when UTF-8 mode was not set.]
3080
3081 5. A negated single-character class was not being recognized as fixed-length
3082 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
3083 compile error "lookbehind assertion is not fixed length".
3084
3085 6. The RunPerlTest auxiliary script was showing an unexpected difference
3086 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
3087 write a Perl script that can interpret lines of an input file either as
3088 byte characters or as UTF-8, which is what "perltest" was being required to
3089 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
3090 can't do is switch easily at run time between having the "use utf8;" pragma
3091 or not. In the end, I fudged it by using the RunPerlTest script to insert
3092 "use utf8;" explicitly for the UTF-8 tests.
3093
3094 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
3095 the end of the subject string, contrary to the documentation and to what
3096 Perl does. This was true of both matching functions. Now it matches only at
3097 the start of the subject and immediately after *internal* newlines.
3098
3099 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
3100 a pointer to an int instead of a pointer to an unsigned long int. This
3101 caused problems on 64-bit systems.
3102
3103 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
3104 instance of the 'standard' template library not being so standard".
3105
3106 10. There was no check on the number of named subpatterns nor the maximum
3107 length of a subpattern name. The product of these values is used to compute
3108 the size of the memory block for a compiled pattern. By supplying a very
3109 long subpattern name and a large number of named subpatterns, the size
3110 computation could be caused to overflow. This is now prevented by limiting
3111 the length of names to 32 characters, and the number of named subpatterns
3112 to 10,000.
3113
3114 11. Subpatterns that are repeated with specific counts have to be replicated in
3115 the compiled pattern. The size of memory for this was computed from the
3116 length of the subpattern and the repeat count. The latter is limited to
3117 65535, but there was no limit on the former, meaning that integer overflow
3118 could in principle occur. The compiled length of a repeated subpattern is
3119 now limited to 30,000 bytes in order to prevent this.
3120
3121 12. Added the optional facility to have named substrings with the same name.
3122
3123 13. Added the ability to use a named substring as a condition, using the
3124 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
3125 are numbers (not recommended). Forward references are permitted.
3126
3127 14. Added forward references in named backreferences (if you see what I mean).
3128
3129 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
3130 pattern could run off the end of the subject. For example, the pattern
3131 "(?s)(.{1,5})"8 did this with the subject "ab".
3132
3133 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
3134 PCRE_CASELESS was set when matching characters that were quantified with ?
3135 or *.
3136
3137 17. A character class other than a single negated character that had a minimum
3138 but no maximum quantifier - for example [ab]{6,} - was not handled
3139 correctly by pce_dfa_exec(). It would match only one character.
3140
3141 18. A valid (though odd) pattern that looked like a POSIX character
3142 class but used an invalid character after [ (for example [[,abc,]]) caused
3143 pcre_compile() to give the error "Failed: internal error: code overflow" or
3144 in some cases to crash with a glibc free() error. This could even happen if
3145 the pattern terminated after [[ but there just happened to be a sequence of
3146 letters, a binary zero, and a closing ] in the memory that followed.
3147
3148 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
3149 over the years. Originally (before any Unicode support), just the bottom 8
3150 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
3151 output from "man perlunicode" includes this:
3152
3153 The regular expression compiler produces polymorphic opcodes. That
3154 is, the pattern adapts to the data and automatically switches to
3155 the Unicode character scheme when presented with Unicode data--or
3156 instead uses a traditional byte scheme when presented with byte
3157 data.
3158
3159 Sadly, a wide octal escape does not cause a switch, and in a string with
3160 no other multibyte characters, these octal escapes are treated as before.
3161 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
3162 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
3163 Unicode string.
3164
3165 I have not perpetrated such confusion in PCRE. Up till now, it took just
3166 the bottom 8 bits, as in old Perl. I have now made octal escapes with
3167 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
3168 translate to the appropriate multibyte character.
3169
3170 29. Applied some refactoring to reduce the number of warnings from Microsoft
3171 and Borland compilers. This has included removing the fudge introduced
3172 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
3173 a warning about an unused variable.
3174
3175 21. PCRE has not included VT (character 0x0b) in the set of whitespace
3176 characters since release 4.0, because Perl (from release 5.004) does not.
3177 [Or at least, is documented not to: some releases seem to be in conflict
3178 with the documentation.] However, when a pattern was studied with
3179 pcre_study() and all its branches started with \s, PCRE still included VT
3180 as a possible starting character. Of course, this did no harm; it just
3181 caused an unnecessary match attempt.
3182
3183 22. Removed a now-redundant internal flag bit that recorded the fact that case
3184 dependency changed within the pattern. This was once needed for "required
3185 byte" processing, but is no longer used. This recovers a now-scarce options
3186 bit. Also moved the least significant internal flag bit to the most-
3187 significant bit of the word, which was not previously used (hangover from
3188 the days when it was an int rather than a uint) to free up another bit for
3189 the future.
3190
3191 23. Added support for CRLF line endings as well as CR and LF. As well as the
3192 default being selectable at build time, it can now be changed at runtime
3193 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
3194 specify that it is scanning data with non-default line endings.
3195
3196 24. Changed the definition of CXXLINK to make it agree with the definition of
3197 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
3198
3199 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
3200 recursions. This makes a big different to stack usage for some patterns.
3201
3202 26. If a subpattern containing a named recursion or subroutine reference such
3203 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
3204 the space required for the compiled pattern went wrong and gave too small a
3205 value. Depending on the environment, this could lead to "Failed: internal
3206 error: code overflow at offset 49" or "glibc detected double free or
3207 corruption" errors.
3208
3209 27. Applied patches from Google (a) to support the new newline modes and (b) to
3210 advance over multibyte UTF-8 characters in GlobalReplace.
3211
3212 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
3213 difference for some implementation of PCRE in some Windows version.
3214
3215 29. Added some extra testing facilities to pcretest:
3216
3217 \q<number> in a data line sets the "match limit" value
3218 \Q<number> in a data line sets the "match recursion limt" value
3219 -S <number> sets the stack size, where <number> is in megabytes
3220
3221 The -S option isn't available for Windows.
3222
3223
3224 Version 6.6 06-Feb-06
3225 ---------------------
3226
3227 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
3228 in pcreposix.h. I have copied the definition from pcre.h.
3229
3230 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
3231 because pcre.h is no longer a built file.
3232
3233 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
3234 not normally included in the compiled code.
3235
3236
3237 Version 6.5 01-Feb-06
3238 ---------------------
3239
3240 1. When using the partial match feature with pcre_dfa_exec(), it was not
3241 anchoring the second and subsequent partial matches at the new starting
3242 point. This could lead to incorrect results. For example, with the pattern
3243 /1234/, partially matching against "123" and then "a4" gave a match.
3244
3245 2. Changes to pcregrep:
3246
3247 (a) All non-match returns from pcre_exec() were being treated as failures
3248 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
3249 error message is output. Some extra information is given for the
3250 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
3251 probably the only errors that are likely to be caused by users (by
3252 specifying a regex that has nested indefinite repeats, for instance).
3253 If there are more than 20 of these errors, pcregrep is abandoned.
3254
3255 (b) A binary zero was treated as data while matching, but terminated the
3256 output line if it was written out. This has been fixed: binary zeroes
3257 are now no different to any other data bytes.
3258
3259 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3260 used to set a locale for matching. The --locale=xxxx long option has
3261 been added (no short equivalent) to specify a locale explicitly on the
3262 pcregrep command, overriding the environment variables.
3263
3264 (d) When -B was used with -n, some line numbers in the output were one less
3265 than they should have been.
3266
3267 (e) Added the -o (--only-matching) option.
3268
3269 (f) If -A or -C was used with -c (count only), some lines of context were
3270 accidentally printed for the final match.
3271
3272 (g) Added the -H (--with-filename) option.
3273
3274 (h) The combination of options -rh failed to suppress file names for files
3275 that were found from directory arguments.
3276
3277 (i) Added the -D (--devices) and -d (--directories) options.
3278
3279 (j) Added the -F (--fixed-strings) option.
3280
3281 (k) Allow "-" to be used as a file name for -f as well as for a data file.
3282
3283 (l) Added the --colo(u)r option.
3284
3285 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3286 is not present by default.
3287
3288 3. A nasty bug was discovered in the handling of recursive patterns, that is,
3289 items such as (?R) or (?1), when the recursion could match a number of
3290 alternatives. If it matched one of the alternatives, but subsequently,
3291 outside the recursion, there was a failure, the code tried to back up into
3292 the recursion. However, because of the way PCRE is implemented, this is not
3293 possible, and the result was an incorrect result from the match.
3294
3295 In order to prevent this happening, the specification of recursion has
3296 been changed so that all such subpatterns are automatically treated as
3297 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3298
3299 4. I had overlooked the fact that, in some locales, there are characters for
3300 which isalpha() is true but neither isupper() nor islower() are true. In
3301 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3302 and ordfeminine) are like this. This affected the treatment of \w and \W
3303 when they appeared in character classes, but not when they appeared outside
3304 a character class. The bit map for "word" characters is now created
3305 separately from the results of isalnum() instead of just taking it from the
3306 upper, lower, and digit maps. (Plus the underscore character, of course.)
3307
3308 5. The above bug also affected the handling of POSIX character classes such as
3309 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3310 permanent tables. Instead, the bit maps for such a class were previously
3311 created as the appropriate unions of the upper, lower, and digit bitmaps.
3312 Now they are created by subtraction from the [[:word:]] class, which has
3313 its own bitmap.
3314
3315 6. The [[:blank:]] character class matches horizontal, but not vertical space.
3316 It is created by subtracting the vertical space characters (\x09, \x0a,
3317 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3318 subtraction was done in the overall bitmap for a character class, meaning
3319 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3320 be recognized. This bug has been fixed.
3321
3322 7. Patches from the folks at Google:
3323
3324 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3325 real life, but is still worth protecting against".
3326
3327 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3328 regular expressions".
3329
3330 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3331 have it.
3332
3333 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3334 "configure" and the latter not, in order to fix a problem somebody had
3335 with compiling the Arg class on HP-UX.
3336
3337 (e) Improve the error-handling of the C++ wrapper a little bit.
3338
3339 (f) New tests for checking recursion limiting.
3340
3341 8. The pcre_memmove() function, which is used only if the environment does not
3342 have a standard memmove() function (and is therefore rarely compiled),
3343 contained two bugs: (a) use of int instead of size_t, and (b) it was not
3344 returning a result (though PCRE never actually uses the result).
3345
3346 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3347 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3348 returned instead of calling malloc() with an overflowing number that would
3349 most likely cause subsequent chaos.
3350
3351 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3352
3353 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3354 with this option is matched, the nmatch and pmatch options of regexec() are
3355 ignored.
3356
3357 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3358 provided in case anyone wants to the the POSIX interface with UTF-8
3359 strings.
3360
3361 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3362 C++ linking (needed for some HP-UX environments).
3363
3364 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3365 (unused parameter) and in the pcre_printint() function (omitted "default"
3366 switch label when the default is to do nothing).
3367
3368 15. Added some code to make it possible, when PCRE is compiled as a C++
3369 library, to replace subject pointers for pcre_exec() with a smart pointer
3370 class, thus making it possible to process discontinuous strings.
3371
3372 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3373 much the same function. They were added by different people who were trying
3374 to make PCRE easy to compile on non-Unix systems. It has been suggested
3375 that PCRE_EXPORT be abolished now that there is more automatic apparatus
3376 for compiling on Windows systems. I have therefore replaced it with
3377 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3378 defaults to "extern" for C or "extern C" for C++, which works fine on
3379 Unix-like systems. It is now possible to override the value of PCRE_DATA_
3380 SCOPE with something explicit in config.h. In addition:
3381
3382 (a) pcreposix.h still had just "extern" instead of either of these macros;
3383 I have replaced it with PCRE_DATA_SCOPE.
3384
3385 (b) Functions such as _pcre_xclass(), which are internal to the library,
3386 but external in the C sense, all had PCRE_EXPORT in their definitions.
3387 This is apparently wrong for the Windows case, so I have removed it.
3388 (It makes no difference on Unix-like systems.)
3389
3390 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3391 of recursive calls to match(). This is different to MATCH_LIMIT because
3392 that limits the total number of calls to match(), not all of which increase
3393 the depth of recursion. Limiting the recursion depth limits the amount of
3394 stack (or heap if NO_RECURSE is set) that is used. The default can be set
3395 when PCRE is compiled, and changed at run time. A patch from Google adds
3396 this functionality to the C++ interface.
3397
3398 18. Changes to the handling of Unicode character properties:
3399
3400 (a) Updated the table to Unicode 4.1.0.
3401
3402 (b) Recognize characters that are not in the table as "Cn" (undefined).
3403
3404 (c) I revised the way the table is implemented to a much improved format
3405 which includes recognition of ranges. It now supports the ranges that
3406 are defined in UnicodeData.txt, and it also amalgamates other
3407 characters into ranges. This has reduced the number of entries in the
3408 table from around 16,000 to around 3,000, thus reducing its size
3409 considerably. I realized I did not need to use a tree structure after
3410 all - a binary chop search is just as efficient. Having reduced the
3411 number of entries, I extended their size from 6 bytes to 8 bytes to
3412 allow for more data.
3413
3414 (d) Added support for Unicode script names via properties such as \p{Han}.
3415
3416 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3417 matching that character.
3418
3419 20. When matching a repeated Unicode property with a minimum greater than zero,
3420 (for example \pL{2,}), PCRE could look past the end of the subject if it
3421 reached it while seeking the minimum number of characters. This could
3422 happen only if some of the characters were more than one byte long, because
3423 there is a check for at least the minimum number of bytes.
3424
3425 21. Refactored the implementation of \p and \P so as to be more general, to
3426 allow for more different types of property in future. This has changed the
3427 compiled form incompatibly. Anybody with saved compiled patterns that use
3428 \p or \P will have to recompile them.
3429
3430 22. Added "Any" and "L&" to the supported property types.
3431
3432 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3433 but give a compile time error if the value is greater than 0xff.
3434
3435 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3436 accidentally not being installed or uninstalled.
3437
3438 25. The pcre.h file was built from pcre.h.in, but the only changes that were
3439 made were to insert the current release number. This seemed silly, because
3440 it made things harder for people building PCRE on systems that don't run
3441 "configure". I have turned pcre.h into a distributed file, no longer built
3442 by "configure", with the version identification directly included. There is
3443 no longer a pcre.h.in file.
3444
3445 However, this change necessitated a change to the pcre-config script as
3446 well. It is built from pcre-config.in, and one of the substitutions was the
3447 release number. I have updated configure.ac so that ./configure now finds
3448 the release number by grepping pcre.h.
3449
3450 26. Added the ability to run the tests under valgrind.
3451
3452
3453 Version 6.4 05-Sep-05
3454 ---------------------
3455
3456 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3457 "--" to be printed when multiple files were scanned, even when none of the
3458 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3459 consider it to be a bug, and have restored the previous behaviour.
3460
3461 2. A couple of code tidies to get rid of compiler warnings.
3462
3463 3. The pcretest program used to cheat by referring to symbols in the library
3464 whose names begin with _pcre_. These are internal symbols that are not
3465 really supposed to be visible externally, and in some environments it is
3466 possible to suppress them. The cheating is now confined to including
3467 certain files from the library's source, which is a bit cleaner.
3468
3469 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3470 file's purpose clearer.
3471
3472 5. Reorganized pcre_ucp_findchar().
3473
3474
3475 Version 6.3 15-Aug-05
3476 ---------------------
3477
3478 1. The file libpcre.pc.in did not have general read permission in the tarball.
3479
3480 2. There were some problems when building without C++ support:
3481
3482 (a) If C++ support was not built, "make install" and "make test" still
3483 tried to test it.
3484
3485 (b) There were problems when the value of CXX was explicitly set. Some
3486 changes have been made to try to fix these, and ...
3487
3488 (c) --disable-cpp can now be used to explicitly disable C++ support.
3489
3490 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3491 backslash in a target when C++ was disabled. This confuses some
3492 versions of "make", apparently. Using an intermediate variable solves
3493 this. (Same for CPP_LOBJ.)
3494
3495 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3496 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3497 necessary on certain architectures.
3498
3499 4. Added a setting of -export-symbols-regex to the link command to remove
3500 those symbols that are exported in the C sense, but actually are local
3501 within the library, and not documented. Their names all begin with
3502 "_pcre_". This is not a perfect job, because (a) we have to except some
3503 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3504 available (and never for static libraries). I have made a note to try to
3505 find a way round (a) in the future.
3506
3507
3508 Version 6.2 01-Aug-05
3509 ---------------------
3510
3511 1. There was no test for integer overflow of quantifier values. A construction
3512 such as {1111111111111111} would give undefined results. What is worse, if
3513 a minimum quantifier for a parenthesized subpattern overflowed and became
3514 negative, the calculation of the memory size went wrong. This could have
3515 led to memory overwriting.
3516
3517 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3518
3519 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3520 operating environments where this matters.
3521
3522 4. Applied Giuseppe Maxia's patch to add additional features for controlling
3523 PCRE options from within the C++ wrapper.
3524
3525 5. Named capturing subpatterns were not being correctly counted when a pattern
3526 was compiled. This caused two problems: (a) If there were more than 100
3527 such subpatterns, the calculation of the memory needed for the whole
3528 compiled pattern went wrong, leading to an overflow error. (b) Numerical
3529 back references of the form \12, where the number was greater than 9, were
3530 not recognized as back references, even though there were sufficient
3531 previous subpatterns.
3532
3533 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3534 versions of gcc, e.g. 2.95.4.
3535
3536
3537 Version 6.1 21-Jun-05
3538 ---------------------
3539
3540 1. There was one reference to the variable "posix" in pcretest.c that was not
3541 surrounded by "#if !defined NOPOSIX".
3542
3543 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3544 the cross-check on the old pcre_info() function, for the benefit of the
3545 cut-down version of PCRE that is currently imported into Exim.
3546
3547 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3548 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3549 patterns that start (?i) but I don't think that matters. The use of (?i) is
3550 just an example; this all applies to the other options as well.
3551
3552 4. Since libtool seems to echo the compile commands it is issuing, the output
3553 from "make" can be reduced a bit by putting "@" in front of each libtool
3554 compile command.
3555
3556 5. Patch from the folks at Google for configure.in to be a bit more thorough
3557 in checking for a suitable C++ installation before trying to compile the
3558 C++ stuff. This should fix a reported problem when a compiler was present,
3559 but no suitable headers.
3560
3561 6. The man pages all had just "PCRE" as their title. I have changed them to
3562 be the relevant file name. I have also arranged that these names are
3563 retained in the file doc/pcre.txt, which is a concatenation in text format
3564 of all the man pages except the little individual ones for each function.
3565
3566 7. The NON-UNIX-USE file had not been updated for the different set of source
3567 files that come with release 6. I also added a few comments about the C++
3568 wrapper.
3569
3570
3571 Version 6.0 07-Jun-05
3572 ---------------------
3573
3574 1. Some minor internal re-organization to help with my DFA experiments.
3575
3576 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3577 didn't matter for the library itself when fully configured, but did matter
3578 when compiling without UCP support, or within Exim, where the ucp files are
3579 not imported.
3580
3581 3. Refactoring of the library code to split up the various functions into
3582 different source modules. The addition of the new DFA matching code (see
3583 below) to a single monolithic source would have made it really too
3584 unwieldy, quite apart from causing all the code to be include in a
3585 statically linked application, when only some functions are used. This is
3586 relevant even without the DFA addition now that patterns can be compiled in
3587 one application and matched in another.
3588
3589 The downside of splitting up is that there have to be some external
3590 functions and data tables that are used internally in different modules of
3591 the library but which are not part of the API. These have all had their
3592 names changed to start with "_pcre_" so that they are unlikely to clash
3593 with other external names.
3594
3595 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3596 a different (DFA) algorithm. Although it is slower than the original
3597 function, it does have some advantages for certain types of matching
3598 problem.
3599
3600 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3601 including restarting after a partial match.
3602
3603 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3604 defined when compiling for Windows was sent to me. I have put it into the
3605 code, though I have no means of testing or verifying it.
3606
3607 7. Added the pcre_refcount() auxiliary function.
3608
3609 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3610 match before or at the first newline in the subject string. In pcretest,
3611 the /f option on a pattern can be used to set this.
3612
3613 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3614 would behave wrongly. This has been present in PCRE since release 4.0.
3615
3616 10. A number of changes to the pcregrep command:
3617
3618 (a) Refactored how -x works; insert ^(...)$ instead of setting
3619 PCRE_ANCHORED and checking the length, in preparation for adding
3620 something similar for -w.
3621
3622 (b) Added the -w (match as a word) option.
3623
3624 (c) Refactored the way lines are read and buffered so as to have more
3625 than one at a time available.
3626
3627 (d) Implemented a pcregrep test script.
3628
3629 (e) Added the -M (multiline match) option. This allows patterns to match
3630 over several lines of the subject. The buffering ensures that at least
3631 8K, or the rest of the document (whichever is the shorter) is available
3632 for matching (and similarly the previous 8K for lookbehind assertions).
3633
3634 (f) Changed the --help output so that it now says
3635
3636 -w, --word-regex(p)
3637
3638 instead of two lines, one with "regex" and the other with "regexp"
3639 because that confused at least one person since the short forms are the
3640 same. (This required a bit of code, as the output is generated
3641 automatically from a table. It wasn't just a text change.)
3642
3643 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3644 option but starts with a hyphen. Could be a pattern or a path name
3645 starting with a hyphen, for instance.
3646
3647 (h) "-" can be given as a file name to represent stdin.
3648
3649 (i) When file names are being printed, "(standard input)" is used for
3650 the standard input, for compatibility with GNU grep. Previously
3651 "<stdin>" was used.
3652
3653 (j) The option --label=xxx can be used to supply a name to be used for
3654 stdin when file names are being printed. There is no short form.
3655
3656 (k) Re-factored the options decoding logic because we are going to add
3657 two more options that take data. Such options can now be given in four
3658 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3659
3660 (l) Added the -A, -B, and -C options for requesting that lines of context
3661 around matches be printed.
3662
3663 (m) Added the -L option to print the names of files that do not contain
3664 any matching lines, that is, the complement of -l.
3665
3666 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3667 continue to scan other files.
3668
3669 (o) The -s option was incorrectly implemented. For compatibility with other
3670 greps, it now suppresses the error message for a non-existent or non-
3671 accessible file (but not the return code). There is a new option called
3672 -q that suppresses the output of matching lines, which was what -s was
3673 previously doing.
3674
3675 (p) Added --include and --exclude options to specify files for inclusion
3676 and exclusion when recursing.
3677
3678 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3679 Hopefully, it now does.
3680
3681 12. Missing cast in pcre_study().
3682
3683 13. Added an "uninstall" target to the makefile.
3684
3685 14. Replaced "extern" in the function prototypes in Makefile.in with
3686 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3687 world, but is set differently for Windows.
3688
3689 15. Added a second compiling function called pcre_compile2(). The only
3690 difference is that it has an extra argument, which is a pointer to an
3691 integer error code. When there is a compile-time failure, this is set
3692 non-zero, in addition to the error test pointer being set to point to an
3693 error message. The new argument may be NULL if no error number is required
3694 (but then you may as well call pcre_compile(), which is now just a
3695 wrapper). This facility is provided because some applications need a
3696 numeric error indication, but it has also enabled me to tidy up the way
3697 compile-time errors are handled in the POSIX wrapper.
3698
3699 16. Added VPATH=.libs to the makefile; this should help when building with one
3700 prefix path and installing with another. (Or so I'm told by someone who
3701 knows more about this stuff than I do.)
3702
3703 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3704 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3705 match everything, including newlines. This is not POSIX-compatible, but
3706 somebody wanted the feature. From pcretest it can be activated by using
3707 both the P and the s flags.
3708
3709 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3710
3711 19. libpcre.pc was being incorrectly installed as executable.
3712
3713 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3714 it now also looks for '\r' so that it will work unmodified on Windows.
3715
3716 21. Added Google's contributed C++ wrapper to the distribution.
3717
3718 22. Added some untidy missing memory free() calls in pcretest, to keep
3719 Electric Fence happy when testing.
3720
3721
3722
3723 Version 5.0 13-Sep-04
3724 ---------------------
3725
3726 1. Internal change: literal characters are no longer packed up into items
3727 containing multiple characters in a single byte-string. Each character
3728 is now matched using a separate opcode. However, there may be more than one
3729 byte in the character in UTF-8 mode.
3730
3731 2. The pcre_callout_block structure has two new fields: pattern_position and
3732 next_item_length. These contain the offset in the pattern to the next match
3733 item, and its length, respectively.
3734
3735 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3736 insertion of callouts before each pattern item. Added the /C option to
3737 pcretest to make use of this.
3738
3739 4. On the advice of a Windows user, the lines
3740
3741 #if defined(_WIN32) || defined(WIN32)
3742 _setmode( _fileno( stdout ), 0x8000 );
3743 #endif /* defined(_WIN32) || defined(WIN32) */
3744
3745 have been added to the source of pcretest. This apparently does useful
3746 magic in relation to line terminators.
3747
3748 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3749 for the benefit of those environments where the "b" makes a difference.
3750
3751 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3752 to know about it. I have put a hack into configure.in that adds in code
3753 to set GCC=yes if CC=icc. This seems to end up at a point in the
3754 generated configure script that is early enough to affect the setting of
3755 compiler options, which is what is needed, but I have no means of testing
3756 whether it really works. (The user who reported this had patched the
3757 generated configure script, which of course I cannot do.)
3758
3759 LATER: After change 22 below (new libtool files), the configure script
3760 seems to know about icc (and also ecc). Therefore, I have commented out
3761 this hack in configure.in.
3762
3763 7. Added support for pkg-config (2 patches were sent in).
3764
3765 8. Negated POSIX character classes that used a combination of internal tables
3766 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3767 [[:^ascii]]. Typically, they would match almost any characters. The other
3768 POSIX classes were not broken in this way.
3769
3770 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3771 to find the match, as PCRE was deluded into thinking that the match had to
3772 start at the start point or following a newline. The same bug applied to
3773 patterns with negative forward assertions or any backward assertions
3774 preceding ".*" at the start, unless the pattern required a fixed first
3775 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3776
3777 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3778 starting at the last subject character, bytes beyond the end of the subject
3779 string were read.
3780
3781 11. Renamed the variable "class" as "classbits" to make life easier for C++
3782 users. (Previously there was a macro definition, but it apparently wasn't
3783 enough.)
3784
3785 12. Added the new field "tables" to the extra data so that tables can be passed
3786 in at exec time, or the internal tables can be re-selected. This allows
3787 a compiled regex to be saved and re-used at a later time by a different
3788 program that might have everything at different addresses.
3789
3790 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3791 -R library as well as a -L library.
3792
3793 14. The debugging options of pcretest (-d on the command line or D on a
3794 pattern) showed incorrect output for anything following an extended class
3795 that contained multibyte characters and which was followed by a quantifier.
3796
3797 15. Added optional support for general category Unicode character properties
3798 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3799 support. It adds about 90K to the size of the library. The meanings of the
3800 inbuilt class escapes such as \d and \s have NOT been changed.
3801
3802 16. Updated pcredemo.c to include calls to free() to release the memory for the
3803 compiled pattern.
3804
3805 17. The generated file chartables.c was being created in the source directory
3806 instead of in the building directory. This caused the build to fail if the
3807 source directory was different from the building directory, and was
3808 read-only.
3809
3810 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3811 file. No doubt somebody will tell me if they don't make sense... Also added
3812 Dan Mooney's comments about building on OpenVMS.
3813
3814 19. Added support for partial matching via the PCRE_PARTIAL option for
3815 pcre_exec() and the \P data escape in pcretest.
3816
3817 20. Extended pcretest with 3 new pattern features:
3818
3819 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3820 write the compiled pattern to the file whose name is "rest-of-line".
3821 This is a straight binary dump of the data, with the saved pointer to
3822 the character tables forced to be NULL. The study data, if any, is
3823 written too. After writing, pcretest reads a new pattern.
3824
3825 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3826 compiled pattern from the given file. There must not be any
3827 occurrences of "<" in the file name (pretty unlikely); if there are,
3828 pcretest will instead treat the initial "<" as a pattern delimiter.
3829 After reading in the pattern, pcretest goes on to read data lines as
3830 usual.
3831
3832 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3833 and 16-bit fields in a compiled pattern, to simulate a pattern that
3834 was compiled on a host of opposite endianness.
3835
3836 21. The pcre-exec() function can now cope with patterns that were compiled on
3837 hosts of opposite endianness, with this restriction:
3838
3839 As for any compiled expression that is saved and used later, the tables
3840 pointer field cannot be preserved; the extra_data field in the arguments
3841 to pcre_exec() should be used to pass in a tables address if a value
3842 other than the default internal tables were used at compile time.
3843
3844 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3845 now diagnosed as an error. Previously, most of the time, a negative number
3846 would have been treated as zero, but if in addition "ovector" was passed as
3847 NULL, a crash could occur.
3848
3849 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3850 new versions from the libtool 1.5 distribution (the last one is a copy of
3851 a file called libtool.m4). This seems to have fixed the need to patch
3852 "configure" to support Darwin 1.3 (which I used to do). However, I still
3853 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3854 workstation).
3855
3856 24. Changed the PCRE licence to be the more standard "BSD" licence.
3857
3858
3859 Version 4.5 01-Dec-03
3860 ---------------------
3861
3862 1. There has been some re-arrangement of the code for the match() function so
3863 that it can be compiled in a version that does not call itself recursively.
3864 Instead, it keeps those local variables that need separate instances for
3865 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3866 needs to "recurse". Keeping track of where control must go is done by means
3867 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3868 hide most of the details from the main code, and operates only if
3869 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3870 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3871 operating.
3872
3873 To make it easier for callers to provide specially tailored get/free
3874 functions for this usage, two new functions, pcre_stack_malloc, and
3875 pcre_stack_free, are used. They are always called in strict stacking order,
3876 and the size of block requested is always the same.
3877
3878 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3879 PCRE has been compiled to use the stack or the heap for recursion. The
3880 -C option of pcretest uses this to show which version is compiled.
3881
3882 A new data escape \S, is added to pcretest; it causes the amounts of store
3883 obtained and freed by both kinds of malloc/free at match time to be added
3884 to the output.
3885
3886 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3887 what's available on my current Linux desktop machine.
3888
3889 3. When matching a UTF-8 string, the test for a valid string at the start has
3890 been extended. If start_offset is not zero, PCRE now checks that it points
3891 to a byte that is the start of a UTF-8 character. If not, it returns
3892 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3893 this is necessary because there may be backward assertions in the pattern.
3894 When matching the same subject several times, it may save resources to use
3895 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3896
3897 4. The code for checking the validity of UTF-8 strings has been tightened so
3898 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3899 containing "overlong sequences".
3900
3901 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3902 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3903 should have been "&", but it just so happened that all the cases this let
3904 through by mistake were picked up later in the function.
3905
3906 6. I had used a variable called "isblank" - this is a C99 function, causing
3907 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3908
3909 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3910 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3911 so the version is shown; (c) stop "make test" echoing "./RunTest".
3912
3913 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3914
3915 9. The prototype for memmove() for systems that don't have it was using
3916 size_t, but the inclusion of the header that defines size_t was later. I've
3917 moved the #includes for the C headers earlier to avoid this.
3918
3919 10. Added some adjustments to the code to make it easier to compiler on certain
3920 special systems:
3921
3922 (a) Some "const" qualifiers were missing.
3923 (b) Added the macro EXPORT before all exported functions; by default this
3924 is defined to be empty.
3925 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3926 that it reads its output file name as an argument instead of writing
3927 to the standard output and assuming this can be redirected.
3928
3929 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3930 class containing characters with values greater than 255, PCRE compilation
3931 went into a loop.
3932
3933 12. A recursive reference to a subpattern that was within another subpattern
3934 that had a minimum quantifier of zero caused PCRE to crash. For example,
3935 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3936 recursion. If the recursively-called subpattern itself had a zero repeat,
3937 that was OK.
3938
3939 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3940 buffer into which it was copied (for escape processing) was still set at
3941 1024, so long lines caused crashes.
3942
3943 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3944 "internal error: code overflow...". This applied to any character class
3945 that was followed by a possessive quantifier.
3946
3947 15. Modified the Makefile to add libpcre.la as a prerequisite for
3948 libpcreposix.la because I was told this is needed for a parallel build to
3949 work.
3950
3951 16. If a pattern that contained .* following optional items at the start was
3952 studied, the wrong optimizing data was generated, leading to matching
3953 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3954 matching string must start with a or b or c. The correct conclusion for
3955 this pattern is that a match can start with any character.
3956
3957
3958 Version 4.4 13-Aug-03
3959 ---------------------
3960
3961 1. In UTF-8 mode, a character class containing characters with values between
3962 127 and 255 was not handled correctly if the compiled pattern was studied.
3963 In fixing this, I have also improved the studying algorithm for such
3964 classes (slightly).
3965
3966 2. Three internal functions had redundant arguments passed to them. Removal
3967 might give a very teeny performance improvement.
3968
3969 3. Documentation bug: the value of the capture_top field in a callout is *one
3970 more than* the number of the hightest numbered captured substring.
3971
3972 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3973 in incorrectly linking with a previously installed version. They now link
3974 explicitly with libpcre.la.
3975
3976 5. configure.in no longer needs to recognize Cygwin specially.
3977
3978 6. A problem in pcre.in for Windows platforms is fixed.
3979
3980 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3981 pcretest, it used to include the size of the study block as part of its
3982 output. Unfortunately, the structure contains a field that has a different
3983 size on different hardware architectures. This meant that the tests that
3984 showed this size failed. As the block is currently always of a fixed size,
3985 this information isn't actually particularly useful in pcretest output, so
3986 I have just removed it.
3987
3988 8. Three pre-processor statements accidentally did not start in column 1.
3989 Sadly, there are *still* compilers around that complain, even though
3990 standard C has not required this for well over a decade. Sigh.
3991
3992 9. In pcretest, the code for checking callouts passed small integers in the
3993 callout_data field, which is a void * field. However, some picky compilers
3994 complained about the casts involved for this on 64-bit systems. Now
3995 pcretest passes the address of the small integer instead, which should get
3996 rid of the warnings.
3997
3998 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3999 both compile and run time, and gives an error if an invalid UTF-8 sequence
4000 is found. There is a option for disabling this check in cases where the
4001 string is known to be correct and/or the maximum performance is wanted.
4002
4003 11. In response to a bug report, I changed one line in Makefile.in from
4004
4005 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
4006 to
4007 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
4008
4009 to look similar to other lines, but I have no way of telling whether this
4010 is the right thing to do, as I do not use Windows. No doubt I'll get told
4011 if it's wrong...
4012
4013
4014 Version 4.3 21-May-03
4015 ---------------------
4016
4017 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
4018 Makefile.
4019
4020 2. Some refactoring to improve the quality of the code:
4021
4022 (i) The utf8_table... variables are now declared "const".
4023
4024 (ii) The code for \cx, which used the "case flipping" table to upper case
4025 lower case letters, now just substracts 32. This is ASCII-specific,
4026 but the whole concept of \cx is ASCII-specific, so it seems
4027 reasonable.
4028
4029 (iii) PCRE was using its character types table to recognize decimal and
4030 hexadecimal digits in the pattern. This is silly, because it handles
4031 only 0-9, a-f, and A-F, but the character types table is locale-
4032 specific, which means strange things might happen. A private
4033 table is now used for this - though it costs 256 bytes, a table is
4034 much faster than multiple explicit tests. Of course, the standard
4035 character types table is still used for matching digits in subject
4036 strings against \d.
4037
4038 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
4039 ending in _t are). So I've renamed it as ESC_tee.
4040
4041 3. The first argument for regexec() in the POSIX wrapper should have been
4042 defined as "const".
4043
4044 4. Changed pcretest to use malloc() for its buffers so that they can be
4045 Electric Fenced for debugging.
4046
4047 5. There were several places in the code where, in UTF-8 mode, PCRE would try
4048 to read one or more bytes before the start of the subject string. Often this
4049 had no effect on PCRE's behaviour, but in some circumstances it could
4050 provoke a segmentation fault.
4051
4052 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
4053 to try to read one or more bytes before the start of the subject string.
4054
4055 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
4056 UTF-8 support could misbehave in various ways if the subject string
4057 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
4058 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
4059 back over UTF-8 characters.)
4060
4061
4062 Version 4.2 14-Apr-03
4063 ---------------------
4064
4065 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
4066
4067 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
4068 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
4069 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
4070 [WIN_PREFIX]: new variable, "cyg" for Cygwin
4071 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
4072 and BUILD_EXEEXT
4073 Note: automatic setting of the BUILD variables is not yet working
4074 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
4075 compile-time but not at link-time
4076 [LINK]: use for linking executables only
4077 make different versions for Windows and non-Windows
4078 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
4079 libraries
4080 [LINK_FOR_BUILD]: new variable
4081 [OBJEXT]: use throughout
4082 [EXEEXT]: use throughout
4083 <winshared>: new target
4084 <wininstall>: new target
4085 <dftables.o>: use native compiler
4086 <dftables>: use native linker
4087 <install>: handle Windows platform correctly
4088 <clean>: ditto
4089 <check>: ditto
4090 copy DLL to top builddir before testing
4091
4092 As part of these changes, -no-undefined was removed again. This was reported
4093 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
4094 in any case.
4095
4096 3. Some tidies to get rid of compiler warnings:
4097
4098 . In the match_data structure, match_limit was an unsigned long int, whereas
4099 match_call_count was an int. I've made them both unsigned long ints.
4100
4101 . In pcretest the fact that a const uschar * doesn't automatically cast to
4102 a void * provoked a warning.
4103
4104 . Turning on some more compiler warnings threw up some "shadow" variables
4105 and a few more missing casts.
4106
4107 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
4108 option, a class that contained a single character with a value between 128
4109 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
4110
4111 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
4112 option, a class that contained several characters, but with at least one
4113 whose value was between 128 and 255 caused PCRE to crash.
4114
4115
4116 Version 4.1 12-Mar-03
4117 ---------------------
4118
4119 1. Compiling with gcc -pedantic found a couple of places where casts were
4120 needed, and a string in dftables.c that was longer than standard compilers are
4121 required to support.
4122
4123 2. Compiling with Sun's compiler found a few more places where the code could
4124 be tidied up in order to avoid warnings.
4125
4126 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
4127 first of these names is deprecated in the latest Autoconf in favour of the name
4128 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
4129 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
4130 analogy I have changed it to CFLAGS_FOR_BUILD.
4131
4132 4. Added -no-undefined to the linking command in the Makefile, because this is
4133 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
4134 linking step for the pcreposix library.
4135
4136 5. PCRE was failing to diagnose the case of two named groups with the same
4137 name.
4138
4139 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
4140 literal character that is needed in the subject for a match, and scans along to
4141 ensure that it is present before embarking on the full matching process. This
4142 saves time in cases of nested unlimited repeats that are never going to match.
4143 Problem: the scan can take a lot of time if the subject is very long (e.g.
4144 megabytes), thus penalizing straightforward matches. It is now done only if the
4145 amount of subject to be scanned is less than 1000 bytes.
4146
4147 7. A lesser problem with the same optimization is that it was recording the
4148 first character of an anchored pattern as "needed", thus provoking a search
4149 right along the subject, even when the first match of the pattern was going to
4150 fail. The "needed" character is now not set for anchored patterns, unless it
4151 follows something in the pattern that is of non-fixed length. Thus, it still
4152 fulfils its original purpose of finding quick non-matches in cases of nested
4153 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
4154
4155
4156 Version 4.0 17-Feb-03
4157 ---------------------
4158
4159 1. If a comment in an extended regex that started immediately after a meta-item
4160 extended to the end of string, PCRE compiled incorrect data. This could lead to
4161 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
4162
4163 2. Moved to autoconf 2.53 and libtool 1.4.2.
4164
4165 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
4166 the special perltest8 script is no longer needed - all the tests can be run
4167 from a single perltest script.
4168
4169 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
4170 by \s. It has now been removed in PCRE. This means it isn't recognized as
4171 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
4172 class [:space:] *does* include VT, thereby creating a mess.
4173
4174 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
4175 space and tab.
4176
4177 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
4178 its new features into the main test script, reducing the number of scripts.
4179
4180 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
4181 were backward compatible, and made the (?i) apply to the whole pattern, as if
4182 /i were given. Now it behaves more logically, and applies the option setting
4183 only to what follows. PCRE has been changed to follow suit. However, if it
4184 finds options settings right at the start of the pattern, it extracts them into
4185 the global options, as before. Thus, they show up in the info data.
4186
4187 8. Added support for the \Q...\E escape sequence. Characters in between are
4188 treated as literals. This is slightly different from Perl in that $ and @ are
4189 also handled as literals inside the quotes. In Perl, they will cause variable
4190 interpolation. Note the following examples:
4191
4192 Pattern PCRE matches Perl matches
4193
4194 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
4195 \Qabc\$xyz\E abc\$xyz abc\$xyz
4196 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
4197
4198 For compatibility with Perl, \Q...\E sequences are recognized inside character
4199 classes as well as outside them.
4200
4201 9. Re-organized 3 code statements in pcretest to avoid "overflow in
4202 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
4203 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
4204 signed/unsigned warnings.
4205
4206 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
4207 option for pcretest, so I've replaced it by a simple function that does just
4208 that job.
4209
4210 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
4211 "pcregrep -".
4212
4213 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
4214 Java package. This provides some syntactic sugar for simple cases of what my
4215 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
4216 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
4217 item, you can use this simplified notation. Note that only makes sense with
4218 greedy quantifiers. Consequently, the use of the possessive quantifier forces
4219 greediness, whatever the setting of the PCRE_UNGREEDY option.
4220
4221 13. A change of greediness default within a pattern was not taking effect at
4222 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
4223 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
4224 was abstracted outside.
4225
4226 14. PCRE now supports the \G assertion. It is true when the current matching
4227 position is at the start point of the match. This differs from \A when the
4228 starting offset is non-zero. Used with the /g option of pcretest (or similar
4229 code), it works in the same way as it does for Perl's /g option. If all
4230 alternatives of a regex begin with \G, the expression is anchored to the start
4231 match position, and the "anchored" flag is set in the compiled expression.
4232
4233 15. Some bugs concerning the handling of certain option changes within patterns
4234 have been fixed. These applied to options other than (?ims). For example,
4235 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
4236 been the other way round. Some of this was related to change 7 above.
4237
4238 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
4239 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
4240 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
4241 POSIX classes only within a class (e.g. /[[:alpha:]]/).
4242
4243 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
4244 mode. Unlike ".", it always matches newline, whatever the setting of
4245 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
4246 assertions. Perl allows it, but it doesn't (in general) work because it can't
4247 calculate the length of the lookbehind. At least, that's the case for Perl
4248 5.8.0 - I've been told they are going to document that it doesn't work in
4249 future.
4250
4251 18. Added an error diagnosis for escapes that PCRE does not support: these are
4252 \L, \l, \N, \P, \p, \U, \u, and \X.
4253
4254 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
4255 reading past the end of the pattern in cases such as /[abcd/.
4256
4257 20. PCRE was getting more memory than necessary for patterns with classes that
4258 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
4259
4260 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
4261 compiling PCRE for use with Virtual Pascal.
4262
4263 22. Small fix to the Makefile to make it work properly if the build is done
4264 outside the source tree.
4265
4266 23. Added a new extension: a condition to go with recursion. If a conditional
4267 subpattern starts with (?(R) the "true" branch is used if recursion has
4268 happened, whereas the "false" branch is used only at the top level.
4269
4270 24. When there was a very long string of literal characters (over 255 bytes
4271 without UTF support, over 250 bytes with UTF support), the computation of how
4272 much memory was required could be incorrect, leading to segfaults or other
4273 strange effects.
4274
4275 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
4276 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
4277 there was a subsequent back reference to those brackets. This meant that, for
4278 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
4279 possible to check for precisely this case. All we can do is abandon the
4280 optimization if .* occurs inside capturing brackets when there are any back
4281 references whatsoever. (See below for a better fix that came later.)
4282
4283 26. The handling of the optimization for finding the first character of a
4284 non-anchored pattern, and for finding a character that is required later in the
4285 match were failing in some cases. This didn't break the matching; it just
4286 failed to optimize when it could. The way this is done has been re-implemented.
4287
4288 27. Fixed typo in error message for invalid (?R item (it said "(?p").
4289
4290 28. Added a new feature that provides some of the functionality that Perl
4291 provides with (?{...}). The facility is termed a "callout". The way it is done
4292 in PCRE is for the caller to provide an optional function, by setting
4293 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
4294 global variable. By default it is unset, which disables all calling out. To get
4295 the function called, the regex must include (?C) at appropriate points. This
4296 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
4297 This provides a means of identifying different callout points. When PCRE
4298 reaches such a point in the regex, if pcre_callout has been set, the external
4299 function is called. It is provided with data in a structure called
4300 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
4301 matching continues; if it returns a non-zero value, the match at the current
4302 point fails. However, backtracking will occur if possible. [This was changed
4303 later and other features added - see item 49 below.]
4304
4305 29. pcretest is upgraded to test the callout functionality. It provides a
4306 callout function that displays information. By default, it shows the start of
4307 the match and the current position in the text. There are some new data escapes
4308 to vary what happens:
4309
4310 \C+ in addition, show current contents of captured substrings
4311 \C- do not supply a callout function
4312 \C!n return 1 when callout number n is reached
4313 \C!n!m return 1 when callout number n is reached for the mth time
4314
4315 30. If pcregrep was called with the -l option and just a single file name, it
4316 output "<stdin>" if a match was found, instead of the file name.
4317
4318 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
4319 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
4320 pcre_exec(). This saves a malloc/free per call. The default value of
4321 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
4322 when configuring.
4323
4324 32. The default maximum size of a compiled pattern is 64K. There have been a
4325 few cases of people hitting this limit. The code now uses macros to handle the
4326 storing of links as offsets within the compiled pattern. It defaults to 2-byte
4327 links, but this can be changed to 3 or 4 bytes by --with-link-size when
4328 configuring. Tests 2 and 5 work only with 2-byte links because they output
4329 debugging information about compiled patterns.
4330
4331 33. Internal code re-arrangements:
4332
4333 (a) Moved the debugging function for printing out a compiled regex into
4334 its own source file (printint.c) and used #include to pull it into
4335 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
4336 separate copies.
4337
4338 (b) Defined the list of op-code names for debugging as a macro in
4339 internal.h so that it is next to the definition of the opcodes.
4340
4341 (c) Defined a table of op-code lengths for simpler skipping along compiled
4342 code. This is again a macro in internal.h so that it is next to the
4343 definition of the opcodes.
4344
4345 34. Added support for recursive calls to individual subpatterns, along the
4346 lines of Robin Houston's patch (but implemented somewhat differently).
4347
4348 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
4349 allow it to read and process whole directories in Win32. This code was
4350 contributed by Lionel Fourquaux; it has not been tested by me.
4351
4352 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
4353 used to name a group. Names consist of alphanumerics and underscores, and must
4354 be unique. Back references use the syntax (?P=name) and recursive calls use
4355 (?P>name) which is a PCRE extension to the Python extension. Groups still have
4356 numbers. The function pcre_fullinfo() can be used after compilation to extract
4357 a name/number map. There are three relevant calls:
4358
4359 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
4360 PCRE_INFO_NAMECOUNT yields the number of entries
4361 PCRE_INFO_NAMETABLE yields a pointer to the map.
4362
4363 The map is a vector of fixed-size entries. The size of each entry depends on
4364 the length of the longest name used. The first two bytes of each entry are the
4365 group number, most significant byte first. There follows the corresponding
4366 name, zero terminated. The names are in alphabetical order.
4367
4368 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
4369 case instead of 255. Making it the same both with and without UTF-8 support
4370 means that the same test output works with both.
4371
4372 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
4373 calling malloc() with a zero argument.
4374
4375 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
4376 optimization. I've improved things by keeping a bitmap of backreferences with
4377 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
4378 fact referenced, the optimization can be applied. It is unlikely that a
4379 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
4380 the match to follow \n) will appear inside brackets with a number greater than
4381 31, but if it does, any back reference > 31 suppresses the optimization.
4382
4383 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
4384 of disabling numbered capturing parentheses. Any opening parenthesis that is
4385 not followed by ? behaves as if it were followed by ?: but named parentheses
4386 can still be used for capturing (and they will acquire numbers in the usual
4387 way).
4388
4389 41. Redesigned the return codes from the match() function into yes/no/error so
4390 that errors can be passed back from deep inside the nested calls. A malloc
4391 failure while inside a recursive subpattern call now causes the
4392 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
4393
4394 42. It is now possible to set a limit on the number of times the match()
4395 function is called in a call to pcre_exec(). This facility makes it possible to
4396 limit the amount of recursion and backtracking, though not in a directly
4397 obvious way, because the match() function is used in a number of different
4398 circumstances. The count starts from zero for each position in the subject
4399 string (for non-anchored patterns). The default limit is, for compatibility, a
4400 large number, namely 10 000 000. You can change this in two ways:
4401
4402 (a) When configuring PCRE before making, you can use --with-match-limit=n
4403 to set a default value for the compiled library.
4404
4405 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
4406 a different value is set. See 45 below.
4407
4408 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
4409
4410 43. Added a new function pcre_config(int, void *) to enable run-time extraction
4411 of things that can be changed at compile time. The first argument specifies
4412 what is wanted and the second points to where the information is to be placed.
4413 The current list of available information is:
4414
4415 PCRE_CONFIG_UTF8
4416
4417 The output is an integer that is set to one if UTF-8 support is available;
4418 otherwise it is set to zero.
4419
4420 PCRE_CONFIG_NEWLINE
4421
4422 The output is an integer that it set to the value of the code that is used for
4423 newline. It is either LF (10) or CR (13).
4424
4425 PCRE_CONFIG_LINK_SIZE
4426
4427 The output is an integer that contains the number of bytes used for internal
4428 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
4429
4430 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
4431
4432 The output is an integer that contains the threshold above which the POSIX
4433 interface uses malloc() for output vectors. See item 31 above.
4434
4435 PCRE_CONFIG_MATCH_LIMIT
4436
4437 The output is an unsigned integer that contains the default limit of the number
4438 of match() calls in a pcre_exec() execution. See 42 above.
4439
4440 44. pcretest has been upgraded by the addition of the -C option. This causes it
4441 to extract all the available output from the new pcre_config() function, and to
4442 output it. The program then exits immediately.
4443
4444 45. A need has arisen to pass over additional data with calls to pcre_exec() in
4445 order to support additional features. One way would have been to define
4446 pcre_exec2() (for example) with extra arguments, but this would not have been
4447 extensible, and would also have required all calls to the original function to
4448 be mapped to the new one. Instead, I have chosen to extend the mechanism that
4449 is used for passing in "extra" data from pcre_study().
4450
4451 The pcre_extra structure is now exposed and defined in pcre.h. It currently
4452 contains the following fields:
4453
4454 flags a bitmap indicating which of the following fields are set
4455 study_data opaque data from pcre_study()
4456 match_limit a way of specifying a limit on match() calls for a specific
4457 call to pcre_exec()
4458 callout_data data for callouts (see 49 below)
4459
4460 The flag bits are also defined in pcre.h, and are
4461
4462 PCRE_EXTRA_STUDY_DATA
4463 PCRE_EXTRA_MATCH_LIMIT
4464 PCRE_EXTRA_CALLOUT_DATA
4465
4466 The pcre_study() function now returns one of these new pcre_extra blocks, with
4467 the actual study data pointed to by the study_data field, and the
4468 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
4469 before. That is, this change is entirely upwards-compatible and requires no
4470 change to existing code.
4471
4472 If you want to pass in additional data to pcre_exec(), you can either place it
4473 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
4474 block.
4475
4476 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
4477 data string contains the escape sequence \M, pcretest calls pcre_exec() several
4478 times with different match limits, until it finds the minimum value needed for
4479 pcre_exec() to complete. The value is then output. This can be instructive; for
4480 most simple matches the number is quite small, but for pathological cases it
4481 gets very large very quickly.
4482
4483 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
4484 returns the size of the data block pointed to by the study_data field in a
4485 pcre_extra block, that is, the value that was passed as the argument to
4486 pcre_malloc() when PCRE was getting memory in which to place the information
4487 created by pcre_study(). The fourth argument should point to a size_t variable.
4488 pcretest has been extended so that this information is shown after a successful
4489 pcre_study() call when information about the compiled regex is being displayed.
4490
4491 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
4492 because what follows is always an absolute path. (Later: it turns out that this
4493 is more than cosmetic for MinGW, because it doesn't like empty path
4494 components.)
4495
4496 49. Some changes have been made to the callout feature (see 28 above):
4497
4498 (i) A callout function now has three choices for what it returns:
4499
4500 0 => success, carry on matching
4501 > 0 => failure at this point, but backtrack if possible
4502 < 0 => serious error, return this value from pcre_exec()
4503
4504 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
4505 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
4506 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
4507 use by callout functions. It will never be used by PCRE itself.
4508
4509 (ii) The pcre_extra structure (see 45 above) has a void * field called
4510 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
4511 pcre_callout_block structure has a field of the same name. The contents of
4512 the field passed in the pcre_extra structure are passed to the callout
4513 function in the corresponding field in the callout block. This makes it
4514 easier to use the same callout-containing regex from multiple threads. For
4515 testing, the pcretest program has a new data escape
4516
4517 \C*n pass the number n (may be negative) as callout_data
4518
4519 If the callout function in pcretest receives a non-zero value as
4520 callout_data, it returns that value.
4521
4522 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
4523 there were some redundant $(CFLAGS) in commands that are now specified as
4524 $(LINK), which already includes $(CFLAGS).
4525
4526 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
4527 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
4528 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
4529 one-byte characters throughout. Note that case-insensitive matching applies
4530 only to characters whose values are less than 256. PCRE doesn't support the
4531 notion of cases for higher-valued characters.
4532
4533 (i) A character class whose characters are all within 0-255 is handled as
4534 a bit map, and the map is inverted for negative classes. Previously, a
4535 character > 255 always failed to match such a class; however it should
4536 match if the class was a negative one (e.g. [^ab]). This has been fixed.
4537
4538 (ii) A negated character class with a single character < 255 is coded as
4539 "not this character" (OP_NOT). This wasn't working properly when the test
4540 character was multibyte, either singly or repeated.
4541
4542 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
4543 mode, for example: \x{100}{2,3}.
4544
4545 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
4546 singly or repeated) now correctly test multibyte characters. However,
4547 PCRE doesn't recognize any characters with values greater than 255 as
4548 digits, spaces, or word characters. Such characters always match \D, \S,
4549 and \W, and never match \d, \s, or \w.
4550
4551 (v) Classes may now contain characters and character ranges with values
4552 greater than 255. For example: [ab\x{100}-\x{400}].
4553
4554 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
4555 PCRE in UTF-8 mode.
4556
4557 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
4558 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
4559 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
4560 value.)
4561
4562 53. The single man page has become too large. I have therefore split it up into
4563 a number of separate man pages. These also give rise to individual HTML pages;
4564 these are now put in a separate directory, and there is an index.html page that
4565 lists them all. Some hyperlinking between the pages has been installed.
4566
4567 54. Added convenience functions for handling named capturing parentheses.
4568
4569 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4570 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4571 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4572 are faulted.
4573
4574 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4575 calling configure. These values are used when compiling the dftables.c program
4576 which is run to generate the source of the default character tables. They
4577 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4578 you will need to set these values.
4579
4580 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4581
4582
4583 Version 3.9 02-Jan-02
4584 ---------------------
4585
4586 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4587
4588 2. If --disable-static was given, the building process failed when trying to
4589 build pcretest and pcregrep. (For some reason it was using libtool to compile
4590 them, which is not right, as they aren't part of the library.)
4591
4592
4593 Version 3.8 18-Dec-01
4594 ---------------------
4595
4596 1. The experimental UTF-8 code was completely screwed up. It was packing the
4597 bytes in the wrong order. How dumb can you get?
4598
4599
4600 Version 3.7 29-Oct-01
4601 ---------------------
4602
4603 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4604 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4605 this didn't happen under Solaris 8, where I normally test things.
4606
4607 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4608 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4609 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4610 this fix an hour or so after the initial 3.7 release.)
4611
4612
4613 Version 3.6 23-Oct-01
4614 ---------------------
4615
4616 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4617 offsets passed as NULL with zero offset count.
4618
4619 2. The config.guess and config.sub files had not been updated when I moved to
4620 the latest autoconf.
4621
4622
4623 Version 3.5 15-Aug-01
4624 ---------------------
4625
4626 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4627 had been forgotten.
4628
4629 2. By using declared but undefined structures, we can avoid using "void"
4630 definitions in pcre.h while keeping the internal definitions of the structures
4631 private.
4632
4633 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4634 user point of view, this means that both static and shared libraries are built
4635 by default, but this can be individually controlled. More of the work of
4636 handling this static/shared cases is now inside libtool instead of PCRE's make
4637 file.
4638
4639 4. The pcretest utility is now installed along with pcregrep because it is
4640 useful for users (to test regexs) and by doing this, it automatically gets
4641 relinked by libtool. The documentation has been turned into a man page, so
4642 there are now .1, .txt, and .html versions in /doc.
4643
4644 5. Upgrades to pcregrep:
4645 (i) Added long-form option names like gnu grep.
4646 (ii) Added --help to list all options with an explanatory phrase.
4647 (iii) Added -r, --recursive to recurse into sub-directories.
4648 (iv) Added -f, --file to read patterns from a file.
4649
4650 6. pcre_exec() was referring to its "code" argument before testing that
4651 argument for NULL (and giving an error if it was NULL).
4652
4653 7. Upgraded Makefile.in to allow for compiling in a different directory from
4654 the source directory.
4655
4656 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4657 options bits, the pointer it was passed was to an int instead of to an unsigned
4658 long int. This mattered only on 64-bit systems.
4659
4660 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4661 generated) instead of pcre.in, which it its source. Also made the same change
4662 in several of the .c files.
4663
4664 10. A new release of gcc defines printf() as a macro, which broke pcretest
4665 because it had an ifdef in the middle of a string argument for printf(). Fixed
4666 by using separate calls to printf().
4667
4668 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4669 script, to force use of CR or LF instead of \n in the source. On non-Unix
4670 systems, the value can be set in config.h.
4671
4672 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4673 absolute limit. Changed the text of the error message to make this clear, and
4674 likewise updated the man page.
4675
4676 13. The limit of 99 on the number of capturing subpatterns has been removed.
4677 The new limit is 65535, which I hope will not be a "real" limit.
4678
4679
4680 Version 3.4 22-Aug-00
4681 ---------------------
4682
4683 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4684
4685 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4686
4687
4688 Version 3.3 01-Aug-00
4689 ---------------------
4690
4691 1. If an octal character was given, but the value was greater than \377, it
4692 was not getting masked to the least significant bits, as documented. This could
4693 lead to crashes in some systems.
4694
4695 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4696 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4697
4698 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4699 These just pass their arguments on to (pcre_free)(), but they are provided
4700 because some uses of PCRE bind it to non-C systems that can call its functions,
4701 but cannot call free() or pcre_free() directly.
4702
4703 4. Add "make test" as a synonym for "make check". Corrected some comments in
4704 the Makefile.
4705
4706 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4707 Makefile.
4708
4709 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4710 command called pgrep for grepping around the active processes.
4711
4712 7. Added the beginnings of support for UTF-8 character strings.
4713
4714 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4715 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4716 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4717 out for the ar command.)
4718
4719
4720 Version 3.2 12-May-00
4721 ---------------------
4722
4723 This is purely a bug fixing release.
4724
4725 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4726 of ZA. This was just one example of several cases that could provoke this bug,
4727 which was introduced by change 9 of version 2.00. The code for breaking
4728 infinite loops after an iteration that matches an empty string was't working
4729 correctly.
4730
4731 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4732 when matched against abbab (for example). After matching an empty string, it
4733 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4734 caused it to match further down the string than it should.
4735
4736 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4737 was there because it doesn't seem to be needed, and it causes trouble on some
4738 systems, as it is not a Standard C header. It has been removed.
4739
4740 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4741 were reported on the Macintosh. The changes were from
4742
4743 while ((c = *(++ptr)) != 0 && c != '\n');
4744 to
4745 while ((c = *(++ptr)) != 0 && c != '\n') ;
4746
4747 Totally extraordinary, but if that's what it takes...
4748
4749 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4750 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4751 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4752 assumes the way PCRE uses memmove() (always moving upwards).
4753
4754 6. PCRE is being used in one environment where strchr() is not available. There
4755 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4756 faster code anyway.
4757
4758
4759 Version 3.1 09-Feb-00
4760 ---------------------
4761
4762 The only change in this release is the fixing of some bugs in Makefile.in for
4763 the "install" target:
4764
4765 (1) It was failing to install pcreposix.h.
4766
4767 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4768
4769
4770 Version 3.0 01-Feb-00
4771 ---------------------
4772
4773 1. Add support for the /+ modifier to perltest (to output $` like it does in
4774 pcretest).
4775
4776 2. Add support for the /g modifier to perltest.
4777
4778 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4779 matches null strings.
4780
4781 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4782 pattern. Perl treats empty patterns specially - it reuses the most recent
4783 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4784 effect.
4785
4786 5. The POSIX interface was broken in that it was just handing over the POSIX
4787 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4788 required a bigger vector, with some working space on the end. This means that
4789 the POSIX wrapper now has to get and free some memory, and copy the results.
4790
4791 6. Added some simple autoconf support, placing the test data and the
4792 documentation in separate directories, re-organizing some of the
4793 information files, and making it build pcre-config (a GNU standard). Also added
4794 libtool support for building PCRE as a shared library, which is now the
4795 default.
4796
4797 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4798 09 are not valid octal constants. Single digits will be used for minor values
4799 less than 10.
4800
4801 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4802 existing programs that set these in the POSIX interface can use PCRE without
4803 modification.
4804
4805 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4806 return all that pcre_info() returns, plus additional data. The pcre_info()
4807 function is retained for compatibility, but is considered to be obsolete.
4808
4809 10. Added experimental recursion feature (?R) to handle one common case that
4810 Perl 5.6 will be able to do with (?p{...}).
4811
4812 11. Added support for POSIX character classes like [:alpha:], which Perl is
4813 adopting.
4814
4815
4816 Version 2.08 31-Aug-99
4817 ----------------------
4818
4819 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4820 trying to match at the startoffset position, but instead was moving forward to
4821 the next newline as if a previous match had failed.
4822
4823 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4824 and could get into a loop if a null string was matched other than at the start
4825 of the subject.
4826
4827 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4828 be distinguished at compile time, and for completeness also added PCRE_DATE.
4829
4830 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4831 in GnuWin32 environments.
4832
4833
4834 Version 2.07 29-Jul-99
4835 ----------------------
4836
4837 1. The documentation is now supplied in plain text form and HTML as well as in
4838 the form of man page sources.
4839
4840 2. C++ compilers don't like assigning (void *) values to other pointer types.
4841 In particular this affects malloc(). Although there is no problem in Standard
4842 C, I've put in casts to keep C++ compilers happy.
4843
4844 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4845 should be (const char *).
4846
4847 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4848 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4849 However, I haven't made this a standard facility. The documentation doesn't
4850 mention it, and the Makefile doesn't support it.
4851
4852 5. The Makefile now contains an "install" target, with editable destinations at
4853 the top of the file. The pcretest program is not installed.
4854
4855 6. pgrep -V now gives the PCRE version number and date.
4856
4857 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4858 causing the entire string to be ignored, instead of just the last character.
4859
4860 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4861 non-matching string, it can take a very, very long time, even for strings of
4862 quite modest length, because of the nested recursion. PCRE now does better in
4863 some of these cases. It does this by remembering the last required literal
4864 character in the pattern, and pre-searching the subject to ensure it is present
4865 before running the real match. In other words, it applies a heuristic to detect
4866 some types of certain failure quickly, and in the above example, if presented
4867 with a string that has no trailing " it gives "no match" very quickly.
4868
4869 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4870 other alternatives are tried instead.
4871
4872
4873 Version 2.06 09-Jun-99
4874 ----------------------
4875
4876 1. Change pcretest's output for amount of store used to show just the code
4877 space, because the remainder (the data block) varies in size between 32-bit and
4878 64-bit systems.
4879
4880 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4881 start matching at. This allows lookbehinds to work when searching for multiple
4882 occurrences in a string.
4883
4884 3. Added additional options to pcretest for testing multiple occurrences:
4885
4886 /+ outputs the rest of the string that follows a match
4887 /g loops for multiple occurrences, using the new startoffset argument
4888 /G loops for multiple occurrences by passing an incremented pointer
4889
4890 4. PCRE wasn't doing the "first character" optimization for patterns starting
4891 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4892 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4893 the letter 'x'. On long subject strings, this gives a significant speed-up.
4894
4895
4896 Version 2.05 21-Apr-99
4897 ----------------------
4898
4899 1. Changed the type of magic_number from int to long int so that it works
4900 properly on 16-bit systems.
4901
4902 2. Fixed a bug which caused patterns starting with .* not to work correctly
4903 when the subject string contained newline characters. PCRE was assuming
4904 anchoring for such patterns in all cases, which is not correct because .* will
4905 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4906 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4907 must be retried after every newline in the subject.
4908
4909
4910 Version 2.04 18-Feb-99
4911 ----------------------
4912
4913 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4914 computation of the store needed to hold the pattern was incorrect (too large).
4915 If such patterns were nested a few deep, this could multiply and become a real
4916 problem.
4917
4918 2. Added /M option to pcretest to show the memory requirement of a specific
4919 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4920
4921 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4922 compiled in such a way that the backtracking after subsequent failure was
4923 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4924 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4925
4926
4927 Version 2.03 02-Feb-99
4928 ----------------------
4929
4930 1. Fixed typo and small mistake in man page.
4931
4932 2. Added 4th condition (GPL supersedes if conflict) and created separate
4933 LICENCE file containing the conditions.
4934
4935 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4936 Perl, that is the internal \ allows the delimiter to be included in the
4937 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4938 the final delimiter, add \ to the end of the pattern (to test the error).
4939
4940 4. Added the convenience functions for extracting substrings after a successful
4941 match. Updated pcretest to make it able to test these functions.
4942
4943
4944 Version 2.02 14-Jan-99
4945 ----------------------
4946
4947 1. Initialized the working variables associated with each extraction so that
4948 their saving and restoring doesn't refer to uninitialized store.
4949
4950 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4951 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4952 fix the problem.
4953
4954 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4955 calls, and wasn't printing the correct value for compiling calls. Increased the
4956 default value of LOOPREPEAT, and the number of significant figures in the
4957 times.
4958
4959 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4960
4961 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4962 a building problem on Windows NT with a FAT file system.
4963
4964
4965 Version 2.01 21-Oct-98
4966 ----------------------
4967
4968 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4969 to character tables built by pcre_maketables() in the current locale. If NULL
4970 is passed, the default tables are used.
4971
4972
4973 Version 2.00 24-Sep-98
4974 ----------------------
4975
4976 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4977 it any more.
4978
4979 2. Allow quantification of (?>) groups, and make it work correctly.
4980
4981 3. The first character computation wasn't working for (?>) groups.
4982
4983 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4984 end of the subject) and add 5.005's \z, which really does match only at the
4985 very end of the subject.
4986
4987 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4988
4989 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4990 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4991 localized options. All options to pcre_study() were also removed.
4992
4993 7. Add other new features from 5.005:
4994
4995 $(?<= positive lookbehind
4996 $(?<! negative lookbehind
4997 (?imsx-imsx) added the unsetting capability
4998 such a setting is global if at outer level; local otherwise
4999 (?imsx-imsx:) non-capturing groups with option setting
5000 (?(cond)re|re) conditional pattern matching
5001
5002 A backreference to itself in a repeated group matches the previous
5003 captured string.
5004
5005 8. General tidying up of studying (both automatic and via "study")
5006 consequential on the addition of new assertions.
5007
5008 9. As in 5.005, unlimited repeated groups that could match an empty substring
5009 are no longer faulted at compile time. Instead, the loop is forcibly broken at
5010 runtime if any iteration does actually match an empty substring.
5011
5012 10. Include the RunTest script in the distribution.
5013
5014 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
5015 discrepancies, some of which were old and were also with respect to 5.004. They
5016 have now been fixed.
5017
5018
5019 Version 1.09 28-Apr-98
5020 ----------------------
5021
5022 1. A negated single character class followed by a quantifier with a minimum
5023 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
5024 program crashes, or just wrong answers. This did not apply to negated classes
5025 containing more than one character, or to minima other than one.
5026
5027
5028 Version 1.08 27-Mar-98
5029 ----------------------
5030
5031 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
5032
5033 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
5034 latter must appear before anything that relies on it in the pattern.
5035
5036
5037 Version 1.07 16-Feb-98
5038 ----------------------
5039
5040 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
5041 repeat of a potentially empty string).
5042
5043
5044 Version 1.06 23-Jan-98
5045 ----------------------
5046
5047 1. Added Markus Oberhumer's little patches for C++.
5048
5049 2. Literal strings longer than 255 characters were broken.
5050
5051
5052 Version 1.05 23-Dec-97
5053 ----------------------
5054
5055 1. Negated character classes containing more than one character were failing if
5056 PCRE_CASELESS was set at run time.
5057
5058
5059 Version 1.04 19-Dec-97
5060 ----------------------
5061
5062 1. Corrected the man page, where some "const" qualifiers had been omitted.
5063
5064 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
5065 input syntax.
5066
5067 3. Fixed memory leak which occurred when a regex with back references was
5068 matched with an offsets vector that wasn't big enough. The temporary memory
5069 that is used in this case wasn't being freed if the match failed.
5070
5071 4. Tidied pcretest to ensure it frees memory that it gets.
5072
5073 5. Temporary memory was being obtained in the case where the passed offsets
5074 vector was exactly big enough.
5075
5076 6. Corrected definition of offsetof() from change 5 below.
5077
5078 7. I had screwed up change 6 below and broken the rules for the use of
5079 setjmp(). Now fixed.
5080
5081
5082 Version 1.03 18-Dec-97
5083 ----------------------
5084
5085 1. A erroneous regex with a missing opening parenthesis was correctly
5086 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
5087 on some systems.
5088
5089 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
5090 it was reported that one broken compiler failed on the former because "code" is
5091 also an independent variable.
5092
5093 3. The erroneous regex a[]b caused an array overrun reference.
5094
5095 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
5096 fail on data ending with that character. (It was going on too far, and checking
5097 the next character, typically a binary zero.) This was specific to the
5098 optimized code for single-character negative classes.
5099
5100 5. Added a contributed patch from the TIN world which does the following:
5101
5102 + Add an undef for memmove, in case the the system defines a macro for it.
5103
5104 + Add a definition of offsetof(), in case there isn't one. (I don't know
5105 the reason behind this - offsetof() is part of the ANSI standard - but
5106 it does no harm).
5107
5108 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
5109 most of the places where whitespace preceded '#'. I have given up and
5110 allowed the remaining 2 cases to be at the margin.
5111
5112 + Rename some variables in pcre to eliminate shadowing. This seems very
5113 pedantic, but does no harm, of course.
5114
5115 6. Moved the call to setjmp() into its own function, to get rid of warnings
5116 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
5117
5118 7. Constructs such as \d{8,} were compiling into the equivalent of
5119 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
5120 outcome, but in this particular case used more store than had been allocated,
5121 which caused the bug to be discovered because it threw up an internal error.
5122
5123 8. The debugging code in both pcre and pcretest for outputting the compiled
5124 form of a regex was going wrong in the case of back references followed by
5125 curly-bracketed repeats.
5126
5127
5128 Version 1.02 12-Dec-97
5129 ----------------------
5130
5131 1. Typos in pcre.3 and comments in the source fixed.
5132
5133 2. Applied a contributed patch to get rid of places where it used to remove
5134 'const' from variables, and fixed some signed/unsigned and uninitialized
5135 variable warnings.
5136
5137 3. Added the "runtest" target to Makefile.
5138
5139 4. Set default compiler flag to -O2 rather than just -O.
5140
5141
5142 Version 1.01 19-Nov-97
5143 ----------------------
5144
5145 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
5146 like /([ab]*)*/, that is, for classes with more than one character in them.
5147
5148 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
5149 as /((?>a*))*/ (a PCRE_EXTRA facility).
5150
5151
5152 Version 1.00 18-Nov-97
5153 ----------------------
5154
5155 1. Added compile-time macros to support systems such as SunOS4 which don't have
5156 memmove() or strerror() but have other things that can be used instead.
5157
5158 2. Arranged that "make clean" removes the executables.
5159
5160
5161 Version 0.99 27-Oct-97
5162 ----------------------
5163
5164 1. Fixed bug in code for optimizing classes with only one character. It was
5165 initializing a 32-byte map regardless, which could cause it to run off the end
5166 of the memory it had got.
5167
5168 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
5169
5170
5171 Version 0.98 22-Oct-97
5172 ----------------------
5173
5174 1. Fixed bug in code for handling temporary memory usage when there are more
5175 back references than supplied space in the ovector. This could cause segfaults.
5176
5177
5178 Version 0.97 21-Oct-97
5179 ----------------------
5180
5181 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
5182
5183 2. Optimized negated single characters not to use a bit map.
5184
5185 3. Brought error texts together as macro definitions; clarified some of them;
5186 fixed one that was wrong - it said "range out of order" when it meant "invalid
5187 escape sequence".
5188
5189 4. Changed some char * arguments to const char *.
5190
5191 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
5192
5193 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
5194 pcretest.
5195
5196
5197 Version 0.96 16-Oct-97
5198 ----------------------
5199
5200 1. Added a simple "pgrep" utility to the distribution.
5201
5202 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
5203 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
5204 where "ddd" means "one or more decimal digits".
5205
5206 3. Fixed serious bug. If a pattern had a back reference, but the call to
5207 pcre_exec() didn't supply a large enough ovector to record the related
5208 identifying subpattern, the match always failed. PCRE now remembers the number
5209 of the largest back reference, and gets some temporary memory in which to save
5210 the offsets during matching if necessary, in order to ensure that
5211 backreferences always work.
5212
5213 4. Increased the compatibility with Perl in a number of ways:
5214
5215 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
5216 to request this handling. The option can be set at compile or exec time.
5217
5218 (b) $ matches before a terminating newline by default; an option
5219 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
5220 mode). The option can be set at compile or exec time.
5221
5222 (c) The handling of \ followed by a digit other than 0 is now supposed to be
5223 the same as Perl's. If the decimal number it represents is less than 10
5224 or there aren't that many previous left capturing parentheses, an octal
5225 escape is read. Inside a character class, it's always an octal escape,
5226 even if it is a single digit.
5227
5228 (d) An escaped but undefined alphabetic character is taken as a literal,
5229 unless PCRE_EXTRA is set. Currently this just reserves the remaining
5230 escapes.
5231
5232 (e) {0} is now permitted. (The previous item is removed from the compiled
5233 pattern).
5234
5235 5. Changed all the names of code files so that the basic parts are no longer
5236 than 10 characters, and abolished the teeny "globals.c" file.
5237
5238 6. Changed the handling of character classes; they are now done with a 32-byte
5239 bit map always.
5240
5241 7. Added the -d and /D options to pcretest to make it possible to look at the
5242 internals of compilation without having to recompile pcre.
5243
5244
5245 Version 0.95 23-Sep-97
5246 ----------------------
5247
5248 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
5249 \x20 at the start of a run of normal characters. These were being treated as
5250 real characters, instead of the source characters being re-checked.
5251
5252
5253 Version 0.94 18-Sep-97
5254 ----------------------
5255
5256 1. The functions are now thread-safe, with the caveat that the global variables
5257 containing pointers to malloc() and free() or alternative functions are the
5258 same for all threads.
5259
5260 2. Get pcre_study() to generate a bitmap of initial characters for non-
5261 anchored patterns when this is possible, and use it if passed to pcre_exec().
5262
5263
5264 Version 0.93 15-Sep-97
5265 ----------------------
5266
5267 1. /(b)|(:+)/ was computing an incorrect first character.
5268
5269 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
5270 but not actually doing anything yet.
5271
5272 3. Treat "-" characters in classes that cannot be part of ranges as literals,
5273 as Perl does (e.g. [-az] or [az-]).
5274
5275 4. Set the anchored flag if a branch starts with .* or .*? because that tests
5276 all possible positions.
5277
5278 5. Split up into different modules to avoid including unneeded functions in a
5279 compiled binary. However, compile and exec are still in one module. The "study"
5280 function is split off.
5281
5282 6. The character tables are now in a separate module whose source is generated
5283 by an auxiliary program - but can then be edited by hand if required. There are
5284 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
5285 toupper() in the code.
5286
5287 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
5288 make them global. Abolish the function for setting them, as the caller can now
5289 set them directly.
5290
5291
5292 Version 0.92 11-Sep-97
5293 ----------------------
5294
5295 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
5296 (e.g. /a{1,3}/) was broken (I mis-optimized it).
5297
5298 2. Caseless matching was not working in character classes if the characters in
5299 the pattern were in upper case.
5300
5301 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
5302
5303 4. Make PCRE_ANCHORED public and accept as a compile option.
5304
5305 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
5306 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
5307 pass them.
5308
5309 6. Give an error if bad option bits passed at compile or run time.
5310
5311 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
5312 pcretest to cause it to pass that flag.
5313
5314 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
5315 options, and the first character, if set.
5316
5317 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
5318
5319
5320 Version 0.91 10-Sep-97
5321 ----------------------
5322
5323 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
5324 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
5325
5326 2. PCRE was looping on encountering an indefinitely repeated back reference to
5327 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
5328 Perl does - treats the match as successful.
5329
5330 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5