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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 53 - (show annotations) (download)
Sat Feb 24 21:39:42 2007 UTC (7 years, 6 months ago) by nigel
File size: 27307 byte(s)
Load pcre-3.5 into code/trunk.

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12