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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 55 - (show annotations) (download)
Sat Feb 24 21:39:46 2007 UTC (7 years, 5 months ago) by nigel
File size: 27580 byte(s)
Load pcre-3.6 into code/trunk.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 3.6 23-Oct-01
5 ---------------------
6
7 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
8 offsets passed as NULL with zero offset count.
9
10 2. The config.guess and config.sub files had not been updated when I moved to
11 the latest autoconf.
12
13
14 Version 3.5 15-Aug-01
15 ---------------------
16
17 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
18 had been forgotten.
19
20 2. By using declared but undefined structures, we can avoid using "void"
21 definitions in pcre.h while keeping the internal definitions of the structures
22 private.
23
24 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
25 user point of view, this means that both static and shared libraries are built
26 by default, but this can be individually controlled. More of the work of
27 handling this static/shared cases is now inside libtool instead of PCRE's make
28 file.
29
30 4. The pcretest utility is now installed along with pcregrep because it is
31 useful for users (to test regexs) and by doing this, it automatically gets
32 relinked by libtool. The documentation has been turned into a man page, so
33 there are now .1, .txt, and .html versions in /doc.
34
35 5. Upgrades to pcregrep:
36 (i) Added long-form option names like gnu grep.
37 (ii) Added --help to list all options with an explanatory phrase.
38 (iii) Added -r, --recursive to recurse into sub-directories.
39 (iv) Added -f, --file to read patterns from a file.
40
41 6. pcre_exec() was referring to its "code" argument before testing that
42 argument for NULL (and giving an error if it was NULL).
43
44 7. Upgraded Makefile.in to allow for compiling in a different directory from
45 the source directory.
46
47 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
48 options bits, the pointer it was passed was to an int instead of to an unsigned
49 long int. This mattered only on 64-bit systems.
50
51 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
52 generated) instead of pcre.in, which it its source. Also made the same change
53 in several of the .c files.
54
55 10. A new release of gcc defines printf() as a macro, which broke pcretest
56 because it had an ifdef in the middle of a string argument for printf(). Fixed
57 by using separate calls to printf().
58
59 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
60 script, to force use of CR or LF instead of \n in the source. On non-Unix
61 systems, the value can be set in config.h.
62
63 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
64 absolute limit. Changed the text of the error message to make this clear, and
65 likewise updated the man page.
66
67 13. The limit of 99 on the number of capturing subpatterns has been removed.
68 The new limit is 65535, which I hope will not be a "real" limit.
69
70
71 Version 3.4 22-Aug-00
72 ---------------------
73
74 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
75
76 2. Diagnose condition (?(0) as an error instead of crashing on matching.
77
78
79 Version 3.3 01-Aug-00
80 ---------------------
81
82 1. If an octal character was given, but the value was greater than \377, it
83 was not getting masked to the least significant bits, as documented. This could
84 lead to crashes in some systems.
85
86 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
87 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
88
89 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
90 These just pass their arguments on to (pcre_free)(), but they are provided
91 because some uses of PCRE bind it to non-C systems that can call its functions,
92 but cannot call free() or pcre_free() directly.
93
94 4. Add "make test" as a synonym for "make check". Corrected some comments in
95 the Makefile.
96
97 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
98 Makefile.
99
100 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
101 command called pgrep for grepping around the active processes.
102
103 7. Added the beginnings of support for UTF-8 character strings.
104
105 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
106 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
107 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
108 out for the ar command.)
109
110
111 Version 3.2 12-May-00
112 ---------------------
113
114 This is purely a bug fixing release.
115
116 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
117 of ZA. This was just one example of several cases that could provoke this bug,
118 which was introduced by change 9 of version 2.00. The code for breaking
119 infinite loops after an iteration that matches an empty string was't working
120 correctly.
121
122 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
123 when matched against abbab (for example). After matching an empty string, it
124 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
125 caused it to match further down the string than it should.
126
127 3. The code contained an inclusion of sys/types.h. It isn't clear why this
128 was there because it doesn't seem to be needed, and it causes trouble on some
129 systems, as it is not a Standard C header. It has been removed.
130
131 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
132 were reported on the Macintosh. The changes were from
133
134 while ((c = *(++ptr)) != 0 && c != '\n');
135 to
136 while ((c = *(++ptr)) != 0 && c != '\n') ;
137
138 Totally extraordinary, but if that's what it takes...
139
140 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
141 available. Added HAVE_BCOPY and an autoconf test for it; if neither
142 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
143 assumes the way PCRE uses memmove() (always moving upwards).
144
145 6. PCRE is being used in one environment where strchr() is not available. There
146 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
147 faster code anyway.
148
149
150 Version 3.1 09-Feb-00
151 ---------------------
152
153 The only change in this release is the fixing of some bugs in Makefile.in for
154 the "install" target:
155
156 (1) It was failing to install pcreposix.h.
157
158 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
159
160
161 Version 3.0 01-Feb-00
162 ---------------------
163
164 1. Add support for the /+ modifier to perltest (to output $` like it does in
165 pcretest).
166
167 2. Add support for the /g modifier to perltest.
168
169 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
170 matches null strings.
171
172 4. Fix perltest so that it doesn't do unwanted things when fed an empty
173 pattern. Perl treats empty patterns specially - it reuses the most recent
174 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
175 effect.
176
177 5. The POSIX interface was broken in that it was just handing over the POSIX
178 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
179 required a bigger vector, with some working space on the end. This means that
180 the POSIX wrapper now has to get and free some memory, and copy the results.
181
182 6. Added some simple autoconf support, placing the test data and the
183 documentation in separate directories, re-organizing some of the
184 information files, and making it build pcre-config (a GNU standard). Also added
185 libtool support for building PCRE as a shared library, which is now the
186 default.
187
188 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
189 09 are not valid octal constants. Single digits will be used for minor values
190 less than 10.
191
192 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
193 existing programs that set these in the POSIX interface can use PCRE without
194 modification.
195
196 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
197 return all that pcre_info() returns, plus additional data. The pcre_info()
198 function is retained for compatibility, but is considered to be obsolete.
199
200 10. Added experimental recursion feature (?R) to handle one common case that
201 Perl 5.6 will be able to do with (?p{...}).
202
203 11. Added support for POSIX character classes like [:alpha:], which Perl is
204 adopting.
205
206
207 Version 2.08 31-Aug-99
208 ----------------------
209
210 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
211 trying to match at the startoffset position, but instead was moving forward to
212 the next newline as if a previous match had failed.
213
214 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
215 and could get into a loop if a null string was matched other than at the start
216 of the subject.
217
218 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
219 be distinguished at compile time, and for completeness also added PCRE_DATE.
220
221 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
222 in GnuWin32 environments.
223
224
225 Version 2.07 29-Jul-99
226 ----------------------
227
228 1. The documentation is now supplied in plain text form and HTML as well as in
229 the form of man page sources.
230
231 2. C++ compilers don't like assigning (void *) values to other pointer types.
232 In particular this affects malloc(). Although there is no problem in Standard
233 C, I've put in casts to keep C++ compilers happy.
234
235 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
236 should be (const char *).
237
238 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
239 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
240 However, I haven't made this a standard facility. The documentation doesn't
241 mention it, and the Makefile doesn't support it.
242
243 5. The Makefile now contains an "install" target, with editable destinations at
244 the top of the file. The pcretest program is not installed.
245
246 6. pgrep -V now gives the PCRE version number and date.
247
248 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
249 causing the entire string to be ignored, instead of just the last character.
250
251 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
252 non-matching string, it can take a very, very long time, even for strings of
253 quite modest length, because of the nested recursion. PCRE now does better in
254 some of these cases. It does this by remembering the last required literal
255 character in the pattern, and pre-searching the subject to ensure it is present
256 before running the real match. In other words, it applies a heuristic to detect
257 some types of certain failure quickly, and in the above example, if presented
258 with a string that has no trailing " it gives "no match" very quickly.
259
260 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
261 other alternatives are tried instead.
262
263
264 Version 2.06 09-Jun-99
265 ----------------------
266
267 1. Change pcretest's output for amount of store used to show just the code
268 space, because the remainder (the data block) varies in size between 32-bit and
269 64-bit systems.
270
271 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
272 start matching at. This allows lookbehinds to work when searching for multiple
273 occurrences in a string.
274
275 3. Added additional options to pcretest for testing multiple occurrences:
276
277 /+ outputs the rest of the string that follows a match
278 /g loops for multiple occurrences, using the new startoffset argument
279 /G loops for multiple occurrences by passing an incremented pointer
280
281 4. PCRE wasn't doing the "first character" optimization for patterns starting
282 with \b or \B, though it was doing it for other lookbehind assertions. That is,
283 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
284 the letter 'x'. On long subject strings, this gives a significant speed-up.
285
286
287 Version 2.05 21-Apr-99
288 ----------------------
289
290 1. Changed the type of magic_number from int to long int so that it works
291 properly on 16-bit systems.
292
293 2. Fixed a bug which caused patterns starting with .* not to work correctly
294 when the subject string contained newline characters. PCRE was assuming
295 anchoring for such patterns in all cases, which is not correct because .* will
296 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
297 DOTALL is set at top level; otherwise it knows that patterns starting with .*
298 must be retried after every newline in the subject.
299
300
301 Version 2.04 18-Feb-99
302 ----------------------
303
304 1. For parenthesized subpatterns with repeats whose minimum was zero, the
305 computation of the store needed to hold the pattern was incorrect (too large).
306 If such patterns were nested a few deep, this could multiply and become a real
307 problem.
308
309 2. Added /M option to pcretest to show the memory requirement of a specific
310 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
311
312 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
313 compiled in such a way that the backtracking after subsequent failure was
314 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
315 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
316
317
318 Version 2.03 02-Feb-99
319 ----------------------
320
321 1. Fixed typo and small mistake in man page.
322
323 2. Added 4th condition (GPL supersedes if conflict) and created separate
324 LICENCE file containing the conditions.
325
326 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
327 Perl, that is the internal \ allows the delimiter to be included in the
328 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
329 the final delimiter, add \ to the end of the pattern (to test the error).
330
331 4. Added the convenience functions for extracting substrings after a successful
332 match. Updated pcretest to make it able to test these functions.
333
334
335 Version 2.02 14-Jan-99
336 ----------------------
337
338 1. Initialized the working variables associated with each extraction so that
339 their saving and restoring doesn't refer to uninitialized store.
340
341 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
342 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
343 fix the problem.
344
345 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
346 calls, and wasn't printing the correct value for compiling calls. Increased the
347 default value of LOOPREPEAT, and the number of significant figures in the
348 times.
349
350 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
351
352 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
353 a building problem on Windows NT with a FAT file system.
354
355
356 Version 2.01 21-Oct-98
357 ----------------------
358
359 1. Changed the API for pcre_compile() to allow for the provision of a pointer
360 to character tables built by pcre_maketables() in the current locale. If NULL
361 is passed, the default tables are used.
362
363
364 Version 2.00 24-Sep-98
365 ----------------------
366
367 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
368 it any more.
369
370 2. Allow quantification of (?>) groups, and make it work correctly.
371
372 3. The first character computation wasn't working for (?>) groups.
373
374 4. Correct the implementation of \Z (it is permitted to match on the \n at the
375 end of the subject) and add 5.005's \z, which really does match only at the
376 very end of the subject.
377
378 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
379
380 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
381 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
382 localized options. All options to pcre_study() were also removed.
383
384 7. Add other new features from 5.005:
385
386 $(?<= positive lookbehind
387 $(?<! negative lookbehind
388 (?imsx-imsx) added the unsetting capability
389 such a setting is global if at outer level; local otherwise
390 (?imsx-imsx:) non-capturing groups with option setting
391 (?(cond)re|re) conditional pattern matching
392
393 A backreference to itself in a repeated group matches the previous
394 captured string.
395
396 8. General tidying up of studying (both automatic and via "study")
397 consequential on the addition of new assertions.
398
399 9. As in 5.005, unlimited repeated groups that could match an empty substring
400 are no longer faulted at compile time. Instead, the loop is forcibly broken at
401 runtime if any iteration does actually match an empty substring.
402
403 10. Include the RunTest script in the distribution.
404
405 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
406 discrepancies, some of which were old and were also with respect to 5.004. They
407 have now been fixed.
408
409
410 Version 1.09 28-Apr-98
411 ----------------------
412
413 1. A negated single character class followed by a quantifier with a minimum
414 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
415 program crashes, or just wrong answers. This did not apply to negated classes
416 containing more than one character, or to minima other than one.
417
418
419 Version 1.08 27-Mar-98
420 ----------------------
421
422 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
423
424 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
425 latter must appear before anything that relies on it in the pattern.
426
427
428 Version 1.07 16-Feb-98
429 ----------------------
430
431 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
432 repeat of a potentially empty string).
433
434
435 Version 1.06 23-Jan-98
436 ----------------------
437
438 1. Added Markus Oberhumer's little patches for C++.
439
440 2. Literal strings longer than 255 characters were broken.
441
442
443 Version 1.05 23-Dec-97
444 ----------------------
445
446 1. Negated character classes containing more than one character were failing if
447 PCRE_CASELESS was set at run time.
448
449
450 Version 1.04 19-Dec-97
451 ----------------------
452
453 1. Corrected the man page, where some "const" qualifiers had been omitted.
454
455 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
456 input syntax.
457
458 3. Fixed memory leak which occurred when a regex with back references was
459 matched with an offsets vector that wasn't big enough. The temporary memory
460 that is used in this case wasn't being freed if the match failed.
461
462 4. Tidied pcretest to ensure it frees memory that it gets.
463
464 5. Temporary memory was being obtained in the case where the passed offsets
465 vector was exactly big enough.
466
467 6. Corrected definition of offsetof() from change 5 below.
468
469 7. I had screwed up change 6 below and broken the rules for the use of
470 setjmp(). Now fixed.
471
472
473 Version 1.03 18-Dec-97
474 ----------------------
475
476 1. A erroneous regex with a missing opening parenthesis was correctly
477 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
478 on some systems.
479
480 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
481 it was reported that one broken compiler failed on the former because "code" is
482 also an independent variable.
483
484 3. The erroneous regex a[]b caused an array overrun reference.
485
486 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
487 fail on data ending with that character. (It was going on too far, and checking
488 the next character, typically a binary zero.) This was specific to the
489 optimized code for single-character negative classes.
490
491 5. Added a contributed patch from the TIN world which does the following:
492
493 + Add an undef for memmove, in case the the system defines a macro for it.
494
495 + Add a definition of offsetof(), in case there isn't one. (I don't know
496 the reason behind this - offsetof() is part of the ANSI standard - but
497 it does no harm).
498
499 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
500 most of the places where whitespace preceded '#'. I have given up and
501 allowed the remaining 2 cases to be at the margin.
502
503 + Rename some variables in pcre to eliminate shadowing. This seems very
504 pedantic, but does no harm, of course.
505
506 6. Moved the call to setjmp() into its own function, to get rid of warnings
507 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
508
509 7. Constructs such as \d{8,} were compiling into the equivalent of
510 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
511 outcome, but in this particular case used more store than had been allocated,
512 which caused the bug to be discovered because it threw up an internal error.
513
514 8. The debugging code in both pcre and pcretest for outputting the compiled
515 form of a regex was going wrong in the case of back references followed by
516 curly-bracketed repeats.
517
518
519 Version 1.02 12-Dec-97
520 ----------------------
521
522 1. Typos in pcre.3 and comments in the source fixed.
523
524 2. Applied a contributed patch to get rid of places where it used to remove
525 'const' from variables, and fixed some signed/unsigned and uninitialized
526 variable warnings.
527
528 3. Added the "runtest" target to Makefile.
529
530 4. Set default compiler flag to -O2 rather than just -O.
531
532
533 Version 1.01 19-Nov-97
534 ----------------------
535
536 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
537 like /([ab]*)*/, that is, for classes with more than one character in them.
538
539 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
540 as /((?>a*))*/ (a PCRE_EXTRA facility).
541
542
543 Version 1.00 18-Nov-97
544 ----------------------
545
546 1. Added compile-time macros to support systems such as SunOS4 which don't have
547 memmove() or strerror() but have other things that can be used instead.
548
549 2. Arranged that "make clean" removes the executables.
550
551
552 Version 0.99 27-Oct-97
553 ----------------------
554
555 1. Fixed bug in code for optimizing classes with only one character. It was
556 initializing a 32-byte map regardless, which could cause it to run off the end
557 of the memory it had got.
558
559 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
560
561
562 Version 0.98 22-Oct-97
563 ----------------------
564
565 1. Fixed bug in code for handling temporary memory usage when there are more
566 back references than supplied space in the ovector. This could cause segfaults.
567
568
569 Version 0.97 21-Oct-97
570 ----------------------
571
572 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
573
574 2. Optimized negated single characters not to use a bit map.
575
576 3. Brought error texts together as macro definitions; clarified some of them;
577 fixed one that was wrong - it said "range out of order" when it meant "invalid
578 escape sequence".
579
580 4. Changed some char * arguments to const char *.
581
582 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
583
584 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
585 pcretest.
586
587
588 Version 0.96 16-Oct-97
589 ----------------------
590
591 1. Added a simple "pgrep" utility to the distribution.
592
593 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
594 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
595 where "ddd" means "one or more decimal digits".
596
597 3. Fixed serious bug. If a pattern had a back reference, but the call to
598 pcre_exec() didn't supply a large enough ovector to record the related
599 identifying subpattern, the match always failed. PCRE now remembers the number
600 of the largest back reference, and gets some temporary memory in which to save
601 the offsets during matching if necessary, in order to ensure that
602 backreferences always work.
603
604 4. Increased the compatibility with Perl in a number of ways:
605
606 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
607 to request this handling. The option can be set at compile or exec time.
608
609 (b) $ matches before a terminating newline by default; an option
610 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
611 mode). The option can be set at compile or exec time.
612
613 (c) The handling of \ followed by a digit other than 0 is now supposed to be
614 the same as Perl's. If the decimal number it represents is less than 10
615 or there aren't that many previous left capturing parentheses, an octal
616 escape is read. Inside a character class, it's always an octal escape,
617 even if it is a single digit.
618
619 (d) An escaped but undefined alphabetic character is taken as a literal,
620 unless PCRE_EXTRA is set. Currently this just reserves the remaining
621 escapes.
622
623 (e) {0} is now permitted. (The previous item is removed from the compiled
624 pattern).
625
626 5. Changed all the names of code files so that the basic parts are no longer
627 than 10 characters, and abolished the teeny "globals.c" file.
628
629 6. Changed the handling of character classes; they are now done with a 32-byte
630 bit map always.
631
632 7. Added the -d and /D options to pcretest to make it possible to look at the
633 internals of compilation without having to recompile pcre.
634
635
636 Version 0.95 23-Sep-97
637 ----------------------
638
639 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
640 \x20 at the start of a run of normal characters. These were being treated as
641 real characters, instead of the source characters being re-checked.
642
643
644 Version 0.94 18-Sep-97
645 ----------------------
646
647 1. The functions are now thread-safe, with the caveat that the global variables
648 containing pointers to malloc() and free() or alternative functions are the
649 same for all threads.
650
651 2. Get pcre_study() to generate a bitmap of initial characters for non-
652 anchored patterns when this is possible, and use it if passed to pcre_exec().
653
654
655 Version 0.93 15-Sep-97
656 ----------------------
657
658 1. /(b)|(:+)/ was computing an incorrect first character.
659
660 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
661 but not actually doing anything yet.
662
663 3. Treat "-" characters in classes that cannot be part of ranges as literals,
664 as Perl does (e.g. [-az] or [az-]).
665
666 4. Set the anchored flag if a branch starts with .* or .*? because that tests
667 all possible positions.
668
669 5. Split up into different modules to avoid including unneeded functions in a
670 compiled binary. However, compile and exec are still in one module. The "study"
671 function is split off.
672
673 6. The character tables are now in a separate module whose source is generated
674 by an auxiliary program - but can then be edited by hand if required. There are
675 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
676 toupper() in the code.
677
678 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
679 make them global. Abolish the function for setting them, as the caller can now
680 set them directly.
681
682
683 Version 0.92 11-Sep-97
684 ----------------------
685
686 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
687 (e.g. /a{1,3}/) was broken (I mis-optimized it).
688
689 2. Caseless matching was not working in character classes if the characters in
690 the pattern were in upper case.
691
692 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
693
694 4. Make PCRE_ANCHORED public and accept as a compile option.
695
696 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
697 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
698 pass them.
699
700 6. Give an error if bad option bits passed at compile or run time.
701
702 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
703 pcretest to cause it to pass that flag.
704
705 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
706 options, and the first character, if set.
707
708 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
709
710
711 Version 0.91 10-Sep-97
712 ----------------------
713
714 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
715 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
716
717 2. PCRE was looping on encountering an indefinitely repeated back reference to
718 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
719 Perl does - treats the match as successful.
720
721 ****

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12