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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 49 - (show annotations) (download)
Sat Feb 24 21:39:33 2007 UTC (7 years, 1 month ago) by nigel
File size: 24604 byte(s)
Load pcre-3.3 into code/trunk.

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12