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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 59 - (show annotations) (download)
Sat Feb 24 21:39:54 2007 UTC (7 years, 9 months ago) by nigel
File size: 28303 byte(s)
Load pcre-3.8 into code/trunk.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 3.8 18-Dec-01
5 ---------------------
6
7 1. The experimental UTF-8 code was completely screwed up. It was packing the
8 bytes in the wrong order. How dumb can you get?
9
10
11 Version 3.7 29-Oct-01
12 ---------------------
13
14 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
15 This caused pcretest, when used on the test data, to segfault. Unfortunately,
16 this didn't happen under Solaris 8, where I normally test things.
17
18 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
19 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
20 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
21 this fix an hour or so after the initial 3.7 release.)
22
23
24 Version 3.6 23-Oct-01
25 ---------------------
26
27 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
28 offsets passed as NULL with zero offset count.
29
30 2. The config.guess and config.sub files had not been updated when I moved to
31 the latest autoconf.
32
33
34 Version 3.5 15-Aug-01
35 ---------------------
36
37 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
38 had been forgotten.
39
40 2. By using declared but undefined structures, we can avoid using "void"
41 definitions in pcre.h while keeping the internal definitions of the structures
42 private.
43
44 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
45 user point of view, this means that both static and shared libraries are built
46 by default, but this can be individually controlled. More of the work of
47 handling this static/shared cases is now inside libtool instead of PCRE's make
48 file.
49
50 4. The pcretest utility is now installed along with pcregrep because it is
51 useful for users (to test regexs) and by doing this, it automatically gets
52 relinked by libtool. The documentation has been turned into a man page, so
53 there are now .1, .txt, and .html versions in /doc.
54
55 5. Upgrades to pcregrep:
56 (i) Added long-form option names like gnu grep.
57 (ii) Added --help to list all options with an explanatory phrase.
58 (iii) Added -r, --recursive to recurse into sub-directories.
59 (iv) Added -f, --file to read patterns from a file.
60
61 6. pcre_exec() was referring to its "code" argument before testing that
62 argument for NULL (and giving an error if it was NULL).
63
64 7. Upgraded Makefile.in to allow for compiling in a different directory from
65 the source directory.
66
67 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
68 options bits, the pointer it was passed was to an int instead of to an unsigned
69 long int. This mattered only on 64-bit systems.
70
71 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
72 generated) instead of pcre.in, which it its source. Also made the same change
73 in several of the .c files.
74
75 10. A new release of gcc defines printf() as a macro, which broke pcretest
76 because it had an ifdef in the middle of a string argument for printf(). Fixed
77 by using separate calls to printf().
78
79 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
80 script, to force use of CR or LF instead of \n in the source. On non-Unix
81 systems, the value can be set in config.h.
82
83 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
84 absolute limit. Changed the text of the error message to make this clear, and
85 likewise updated the man page.
86
87 13. The limit of 99 on the number of capturing subpatterns has been removed.
88 The new limit is 65535, which I hope will not be a "real" limit.
89
90
91 Version 3.4 22-Aug-00
92 ---------------------
93
94 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
95
96 2. Diagnose condition (?(0) as an error instead of crashing on matching.
97
98
99 Version 3.3 01-Aug-00
100 ---------------------
101
102 1. If an octal character was given, but the value was greater than \377, it
103 was not getting masked to the least significant bits, as documented. This could
104 lead to crashes in some systems.
105
106 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
107 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
108
109 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
110 These just pass their arguments on to (pcre_free)(), but they are provided
111 because some uses of PCRE bind it to non-C systems that can call its functions,
112 but cannot call free() or pcre_free() directly.
113
114 4. Add "make test" as a synonym for "make check". Corrected some comments in
115 the Makefile.
116
117 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
118 Makefile.
119
120 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
121 command called pgrep for grepping around the active processes.
122
123 7. Added the beginnings of support for UTF-8 character strings.
124
125 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
126 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
127 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
128 out for the ar command.)
129
130
131 Version 3.2 12-May-00
132 ---------------------
133
134 This is purely a bug fixing release.
135
136 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
137 of ZA. This was just one example of several cases that could provoke this bug,
138 which was introduced by change 9 of version 2.00. The code for breaking
139 infinite loops after an iteration that matches an empty string was't working
140 correctly.
141
142 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
143 when matched against abbab (for example). After matching an empty string, it
144 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
145 caused it to match further down the string than it should.
146
147 3. The code contained an inclusion of sys/types.h. It isn't clear why this
148 was there because it doesn't seem to be needed, and it causes trouble on some
149 systems, as it is not a Standard C header. It has been removed.
150
151 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
152 were reported on the Macintosh. The changes were from
153
154 while ((c = *(++ptr)) != 0 && c != '\n');
155 to
156 while ((c = *(++ptr)) != 0 && c != '\n') ;
157
158 Totally extraordinary, but if that's what it takes...
159
160 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
161 available. Added HAVE_BCOPY and an autoconf test for it; if neither
162 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
163 assumes the way PCRE uses memmove() (always moving upwards).
164
165 6. PCRE is being used in one environment where strchr() is not available. There
166 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
167 faster code anyway.
168
169
170 Version 3.1 09-Feb-00
171 ---------------------
172
173 The only change in this release is the fixing of some bugs in Makefile.in for
174 the "install" target:
175
176 (1) It was failing to install pcreposix.h.
177
178 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
179
180
181 Version 3.0 01-Feb-00
182 ---------------------
183
184 1. Add support for the /+ modifier to perltest (to output $` like it does in
185 pcretest).
186
187 2. Add support for the /g modifier to perltest.
188
189 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
190 matches null strings.
191
192 4. Fix perltest so that it doesn't do unwanted things when fed an empty
193 pattern. Perl treats empty patterns specially - it reuses the most recent
194 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
195 effect.
196
197 5. The POSIX interface was broken in that it was just handing over the POSIX
198 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
199 required a bigger vector, with some working space on the end. This means that
200 the POSIX wrapper now has to get and free some memory, and copy the results.
201
202 6. Added some simple autoconf support, placing the test data and the
203 documentation in separate directories, re-organizing some of the
204 information files, and making it build pcre-config (a GNU standard). Also added
205 libtool support for building PCRE as a shared library, which is now the
206 default.
207
208 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
209 09 are not valid octal constants. Single digits will be used for minor values
210 less than 10.
211
212 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
213 existing programs that set these in the POSIX interface can use PCRE without
214 modification.
215
216 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
217 return all that pcre_info() returns, plus additional data. The pcre_info()
218 function is retained for compatibility, but is considered to be obsolete.
219
220 10. Added experimental recursion feature (?R) to handle one common case that
221 Perl 5.6 will be able to do with (?p{...}).
222
223 11. Added support for POSIX character classes like [:alpha:], which Perl is
224 adopting.
225
226
227 Version 2.08 31-Aug-99
228 ----------------------
229
230 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
231 trying to match at the startoffset position, but instead was moving forward to
232 the next newline as if a previous match had failed.
233
234 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
235 and could get into a loop if a null string was matched other than at the start
236 of the subject.
237
238 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
239 be distinguished at compile time, and for completeness also added PCRE_DATE.
240
241 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
242 in GnuWin32 environments.
243
244
245 Version 2.07 29-Jul-99
246 ----------------------
247
248 1. The documentation is now supplied in plain text form and HTML as well as in
249 the form of man page sources.
250
251 2. C++ compilers don't like assigning (void *) values to other pointer types.
252 In particular this affects malloc(). Although there is no problem in Standard
253 C, I've put in casts to keep C++ compilers happy.
254
255 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
256 should be (const char *).
257
258 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
259 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
260 However, I haven't made this a standard facility. The documentation doesn't
261 mention it, and the Makefile doesn't support it.
262
263 5. The Makefile now contains an "install" target, with editable destinations at
264 the top of the file. The pcretest program is not installed.
265
266 6. pgrep -V now gives the PCRE version number and date.
267
268 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
269 causing the entire string to be ignored, instead of just the last character.
270
271 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
272 non-matching string, it can take a very, very long time, even for strings of
273 quite modest length, because of the nested recursion. PCRE now does better in
274 some of these cases. It does this by remembering the last required literal
275 character in the pattern, and pre-searching the subject to ensure it is present
276 before running the real match. In other words, it applies a heuristic to detect
277 some types of certain failure quickly, and in the above example, if presented
278 with a string that has no trailing " it gives "no match" very quickly.
279
280 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
281 other alternatives are tried instead.
282
283
284 Version 2.06 09-Jun-99
285 ----------------------
286
287 1. Change pcretest's output for amount of store used to show just the code
288 space, because the remainder (the data block) varies in size between 32-bit and
289 64-bit systems.
290
291 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
292 start matching at. This allows lookbehinds to work when searching for multiple
293 occurrences in a string.
294
295 3. Added additional options to pcretest for testing multiple occurrences:
296
297 /+ outputs the rest of the string that follows a match
298 /g loops for multiple occurrences, using the new startoffset argument
299 /G loops for multiple occurrences by passing an incremented pointer
300
301 4. PCRE wasn't doing the "first character" optimization for patterns starting
302 with \b or \B, though it was doing it for other lookbehind assertions. That is,
303 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
304 the letter 'x'. On long subject strings, this gives a significant speed-up.
305
306
307 Version 2.05 21-Apr-99
308 ----------------------
309
310 1. Changed the type of magic_number from int to long int so that it works
311 properly on 16-bit systems.
312
313 2. Fixed a bug which caused patterns starting with .* not to work correctly
314 when the subject string contained newline characters. PCRE was assuming
315 anchoring for such patterns in all cases, which is not correct because .* will
316 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
317 DOTALL is set at top level; otherwise it knows that patterns starting with .*
318 must be retried after every newline in the subject.
319
320
321 Version 2.04 18-Feb-99
322 ----------------------
323
324 1. For parenthesized subpatterns with repeats whose minimum was zero, the
325 computation of the store needed to hold the pattern was incorrect (too large).
326 If such patterns were nested a few deep, this could multiply and become a real
327 problem.
328
329 2. Added /M option to pcretest to show the memory requirement of a specific
330 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
331
332 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
333 compiled in such a way that the backtracking after subsequent failure was
334 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
335 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
336
337
338 Version 2.03 02-Feb-99
339 ----------------------
340
341 1. Fixed typo and small mistake in man page.
342
343 2. Added 4th condition (GPL supersedes if conflict) and created separate
344 LICENCE file containing the conditions.
345
346 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
347 Perl, that is the internal \ allows the delimiter to be included in the
348 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
349 the final delimiter, add \ to the end of the pattern (to test the error).
350
351 4. Added the convenience functions for extracting substrings after a successful
352 match. Updated pcretest to make it able to test these functions.
353
354
355 Version 2.02 14-Jan-99
356 ----------------------
357
358 1. Initialized the working variables associated with each extraction so that
359 their saving and restoring doesn't refer to uninitialized store.
360
361 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
362 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
363 fix the problem.
364
365 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
366 calls, and wasn't printing the correct value for compiling calls. Increased the
367 default value of LOOPREPEAT, and the number of significant figures in the
368 times.
369
370 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
371
372 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
373 a building problem on Windows NT with a FAT file system.
374
375
376 Version 2.01 21-Oct-98
377 ----------------------
378
379 1. Changed the API for pcre_compile() to allow for the provision of a pointer
380 to character tables built by pcre_maketables() in the current locale. If NULL
381 is passed, the default tables are used.
382
383
384 Version 2.00 24-Sep-98
385 ----------------------
386
387 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
388 it any more.
389
390 2. Allow quantification of (?>) groups, and make it work correctly.
391
392 3. The first character computation wasn't working for (?>) groups.
393
394 4. Correct the implementation of \Z (it is permitted to match on the \n at the
395 end of the subject) and add 5.005's \z, which really does match only at the
396 very end of the subject.
397
398 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
399
400 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
401 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
402 localized options. All options to pcre_study() were also removed.
403
404 7. Add other new features from 5.005:
405
406 $(?<= positive lookbehind
407 $(?<! negative lookbehind
408 (?imsx-imsx) added the unsetting capability
409 such a setting is global if at outer level; local otherwise
410 (?imsx-imsx:) non-capturing groups with option setting
411 (?(cond)re|re) conditional pattern matching
412
413 A backreference to itself in a repeated group matches the previous
414 captured string.
415
416 8. General tidying up of studying (both automatic and via "study")
417 consequential on the addition of new assertions.
418
419 9. As in 5.005, unlimited repeated groups that could match an empty substring
420 are no longer faulted at compile time. Instead, the loop is forcibly broken at
421 runtime if any iteration does actually match an empty substring.
422
423 10. Include the RunTest script in the distribution.
424
425 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
426 discrepancies, some of which were old and were also with respect to 5.004. They
427 have now been fixed.
428
429
430 Version 1.09 28-Apr-98
431 ----------------------
432
433 1. A negated single character class followed by a quantifier with a minimum
434 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
435 program crashes, or just wrong answers. This did not apply to negated classes
436 containing more than one character, or to minima other than one.
437
438
439 Version 1.08 27-Mar-98
440 ----------------------
441
442 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
443
444 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
445 latter must appear before anything that relies on it in the pattern.
446
447
448 Version 1.07 16-Feb-98
449 ----------------------
450
451 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
452 repeat of a potentially empty string).
453
454
455 Version 1.06 23-Jan-98
456 ----------------------
457
458 1. Added Markus Oberhumer's little patches for C++.
459
460 2. Literal strings longer than 255 characters were broken.
461
462
463 Version 1.05 23-Dec-97
464 ----------------------
465
466 1. Negated character classes containing more than one character were failing if
467 PCRE_CASELESS was set at run time.
468
469
470 Version 1.04 19-Dec-97
471 ----------------------
472
473 1. Corrected the man page, where some "const" qualifiers had been omitted.
474
475 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
476 input syntax.
477
478 3. Fixed memory leak which occurred when a regex with back references was
479 matched with an offsets vector that wasn't big enough. The temporary memory
480 that is used in this case wasn't being freed if the match failed.
481
482 4. Tidied pcretest to ensure it frees memory that it gets.
483
484 5. Temporary memory was being obtained in the case where the passed offsets
485 vector was exactly big enough.
486
487 6. Corrected definition of offsetof() from change 5 below.
488
489 7. I had screwed up change 6 below and broken the rules for the use of
490 setjmp(). Now fixed.
491
492
493 Version 1.03 18-Dec-97
494 ----------------------
495
496 1. A erroneous regex with a missing opening parenthesis was correctly
497 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
498 on some systems.
499
500 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
501 it was reported that one broken compiler failed on the former because "code" is
502 also an independent variable.
503
504 3. The erroneous regex a[]b caused an array overrun reference.
505
506 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
507 fail on data ending with that character. (It was going on too far, and checking
508 the next character, typically a binary zero.) This was specific to the
509 optimized code for single-character negative classes.
510
511 5. Added a contributed patch from the TIN world which does the following:
512
513 + Add an undef for memmove, in case the the system defines a macro for it.
514
515 + Add a definition of offsetof(), in case there isn't one. (I don't know
516 the reason behind this - offsetof() is part of the ANSI standard - but
517 it does no harm).
518
519 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
520 most of the places where whitespace preceded '#'. I have given up and
521 allowed the remaining 2 cases to be at the margin.
522
523 + Rename some variables in pcre to eliminate shadowing. This seems very
524 pedantic, but does no harm, of course.
525
526 6. Moved the call to setjmp() into its own function, to get rid of warnings
527 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
528
529 7. Constructs such as \d{8,} were compiling into the equivalent of
530 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
531 outcome, but in this particular case used more store than had been allocated,
532 which caused the bug to be discovered because it threw up an internal error.
533
534 8. The debugging code in both pcre and pcretest for outputting the compiled
535 form of a regex was going wrong in the case of back references followed by
536 curly-bracketed repeats.
537
538
539 Version 1.02 12-Dec-97
540 ----------------------
541
542 1. Typos in pcre.3 and comments in the source fixed.
543
544 2. Applied a contributed patch to get rid of places where it used to remove
545 'const' from variables, and fixed some signed/unsigned and uninitialized
546 variable warnings.
547
548 3. Added the "runtest" target to Makefile.
549
550 4. Set default compiler flag to -O2 rather than just -O.
551
552
553 Version 1.01 19-Nov-97
554 ----------------------
555
556 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
557 like /([ab]*)*/, that is, for classes with more than one character in them.
558
559 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
560 as /((?>a*))*/ (a PCRE_EXTRA facility).
561
562
563 Version 1.00 18-Nov-97
564 ----------------------
565
566 1. Added compile-time macros to support systems such as SunOS4 which don't have
567 memmove() or strerror() but have other things that can be used instead.
568
569 2. Arranged that "make clean" removes the executables.
570
571
572 Version 0.99 27-Oct-97
573 ----------------------
574
575 1. Fixed bug in code for optimizing classes with only one character. It was
576 initializing a 32-byte map regardless, which could cause it to run off the end
577 of the memory it had got.
578
579 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
580
581
582 Version 0.98 22-Oct-97
583 ----------------------
584
585 1. Fixed bug in code for handling temporary memory usage when there are more
586 back references than supplied space in the ovector. This could cause segfaults.
587
588
589 Version 0.97 21-Oct-97
590 ----------------------
591
592 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
593
594 2. Optimized negated single characters not to use a bit map.
595
596 3. Brought error texts together as macro definitions; clarified some of them;
597 fixed one that was wrong - it said "range out of order" when it meant "invalid
598 escape sequence".
599
600 4. Changed some char * arguments to const char *.
601
602 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
603
604 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
605 pcretest.
606
607
608 Version 0.96 16-Oct-97
609 ----------------------
610
611 1. Added a simple "pgrep" utility to the distribution.
612
613 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
614 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
615 where "ddd" means "one or more decimal digits".
616
617 3. Fixed serious bug. If a pattern had a back reference, but the call to
618 pcre_exec() didn't supply a large enough ovector to record the related
619 identifying subpattern, the match always failed. PCRE now remembers the number
620 of the largest back reference, and gets some temporary memory in which to save
621 the offsets during matching if necessary, in order to ensure that
622 backreferences always work.
623
624 4. Increased the compatibility with Perl in a number of ways:
625
626 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
627 to request this handling. The option can be set at compile or exec time.
628
629 (b) $ matches before a terminating newline by default; an option
630 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
631 mode). The option can be set at compile or exec time.
632
633 (c) The handling of \ followed by a digit other than 0 is now supposed to be
634 the same as Perl's. If the decimal number it represents is less than 10
635 or there aren't that many previous left capturing parentheses, an octal
636 escape is read. Inside a character class, it's always an octal escape,
637 even if it is a single digit.
638
639 (d) An escaped but undefined alphabetic character is taken as a literal,
640 unless PCRE_EXTRA is set. Currently this just reserves the remaining
641 escapes.
642
643 (e) {0} is now permitted. (The previous item is removed from the compiled
644 pattern).
645
646 5. Changed all the names of code files so that the basic parts are no longer
647 than 10 characters, and abolished the teeny "globals.c" file.
648
649 6. Changed the handling of character classes; they are now done with a 32-byte
650 bit map always.
651
652 7. Added the -d and /D options to pcretest to make it possible to look at the
653 internals of compilation without having to recompile pcre.
654
655
656 Version 0.95 23-Sep-97
657 ----------------------
658
659 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
660 \x20 at the start of a run of normal characters. These were being treated as
661 real characters, instead of the source characters being re-checked.
662
663
664 Version 0.94 18-Sep-97
665 ----------------------
666
667 1. The functions are now thread-safe, with the caveat that the global variables
668 containing pointers to malloc() and free() or alternative functions are the
669 same for all threads.
670
671 2. Get pcre_study() to generate a bitmap of initial characters for non-
672 anchored patterns when this is possible, and use it if passed to pcre_exec().
673
674
675 Version 0.93 15-Sep-97
676 ----------------------
677
678 1. /(b)|(:+)/ was computing an incorrect first character.
679
680 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
681 but not actually doing anything yet.
682
683 3. Treat "-" characters in classes that cannot be part of ranges as literals,
684 as Perl does (e.g. [-az] or [az-]).
685
686 4. Set the anchored flag if a branch starts with .* or .*? because that tests
687 all possible positions.
688
689 5. Split up into different modules to avoid including unneeded functions in a
690 compiled binary. However, compile and exec are still in one module. The "study"
691 function is split off.
692
693 6. The character tables are now in a separate module whose source is generated
694 by an auxiliary program - but can then be edited by hand if required. There are
695 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
696 toupper() in the code.
697
698 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
699 make them global. Abolish the function for setting them, as the caller can now
700 set them directly.
701
702
703 Version 0.92 11-Sep-97
704 ----------------------
705
706 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
707 (e.g. /a{1,3}/) was broken (I mis-optimized it).
708
709 2. Caseless matching was not working in character classes if the characters in
710 the pattern were in upper case.
711
712 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
713
714 4. Make PCRE_ANCHORED public and accept as a compile option.
715
716 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
717 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
718 pass them.
719
720 6. Give an error if bad option bits passed at compile or run time.
721
722 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
723 pcretest to cause it to pass that flag.
724
725 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
726 options, and the first character, if set.
727
728 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
729
730
731 Version 0.91 10-Sep-97
732 ----------------------
733
734 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
735 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
736
737 2. PCRE was looping on encountering an indefinitely repeated back reference to
738 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
739 Perl does - treats the match as successful.
740
741 ****

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12