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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 481 - (show annotations) (download)
Sun Jan 3 17:45:53 2010 UTC (4 years, 10 months ago) by ph10
File size: 179096 byte(s)
Forbid 08 and 09 as minor version numbers.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.01 11-Dec-09
5 ----------------------
6
7 1. If a pattern contained a conditional subpattern with only one branch (in
8 particular, this includes all (DEFINE) patterns), a call to pcre_study()
9 computed the wrong minimum data length (which is of course zero for such
10 subpatterns).
11
12 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
13 the pattern is reset in the first branch, pcre_compile() failed with
14 "internal error: code overflow at offset...". This happened only when
15 the reset was to the original external option setting. (An optimization
16 abstracts leading options settings into an external setting, which was the
17 cause of this.)
18
19 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
20 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
21 assertion pattern did not match (meaning that the assertion was true), it
22 was incorrectly treated as false if the SKIP had been reached during the
23 matching. This also applied to assertions used as conditions.
24
25 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
26 assertion subpattern, including such a pattern used as a condition,
27 unpredictable results occurred, instead of the error return
28 PCRE_ERROR_DFA_UITEM.
29
30 5. The C++ GlobalReplace function was not working like Perl for the special
31 situation when an empty string is matched. It now does the fancy magic
32 stuff that is necessary.
33
34 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
35 removed. (These were left over from very, very early versions of PCRE.)
36
37 7. Some cosmetic changes to the code to make life easier when compiling it
38 as part of something else:
39
40 (a) Change DEBUG to PCRE_DEBUG.
41
42 (b) In pcre_compile(), rename the member of the "branch_chain" structure
43 called "current" as "current_branch", to prevent a collision with the
44 Linux macro when compiled as a kernel module.
45
46 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
47 prevent a collision with the Linux macro when compiled as a kernel
48 module.
49
50 8. In pcre_compile() there are some checks for integer overflows that used to
51 cast potentially large values to (double). This has been changed to that
52 when building, a check for int64_t is made, and if it is found, it is used
53 instead, thus avoiding the use of floating point arithmetic. (There is no
54 other use of FP in PCRE.) If int64_t is not found, the fallback is to
55 double.
56
57 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
58 2005 (difference between two addresses compared to an unsigned value).
59
60 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
61 custom one, because of the following reported problem in Windows:
62
63 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
64 under Win32.
65 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
66 therefore missing the function definition.
67 - The compiler thus generates a "C" signature for the test function.
68 - The linker fails to find the "C" function.
69 - PCRE fails to configure if asked to do so against libbz2.
70
71 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
72 messages were output:
73
74 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
75 rerunning libtoolize, to keep the correct libtool macros in-tree.
76 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
77
78 I have done both of these things.
79
80 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
81 most of the time, it *can* run out if it is given a pattern that contains a
82 runaway infinite recursion. I updated the discussion in the pcrestack man
83 page.
84
85 13. Now that we have gone to the x.xx style of version numbers, the minor
86 version may start with zero. Using 08 or 09 is a bad idea because users
87 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
88 interpreted as invalid octal numbers. I've updated the previous comment in
89 configure.ac, and also added a check that gives an error if 08 or 09 are
90 used.
91
92
93
94 Version 8.00 19-Oct-09
95 ----------------------
96
97 1. The table for translating pcre_compile() error codes into POSIX error codes
98 was out-of-date, and there was no check on the pcre_compile() error code
99 being within the table. This could lead to an OK return being given in
100 error.
101
102 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
103 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
104 in a Windows environment.
105
106 3. The pcregrep --count option prints the count for each file even when it is
107 zero, as does GNU grep. However, pcregrep was also printing all files when
108 --files-with-matches was added. Now, when both options are given, it prints
109 counts only for those files that have at least one match. (GNU grep just
110 prints the file name in this circumstance, but including the count seems
111 more useful - otherwise, why use --count?) Also ensured that the
112 combination -clh just lists non-zero counts, with no names.
113
114 4. The long form of the pcregrep -F option was incorrectly implemented as
115 --fixed_strings instead of --fixed-strings. This is an incompatible change,
116 but it seems right to fix it, and I didn't think it was worth preserving
117 the old behaviour.
118
119 5. The command line items --regex=pattern and --regexp=pattern were not
120 recognized by pcregrep, which required --regex pattern or --regexp pattern
121 (with a space rather than an '='). The man page documented the '=' forms,
122 which are compatible with GNU grep; these now work.
123
124 6. No libpcreposix.pc file was created for pkg-config; there was just
125 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
126
127 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
128 when UCP support is not needed, by modifying the Python script that
129 generates it from Unicode data files. This should not matter if the module
130 is correctly used as a library, but I received one complaint about 50K of
131 unwanted data. My guess is that the person linked everything into his
132 program rather than using a library. Anyway, it does no harm.
133
134 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
135 was a minimum greater than 1 for a wide character in a possessive
136 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
137 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
138 character. Chaos in the form of incorrect output or a compiling loop could
139 result.
140
141 9. The restrictions on what a pattern can contain when partial matching is
142 requested for pcre_exec() have been removed. All patterns can now be
143 partially matched by this function. In addition, if there are at least two
144 slots in the offset vector, the offset of the earliest inspected character
145 for the match and the offset of the end of the subject are set in them when
146 PCRE_ERROR_PARTIAL is returned.
147
148 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
149 synonymous with PCRE_PARTIAL, for backwards compatibility, and
150 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
151 and may be more useful for multi-segment matching.
152
153 11. Partial matching with pcre_exec() is now more intuitive. A partial match
154 used to be given if ever the end of the subject was reached; now it is
155 given only if matching could not proceed because another character was
156 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
157 string "Z", which now yields "no match" instead of "partial match". In the
158 case of pcre_dfa_exec(), "no match" is given if every matching path for the
159 final character ended with (*FAIL).
160
161 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
162 if the pattern had a "must contain" character that was already found in the
163 earlier partial match, unless partial matching was again requested. For
164 example, with the pattern /dog.(body)?/, the "must contain" character is
165 "g". If the first part-match was for the string "dog", restarting with
166 "sbody" failed. This bug has been fixed.
167
168 13. The string returned by pcre_dfa_exec() after a partial match has been
169 changed so that it starts at the first inspected character rather than the
170 first character of the match. This makes a difference only if the pattern
171 starts with a lookbehind assertion or \b or \B (\K is not supported by
172 pcre_dfa_exec()). It's an incompatible change, but it makes the two
173 matching functions compatible, and I think it's the right thing to do.
174
175 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
176 so that the demonstration program is easily available in environments where
177 PCRE has not been installed from source.
178
179 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
180 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
181 library.
182
183 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
184 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
185 is not the first non-POSIX option to be added. Clearly some people find
186 these options useful.
187
188 17. If a caller to the POSIX matching function regexec() passes a non-zero
189 value for nmatch with a NULL value for pmatch, the value of
190 nmatch is forced to zero.
191
192 18. RunGrepTest did not have a test for the availability of the -u option of
193 the diff command, as RunTest does. It now checks in the same way as
194 RunTest, and also checks for the -b option.
195
196 19. If an odd number of negated classes containing just a single character
197 interposed, within parentheses, between a forward reference to a named
198 subpattern and the definition of the subpattern, compilation crashed with
199 an internal error, complaining that it could not find the referenced
200 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
201 [The bug was that it was starting one character too far in when skipping
202 over the character class, thus treating the ] as data rather than
203 terminating the class. This meant it could skip too much.]
204
205 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
206 /g option in pcretest when the pattern contains \K, which makes it possible
207 to have an empty string match not at the start, even when the pattern is
208 anchored. Updated pcretest and pcredemo to use this option.
209
210 21. If the maximum number of capturing subpatterns in a recursion was greater
211 than the maximum at the outer level, the higher number was returned, but
212 with unset values at the outer level. The correct (outer level) value is
213 now given.
214
215 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
216 PCRE did not set those parentheses (unlike Perl). I have now found a way to
217 make it do so. The string so far is captured, making this feature
218 compatible with Perl.
219
220 23. The tests have been re-organized, adding tests 11 and 12, to make it
221 possible to check the Perl 5.10 features against Perl 5.10.
222
223 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
224 pattern matches a fixed length string. PCRE did not allow this; now it
225 does. Neither allows recursion.
226
227 25. I finally figured out how to implement a request to provide the minimum
228 length of subject string that was needed in order to match a given pattern.
229 (It was back references and recursion that I had previously got hung up
230 on.) This code has now been added to pcre_study(); it finds a lower bound
231 to the length of subject needed. It is not necessarily the greatest lower
232 bound, but using it to avoid searching strings that are too short does give
233 some useful speed-ups. The value is available to calling programs via
234 pcre_fullinfo().
235
236 26. While implementing 25, I discovered to my embarrassment that pcretest had
237 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
238 study optimizations had never been tested with that matching function.
239 Oops. What is worse, even when it was passed study data, there was a bug in
240 pcre_dfa_exec() that meant it never actually used it. Double oops. There
241 were also very few tests of studied patterns with pcre_dfa_exec().
242
243 27. If (?| is used to create subpatterns with duplicate numbers, they are now
244 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
245 on the other side of the coin, they are no longer allowed to have different
246 names, because these cannot be distinguished in PCRE, and this has caused
247 confusion. (This is a difference from Perl.)
248
249 28. When duplicate subpattern names are present (necessarily with different
250 numbers, as required by 27 above), and a test is made by name in a
251 conditional pattern, either for a subpattern having been matched, or for
252 recursion in such a pattern, all the associated numbered subpatterns are
253 tested, and the overall condition is true if the condition is true for any
254 one of them. This is the way Perl works, and is also more like the way
255 testing by number works.
256
257
258 Version 7.9 11-Apr-09
259 ---------------------
260
261 1. When building with support for bzlib/zlib (pcregrep) and/or readline
262 (pcretest), all targets were linked against these libraries. This included
263 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
264 libraries. This caused unwanted dependencies to be created. This problem
265 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
266 pcretest is linked with readline.
267
268 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
269 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
270 moved outside it again, because FALSE and TRUE are already defined in AIX,
271 but BOOL is not.
272
273 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
274 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
275
276 4. The pcregrep documentation said spaces were inserted as well as colons (or
277 hyphens) following file names and line numbers when outputting matching
278 lines. This is not true; no spaces are inserted. I have also clarified the
279 wording for the --colour (or --color) option.
280
281 5. In pcregrep, when --colour was used with -o, the list of matching strings
282 was not coloured; this is different to GNU grep, so I have changed it to be
283 the same.
284
285 6. When --colo(u)r was used in pcregrep, only the first matching substring in
286 each matching line was coloured. Now it goes on to look for further matches
287 of any of the test patterns, which is the same behaviour as GNU grep.
288
289 7. A pattern that could match an empty string could cause pcregrep to loop; it
290 doesn't make sense to accept an empty string match in pcregrep, so I have
291 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
292 seems to be how GNU grep behaves.
293
294 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
295 start or after a newline", because the conditional assertion was not being
296 correctly handled. The rule now is that both the assertion and what follows
297 in the first alternative must satisfy the test.
298
299 9. If auto-callout was enabled in a pattern with a conditional group whose
300 condition was an assertion, PCRE could crash during matching, both with
301 pcre_exec() and pcre_dfa_exec().
302
303 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
304 used for matching.
305
306 11. Unicode property support in character classes was not working for
307 characters (bytes) greater than 127 when not in UTF-8 mode.
308
309 12. Added the -M command line option to pcretest.
310
311 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
312
313 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
314
315 16. Added comments and documentation about mis-use of no_arg in the C++
316 wrapper.
317
318 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
319 from Martin Jerabek that uses macro names for all relevant character and
320 string constants.
321
322 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
323 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
324 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
325 these, but not everybody uses configure.
326
327 19. A conditional group that had only one branch was not being correctly
328 recognized as an item that could match an empty string. This meant that an
329 enclosing group might also not be so recognized, causing infinite looping
330 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
331 with the subject "ab", where knowledge that the repeated group can match
332 nothing is needed in order to break the loop.
333
334 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
335 exec(), but without supplying a callout function, matching went wrong.
336
337 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
338 leak if the size of the offset vector was greater than 30. When the vector
339 is smaller, the saved offsets during recursion go onto a local stack
340 vector, but for larger vectors malloc() is used. It was failing to free
341 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
342 error, in fact).
343
344 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
345 heapframe that is used only when UTF-8 support is enabled. This caused no
346 problem, but was untidy.
347
348 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
349 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
350 included within another project.
351
352 24. Steven Van Ingelgem's patches to add more options to the CMake support,
353 slightly modified by me:
354
355 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
356 not building pcregrep.
357
358 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
359 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
360
361 25. Forward references, both numeric and by name, in patterns that made use of
362 duplicate group numbers, could behave incorrectly or give incorrect errors,
363 because when scanning forward to find the reference group, PCRE was not
364 taking into account the duplicate group numbers. A pattern such as
365 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
366
367 26. Changed a few more instances of "const unsigned char *" to USPTR, making
368 the feature of a custom pointer more persuasive (as requested by a user).
369
370 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
371 pcretest.c, inside #ifndefs, because it seems they are sometimes already
372 pre-defined.
373
374 28. Added support for (*UTF8) at the start of a pattern.
375
376 29. Arrange for flags added by the "release type" setting in CMake to be shown
377 in the configuration summary.
378
379
380 Version 7.8 05-Sep-08
381 ---------------------
382
383 1. Replaced UCP searching code with optimized version as implemented for Ad
384 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
385 stage table and inline lookup instead of a function, giving speed ups of 2
386 to 5 times on some simple patterns that I tested. Permission was given to
387 distribute the MultiStage2.py script that generates the tables (it's not in
388 the tarball, but is in the Subversion repository).
389
390 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
391 scripts.
392
393 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
394 a group with a zero qualifier. The result of the study could be incorrect,
395 or the function might crash, depending on the pattern.
396
397 4. Caseless matching was not working for non-ASCII characters in back
398 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
399 It now works when Unicode Property Support is available.
400
401 5. In pcretest, an escape such as \x{de} in the data was always generating
402 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
403 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
404 truncation.
405
406 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
407
408 7. Added two (int) casts to pcregrep when printing the difference of two
409 pointers, in case they are 64-bit values.
410
411 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
412 test 2 if it fails.
413
414 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
415 and a #define of that name to empty if it is not externally set. This is to
416 allow users of MSVC to set it if necessary.
417
418 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
419 the convenience functions in the pcre_get.c source file.
420
421 11. An option change at the start of a pattern that had top-level alternatives
422 could cause overwriting and/or a crash. This command provoked a crash in
423 some environments:
424
425 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
426
427 This potential security problem was recorded as CVE-2008-2371.
428
429 12. For a pattern where the match had to start at the beginning or immediately
430 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
431 pcre_dfa_exec() could read past the end of the passed subject if there was
432 no match. To help with detecting such bugs (e.g. with valgrind), I modified
433 pcretest so that it places the subject at the end of its malloc-ed buffer.
434
435 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
436 exec() might read past the end of the data buffer in UTF-8 mode.
437
438 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
439 the data contained the byte 0x85 as part of a UTF-8 character within its
440 first line. This applied both to normal and DFA matching.
441
442 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
443 /^[^d]*?$/8 failed to match "abc".
444
445 16. Added a missing copyright notice to pcrecpp_internal.h.
446
447 17. Make it more clear in the documentation that values returned from
448 pcre_exec() in ovector are byte offsets, not character counts.
449
450 18. Tidied a few places to stop certain compilers from issuing warnings.
451
452 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
453 supplied by Stefan Weber. I made a further small update for 7.8 because
454 there is a change of source arrangements: the pcre_searchfuncs.c module is
455 replaced by pcre_ucd.c.
456
457
458 Version 7.7 07-May-08
459 ---------------------
460
461 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
462 a string to a long long, pretend we don't even have a long long." This is
463 done by checking for the strtoq, strtoll, and _strtoi64 functions.
464
465 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
466 pre-7.6 versions, which defined a global no_arg variable instead of putting
467 it in the RE class. (See also #8 below.)
468
469 3. Remove a line of dead code, identified by coverity and reported by Nuno
470 Lopes.
471
472 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
473
474 (1) The include/exclude patterns were being applied to the whole pathnames
475 of files, instead of just to the final components.
476
477 (2) If there was more than one level of directory, the subdirectories were
478 skipped unless they satisfied the include/exclude conditions. This is
479 inconsistent with GNU grep (and could even be seen as contrary to the
480 pcregrep specification - which I improved to make it absolutely clear).
481 The action now is always to scan all levels of directory, and just
482 apply the include/exclude patterns to regular files.
483
484 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
485 --exclude_dir in the tests to avoid scanning .svn directories.
486
487 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
488 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
489 doesn't support NULs in patterns.
490
491 7. Added some missing "const"s to declarations of static tables in
492 pcre_compile.c and pcre_dfa_exec.c.
493
494 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
495 caused by fix #2 above. (Subsequently also a second patch to fix the
496 first patch. And a third patch - this was a messy problem.)
497
498 9. Applied Craig's patch to remove the use of push_back().
499
500 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
501 matching function regexec().
502
503 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
504 which, however, unlike Perl's \g{...}, are subroutine calls, not back
505 references. PCRE supports relative numbers with this syntax (I don't think
506 Oniguruma does).
507
508 12. Previously, a group with a zero repeat such as (...){0} was completely
509 omitted from the compiled regex. However, this means that if the group
510 was called as a subroutine from elsewhere in the pattern, things went wrong
511 (an internal error was given). Such groups are now left in the compiled
512 pattern, with a new opcode that causes them to be skipped at execution
513 time.
514
515 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
516 to the way PCRE behaves:
517
518 (a) A lone ] character is dis-allowed (Perl treats it as data).
519
520 (b) A back reference to an unmatched subpattern matches an empty string
521 (Perl fails the current match path).
522
523 (c) A data ] in a character class must be notated as \] because if the
524 first data character in a class is ], it defines an empty class. (In
525 Perl it is not possible to have an empty class.) The empty class []
526 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
527 The negative empty class [^] matches any one character, independently
528 of the DOTALL setting.
529
530 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
531 non-existent subpattern following a character class starting with ']' and
532 containing () gave an internal compiling error instead of "reference to
533 non-existent subpattern". Fortunately, when the pattern did exist, the
534 compiled code was correct. (When scanning forwards to check for the
535 existencd of the subpattern, it was treating the data ']' as terminating
536 the class, so got the count wrong. When actually compiling, the reference
537 was subsequently set up correctly.)
538
539 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
540 it was being rejected as not supported by pcre_dfa_exec(), even though
541 other assertions are supported. I have made pcre_dfa_exec() support
542 (*FAIL).
543
544 16. The implementation of 13c above involved the invention of a new opcode,
545 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
546 cannot be changed at match time, I realized I could make a small
547 improvement to matching performance by compiling OP_ALLANY instead of
548 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
549 on the OP_ANY path.
550
551 17. Compiling pcretest on Windows with readline support failed without the
552 following two fixes: (1) Make the unistd.h include conditional on
553 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
554
555 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
556 ncurses library to be included for pcretest when ReadLine support is
557 requested, but also to allow for it to be overridden. This patch came from
558 Daniel Bergström.
559
560 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
561 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
562 any errors with the current Unicode tables. Thanks to Peter Kankowski for
563 spotting this.
564
565
566 Version 7.6 28-Jan-08
567 ---------------------
568
569 1. A character class containing a very large number of characters with
570 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
571 overflow.
572
573 2. Patch to cut out the "long long" test in pcrecpp_unittest when
574 HAVE_LONG_LONG is not defined.
575
576 3. Applied Christian Ehrlicher's patch to update the CMake build files to
577 bring them up to date and include new features. This patch includes:
578
579 - Fixed PH's badly added libz and libbz2 support.
580 - Fixed a problem with static linking.
581 - Added pcredemo. [But later removed - see 7 below.]
582 - Fixed dftables problem and added an option.
583 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
584 HAVE_LONG_LONG.
585 - Added readline support for pcretest.
586 - Added an listing of the option settings after cmake has run.
587
588 4. A user submitted a patch to Makefile that makes it easy to create
589 "pcre.dll" under mingw when using Configure/Make. I added stuff to
590 Makefile.am that cause it to include this special target, without
591 affecting anything else. Note that the same mingw target plus all
592 the other distribution libraries and programs are now supported
593 when configuring with CMake (see 6 below) instead of with
594 Configure/Make.
595
596 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
597 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
598 exported in the Windows port". It has not yet been confirmed that the patch
599 solves the problem, but it does no harm.
600
601 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
602 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
603 with CMake, and also correct the comment about stack recursion.
604
605 7. Remove the automatic building of pcredemo from the ./configure system and
606 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
607 of a program that users should build themselves after PCRE is installed, so
608 building it automatically is not really right. What is more, it gave
609 trouble in some build environments.
610
611 8. Further tidies to CMakeLists.txt from Sheri and Christian.
612
613
614 Version 7.5 10-Jan-08
615 ---------------------
616
617 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
618 values in parens when parsing an RE using the C++ wrapper."
619
620 2. Negative specials like \S did not work in character classes in UTF-8 mode.
621 Characters greater than 255 were excluded from the class instead of being
622 included.
623
624 3. The same bug as (2) above applied to negated POSIX classes such as
625 [:^space:].
626
627 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
628 defined or documented. It seems to have been a typo for PCRE_STATIC, so
629 I have changed it.
630
631 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
632 first named subpattern) and a construct such as (?&a) would reference the
633 first named subpattern whose name started with "a" (in other words, the
634 length check was missing). Both these problems are fixed. "Subpattern name
635 expected" is now given for (?&) (a zero-length name), and this patch also
636 makes it give the same error for \k'' (previously it complained that that
637 was a reference to a non-existent subpattern).
638
639 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
640 this is right because (?- can be followed by option settings as well as by
641 digits. I have, however, made the messages clearer.
642
643 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
644 than the number used in the conditional) now cause a compile-time error.
645 This is actually not compatible with Perl, which accepts such patterns, but
646 treats the conditional as always being FALSE (as PCRE used to), but it
647 seems to me that giving a diagnostic is better.
648
649 8. Change "alphameric" to the more common word "alphanumeric" in comments
650 and messages.
651
652 9. Fix two occurrences of "backslash" in comments that should have been
653 "backspace".
654
655 10. Remove two redundant lines of code that can never be obeyed (their function
656 was moved elsewhere).
657
658 11. The program that makes PCRE's Unicode character property table had a bug
659 which caused it to generate incorrect table entries for sequences of
660 characters that have the same character type, but are in different scripts.
661 It amalgamated them into a single range, with the script of the first of
662 them. In other words, some characters were in the wrong script. There were
663 thirteen such cases, affecting characters in the following ranges:
664
665 U+002b0 - U+002c1
666 U+0060c - U+0060d
667 U+0061e - U+00612
668 U+0064b - U+0065e
669 U+0074d - U+0076d
670 U+01800 - U+01805
671 U+01d00 - U+01d77
672 U+01d9b - U+01dbf
673 U+0200b - U+0200f
674 U+030fc - U+030fe
675 U+03260 - U+0327f
676 U+0fb46 - U+0fbb1
677 U+10450 - U+1049d
678
679 12. The -o option (show only the matching part of a line) for pcregrep was not
680 compatible with GNU grep in that, if there was more than one match in a
681 line, it showed only the first of them. It now behaves in the same way as
682 GNU grep.
683
684 13. If the -o and -v options were combined for pcregrep, it printed a blank
685 line for every non-matching line. GNU grep prints nothing, and pcregrep now
686 does the same. The return code can be used to tell if there were any
687 non-matching lines.
688
689 14. Added --file-offsets and --line-offsets to pcregrep.
690
691 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
692 infinitely looping recursion. The bug was that positive lookaheads were not
693 being skipped when checking for a possible empty match (negative lookaheads
694 and both kinds of lookbehind were skipped).
695
696 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
697 inclusion of <windows.h> to before rather than after the definition of
698 INVALID_FILE_ATTRIBUTES (patch from David Byron).
699
700 17. Specifying a possessive quantifier with a specific limit for a Unicode
701 character property caused pcre_compile() to compile bad code, which led at
702 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
703 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
704 caused the error; without that there was no problem.
705
706 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
707
708 19. Added --enable-pcretest-libreadline.
709
710 20. In pcrecpp.cc, the variable 'count' was incremented twice in
711 RE::GlobalReplace(). As a result, the number of replacements returned was
712 double what it should be. I removed one of the increments, but Craig sent a
713 later patch that removed the other one (the right fix) and added unit tests
714 that check the return values (which was not done before).
715
716 21. Several CMake things:
717
718 (1) Arranged that, when cmake is used on Unix, the libraries end up with
719 the names libpcre and libpcreposix, not just pcre and pcreposix.
720
721 (2) The above change means that pcretest and pcregrep are now correctly
722 linked with the newly-built libraries, not previously installed ones.
723
724 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
725
726 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
727 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
728 UTF-8 newline character). The key issue is that the pattern starts .*;
729 this means that the match must be either at the beginning, or after a
730 newline. The bug was in the code for advancing after a failed match and
731 checking that the new position followed a newline. It was not taking
732 account of UTF-8 characters correctly.
733
734 23. PCRE was behaving differently from Perl in the way it recognized POSIX
735 character classes. PCRE was not treating the sequence [:...:] as a
736 character class unless the ... were all letters. Perl, however, seems to
737 allow any characters between [: and :], though of course it rejects as
738 unknown any "names" that contain non-letters, because all the known class
739 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
740 for example, whereas PCRE did not - it did not recognize a POSIX character
741 class. This seemed a bit dangerous, so the code has been changed to be
742 closer to Perl. The behaviour is not identical to Perl, because PCRE will
743 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
744 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
745 Perl does, and where it didn't before.
746
747 24. Rewrite so as to remove the single use of %n from pcregrep because in some
748 Windows environments %n is disabled by default.
749
750
751 Version 7.4 21-Sep-07
752 ---------------------
753
754 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
755 means that a class such as [\s] counted as "explicit reference to CR or
756 LF". That isn't really right - the whole point of the change was to try to
757 help when there was an actual mention of one of the two characters. So now
758 the change happens only if \r or \n (or a literal CR or LF) character is
759 encountered.
760
761 2. The 32-bit options word was also used for 6 internal flags, but the numbers
762 of both had grown to the point where there were only 3 bits left.
763 Fortunately, there was spare space in the data structure, and so I have
764 moved the internal flags into a new 16-bit field to free up more option
765 bits.
766
767 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
768 but did not set the internal JCHANGED flag - either of these is enough to
769 control the way the "get" function works - but the PCRE_INFO_JCHANGED
770 facility is supposed to tell if (?J) was ever used, so now (?J) at the
771 start sets both bits.
772
773 4. Added options (at build time, compile time, exec time) to change \R from
774 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
775
776 5. doc/pcresyntax.html was missing from the distribution.
777
778 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
779 compatibility, even though it is no longer used.
780
781 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
782 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
783 windows.h file is present (where different names are used). [This was
784 reversed later after testing - see 16 below.]
785
786 8. Changed all #include <config.h> to #include "config.h". There were also
787 some further <pcre.h> cases that I changed to "pcre.h".
788
789 9. When pcregrep was used with the --colour option, it missed the line ending
790 sequence off the lines that it output.
791
792 10. It was pointed out to me that arrays of string pointers cause lots of
793 relocations when a shared library is dynamically loaded. A technique of
794 using a single long string with a table of offsets can drastically reduce
795 these. I have refactored PCRE in four places to do this. The result is
796 dramatic:
797
798 Originally: 290
799 After changing UCP table: 187
800 After changing error message table: 43
801 After changing table of "verbs" 36
802 After changing table of Posix names 22
803
804 Thanks to the folks working on Gregex for glib for this insight.
805
806 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
807 unicode-properties was also set.
808
809 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
810
811 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
812 checked only for CRLF.
813
814 14. Added casts to pcretest.c to avoid compiler warnings.
815
816 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
817
818 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
819 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
820 entirely. This removes changes made in 7 above.
821
822 17. The CMake files have been updated, and there is now more information about
823 building with CMake in the NON-UNIX-USE document.
824
825
826 Version 7.3 28-Aug-07
827 ---------------------
828
829 1. In the rejigging of the build system that eventually resulted in 7.1, the
830 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
831 brackets there is not right, since it causes compilers to look for an
832 installed pcre.h, not the version that is in the source that is being
833 compiled (which of course may be different). I have changed it back to:
834
835 #include "pcre.h"
836
837 I have a vague recollection that the change was concerned with compiling in
838 different directories, but in the new build system, that is taken care of
839 by the VPATH setting the Makefile.
840
841 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
842 when the subject happened to end in the byte 0x85 (e.g. if the last
843 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
844 characters but of course it shouldn't be taken as a newline when it is part
845 of another character. The bug was that, for an unlimited repeat of . in
846 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
847 characters when looking for a newline.
848
849 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
850
851 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
852 in debug output.
853
854 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
855 long printing in the pcrecpp unittest when running under MinGW.
856
857 6. ESC_K was left out of the EBCDIC table.
858
859 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
860 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
861 limit also applies to "virtual nesting" when a pattern is recursive, and in
862 this case 1000 isn't so big. I have been able to remove this limit at the
863 expense of backing off one optimization in certain circumstances. Normally,
864 when pcre_exec() would call its internal match() function recursively and
865 immediately return the result unconditionally, it uses a "tail recursion"
866 feature to save stack. However, when a subpattern that can match an empty
867 string has an unlimited repetition quantifier, it no longer makes this
868 optimization. That gives it a stack frame in which to save the data for
869 checking that an empty string has been matched. Previously this was taken
870 from the 1000-entry workspace that had been reserved. So now there is no
871 explicit limit, but more stack is used.
872
873 8. Applied Daniel's patches to solve problems with the import/export magic
874 syntax that is required for Windows, and which was going wrong for the
875 pcreposix and pcrecpp parts of the library. These were overlooked when this
876 problem was solved for the main library.
877
878 9. There were some crude static tests to avoid integer overflow when computing
879 the size of patterns that contain repeated groups with explicit upper
880 limits. As the maximum quantifier is 65535, the maximum group length was
881 set at 30,000 so that the product of these two numbers did not overflow a
882 32-bit integer. However, it turns out that people want to use groups that
883 are longer than 30,000 bytes (though not repeat them that many times).
884 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
885 made it possible to implement the integer overflow checks in a much more
886 dynamic way, which I have now done. The artificial limitation on group
887 length has been removed - we now have only the limit on the total length of
888 the compiled pattern, which depends on the LINK_SIZE setting.
889
890 10. Fixed a bug in the documentation for get/copy named substring when
891 duplicate names are permitted. If none of the named substrings are set, the
892 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
893 empty string.
894
895 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
896 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
897 because the ] is interpreted as the first data character and the
898 terminating ] is not found. PCRE has been made compatible with Perl in this
899 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
900 cause memory overwriting.
901
902 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
903 string has been matched (to stop an infinite loop). It was not recognizing
904 a conditional subpattern that could match an empty string if that
905 subpattern was within another subpattern. For example, it looped when
906 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
907 condition was not nested. This bug has been fixed.
908
909 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
910 past the start of the subject in the presence of bytes with the top bit
911 set, for example "\x8aBCD".
912
913 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
914 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
915
916 14. Optimized (?!) to (*FAIL).
917
918 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
919 This restricts code points to be within the range 0 to 0x10FFFF, excluding
920 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
921 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
922 does: it's just the validity check that is more restrictive.
923
924 16. Inserted checks for integer overflows during escape sequence (backslash)
925 processing, and also fixed erroneous offset values for syntax errors during
926 backslash processing.
927
928 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
929 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
930
931 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
932 caused an overrun.
933
934 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
935 something other than just ASCII characters) inside a group that had an
936 unlimited repeat caused a loop at compile time (while checking to see
937 whether the group could match an empty string).
938
939 20. Debugging a pattern containing \p or \P could cause a crash. For example,
940 [\P{Any}] did so. (Error in the code for printing property names.)
941
942 21. An orphan \E inside a character class could cause a crash.
943
944 22. A repeated capturing bracket such as (A)? could cause a wild memory
945 reference during compilation.
946
947 23. There are several functions in pcre_compile() that scan along a compiled
948 expression for various reasons (e.g. to see if it's fixed length for look
949 behind). There were bugs in these functions when a repeated \p or \P was
950 present in the pattern. These operators have additional parameters compared
951 with \d, etc, and these were not being taken into account when moving along
952 the compiled data. Specifically:
953
954 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
955 length.
956
957 (b) An item such as \pL+ within a repeated group could cause crashes or
958 loops.
959
960 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
961 "reference to non-existent subpattern" error.
962
963 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
964
965 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
966 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
967
968 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
969
970 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
971 character were causing crashes (broken optimization).
972
973 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
974 \p or \P) caused a compile-time loop.
975
976 28. More problems have arisen in unanchored patterns when CRLF is a valid line
977 break. For example, the unstudied pattern [\r\n]A does not match the string
978 "\r\nA" because change 7.0/46 below moves the current point on by two
979 characters after failing to match at the start. However, the pattern \nA
980 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
981 the same is true. There doesn't seem any very clean way out of this, but
982 what I have chosen to do makes the common cases work: PCRE now takes note
983 of whether there can be an explicit match for \r or \n anywhere in the
984 pattern, and if so, 7.0/46 no longer applies. As part of this change,
985 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
986 pattern has explicit CR or LF references.
987
988 29. Added (*CR) etc for changing newline setting at start of pattern.
989
990
991 Version 7.2 19-Jun-07
992 ---------------------
993
994 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
995 which is apparently normally available under Windows.
996
997 2. Re-jig the pcregrep tests with different newline settings in an attempt
998 to make them independent of the local environment's newline setting.
999
1000 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1001
1002 4. Some of the "internals" tests were previously cut out when the link size
1003 was not 2, because the output contained actual offsets. The recent new
1004 "Z" feature of pcretest means that these can be cut out, making the tests
1005 usable with all link sizes.
1006
1007 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1008 stack recursion. This gives a massive performance boost under BSD, but just
1009 a small improvement under Linux. However, it saves one field in the frame
1010 in all cases.
1011
1012 6. Added more features from the forthcoming Perl 5.10:
1013
1014 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1015 recursion call. It refers to the nth most recently opened parentheses.
1016
1017 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1018 to be opened parentheses.
1019
1020 (c) Conditions that refer to capturing parentheses can be specified
1021 relatively, for example, (?(-2)... or (?(+3)...
1022
1023 (d) \K resets the start of the current match so that everything before
1024 is not part of it.
1025
1026 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1027
1028 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1029 reference syntax.
1030
1031 (g) (?| introduces a group in which the numbering of parentheses in each
1032 alternative starts with the same number.
1033
1034 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1035
1036 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1037 PCRE_INFO_JCHANGED.
1038
1039 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1040 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1041 for detecting groups that can match an empty string.
1042
1043 9. A pattern with a very large number of alternatives (more than several
1044 hundred) was running out of internal workspace during the pre-compile
1045 phase, where pcre_compile() figures out how much memory will be needed. A
1046 bit of new cunning has reduced the workspace needed for groups with
1047 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1048 workspace instead of running out of the 4096 that are available.
1049
1050 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1051
1052 11. Applied patch from Google to remove an optimization that didn't quite work.
1053 The report of the bug said:
1054
1055 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1056 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1057 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1058
1059 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1060 it matched the wrong number of bytes.
1061
1062
1063 Version 7.1 24-Apr-07
1064 ---------------------
1065
1066 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1067 that is more "standard", making use of automake and other Autotools. There
1068 is some re-arrangement of the files and adjustment of comments consequent
1069 on this.
1070
1071 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1072 for recursive directory scanning broke on some systems because the files
1073 are not scanned in any specific order and on different systems the order
1074 was different. A call to "sort" has been inserted into RunGrepTest for the
1075 approprate test as a short-term fix. In the longer term there may be an
1076 alternative.
1077
1078 3. I had an email from Eric Raymond about problems translating some of PCRE's
1079 man pages to HTML (despite the fact that I distribute HTML pages, some
1080 people do their own conversions for various reasons). The problems
1081 concerned the use of low-level troff macros .br and .in. I have therefore
1082 removed all such uses from the man pages (some were redundant, some could
1083 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1084 HTML has been updated to handle .nf/.fi and to complain if it encounters
1085 .br or .in.
1086
1087 4. Updated comments in configure.ac that get placed in config.h.in and also
1088 arranged for config.h to be included in the distribution, with the name
1089 config.h.generic, for the benefit of those who have to compile without
1090 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1091
1092 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1093 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1094 makevp.bat for the current PCRE, using the additional files
1095 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1096
1097 6. A Windows user reported a minor discrepancy with test 2, which turned out
1098 to be caused by a trailing space on an input line that had got lost in his
1099 copy. The trailing space was an accident, so I've just removed it.
1100
1101 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1102 that is needed.
1103
1104 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1105 as "const" (a) because they are and (b) because it helps the PHP
1106 maintainers who have recently made a script to detect big data structures
1107 in the php code that should be moved to the .rodata section. I remembered
1108 to update Builducptable as well, so it won't revert if ucptable.h is ever
1109 re-created.
1110
1111 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1112 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1113 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1114 support is not required. This saves 1.5-2K of code, which is important in
1115 some applications.
1116
1117 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1118 so as not to refer to the tables, even though these functions will never be
1119 called when UTF-8 support is disabled. Otherwise there are problems with a
1120 shared library.
1121
1122 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1123
1124 (a) It was defining its arguments as char * instead of void *.
1125
1126 (b) It was assuming that all moves were upwards in memory; this was true
1127 a long time ago when I wrote it, but is no longer the case.
1128
1129 The emulated memove() is provided for those environments that have neither
1130 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1131 is clearly not the case, as these two bugs were recently reported.
1132
1133 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1134 and Detrail to create the HTML documentation, the .txt form of the man
1135 pages, and it removes trailing spaces from listed files. It also creates
1136 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1137 case, it wraps all the #defines with #ifndefs. This script should be run
1138 before "make dist".
1139
1140 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1141 with Unicode property support.
1142
1143 (a) For a maximizing quantifier, if the two different cases of the
1144 character were of different lengths in their UTF-8 codings (there are
1145 some cases like this - I found 11), and the matching function had to
1146 back up over a mixture of the two cases, it incorrectly assumed they
1147 were both the same length.
1148
1149 (b) When PCRE was configured to use the heap rather than the stack for
1150 recursion during matching, it was not correctly preserving the data for
1151 the other case of a UTF-8 character when checking ahead for a match
1152 while processing a minimizing repeat. If the check also involved
1153 matching a wide character, but failed, corruption could cause an
1154 erroneous result when trying to check for a repeat of the original
1155 character.
1156
1157 13. Some tidying changes to the testing mechanism:
1158
1159 (a) The RunTest script now detects the internal link size and whether there
1160 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1161 values substituted by "configure". (The RunGrepTest script already did
1162 this for UTF-8.) The configure.ac script no longer substitutes the
1163 relevant variables.
1164
1165 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1166 with length and offset values. This means that the output is different
1167 for different internal link sizes. Test 2 is skipped for link sizes
1168 other than 2 because of this, bypassing the problem. Unfortunately,
1169 there was also a test in test 3 (the locale tests) that used /B and
1170 failed for link sizes other than 2. Rather than cut the whole test out,
1171 I have added a new /Z option to pcretest that replaces the length and
1172 offset values with spaces. This is now used to make test 3 independent
1173 of link size. (Test 2 will be tidied up later.)
1174
1175 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1176 segmentation fault instead of returning the appropriate error message.
1177
1178 15. In multiline mode when the newline sequence was set to "any", the pattern
1179 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1180 This doesn't seem right; it now treats the CRLF combination as the line
1181 ending, and so does not match in that case. It's only a pattern such as ^$
1182 that would hit this one: something like ^ABC$ would have failed after \r
1183 and then tried again after \r\n.
1184
1185 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1186 in an attempt to make files that differ only in their line terminators
1187 compare equal. This works on Linux.
1188
1189 17. Under certain error circumstances pcregrep might try to free random memory
1190 as it exited. This is now fixed, thanks to valgrind.
1191
1192 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1193 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1194 was because its rules for how to advance for /g after matching an empty
1195 string at the end of a line did not allow for this case. They now check for
1196 it specially.
1197
1198 20. pcretest is supposed to handle patterns and data of any length, by
1199 extending its buffers when necessary. It was getting this wrong when the
1200 buffer for a data line had to be extended.
1201
1202 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1203 CRLF as a newline sequence.
1204
1205 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1206 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1207 I have nevertheless tidied it up.
1208
1209 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1210
1211 24. Added a man page for pcre-config.
1212
1213
1214 Version 7.0 19-Dec-06
1215 ---------------------
1216
1217 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1218 moving to gcc 4.1.1.
1219
1220 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1221 sys/time.h, which is documented as needed for this function. It doesn't
1222 seem to matter on Linux, but it showed up on some releases of OS X.
1223
1224 3. It seems that there are systems where bytes whose values are greater than
1225 127 match isprint() in the "C" locale. The "C" locale should be the
1226 default when a C program starts up. In most systems, only ASCII printing
1227 characters match isprint(). This difference caused the output from pcretest
1228 to vary, making some of the tests fail. I have changed pcretest so that:
1229
1230 (a) When it is outputting text in the compiled version of a pattern, bytes
1231 other than 32-126 are always shown as hex escapes.
1232
1233 (b) When it is outputting text that is a matched part of a subject string,
1234 it does the same, unless a different locale has been set for the match
1235 (using the /L modifier). In this case, it uses isprint() to decide.
1236
1237 4. Fixed a major bug that caused incorrect computation of the amount of memory
1238 required for a compiled pattern when options that changed within the
1239 pattern affected the logic of the preliminary scan that determines the
1240 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1241 that the computed length was too small. The symptoms of this bug were
1242 either the PCRE error "internal error: code overflow" from pcre_compile(),
1243 or a glibc crash with a message such as "pcretest: free(): invalid next
1244 size (fast)". Examples of patterns that provoked this bug (shown in
1245 pcretest format) are:
1246
1247 /(?-x: )/x
1248 /(?x)(?-x: \s*#\s*)/
1249 /((?i)[\x{c0}])/8
1250 /(?i:[\x{c0}])/8
1251
1252 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1253 is now done differently.
1254
1255 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1256 wrapper classes; (b) implement a new function in the C++ scanner that is
1257 more efficient than the old way of doing things because it avoids levels of
1258 recursion in the regex matching; (c) add a paragraph to the documentation
1259 for the FullMatch() function.
1260
1261 6. The escape sequence \n was being treated as whatever was defined as
1262 "newline". Not only was this contrary to the documentation, which states
1263 that \n is character 10 (hex 0A), but it also went horribly wrong when
1264 "newline" was defined as CRLF. This has been fixed.
1265
1266 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1267 was being set to -1 for the "end of line" case (supposedly a value that no
1268 character can have). Though this value is never used (the check for end of
1269 line is "zero bytes in current character"), it caused compiler complaints.
1270 I've changed it to 0xffffffff.
1271
1272 8. In pcre_version.c, the version string was being built by a sequence of
1273 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1274 string (as it is for production releases) called a macro with an empty
1275 argument. The C standard says the result of this is undefined. The gcc
1276 compiler treats it as an empty string (which was what was wanted) but it is
1277 reported that Visual C gives an error. The source has been hacked around to
1278 avoid this problem.
1279
1280 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1281 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1282 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1283 of them did).
1284
1285 10. Originally, pcretest opened its input and output without "b"; then I was
1286 told that "b" was needed in some environments, so it was added for release
1287 5.0 to both the input and output. (It makes no difference on Unix-like
1288 systems.) Later I was told that it is wrong for the input on Windows. I've
1289 now abstracted the modes into two macros, to make it easier to fiddle with
1290 them, and removed "b" from the input mode under Windows.
1291
1292 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1293
1294 12. Added -help and --help to pcretest as an official way of being reminded
1295 of the options.
1296
1297 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1298 and pcrecpp.cc because they annoy compilers at high warning levels.
1299
1300 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1301
1302 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1303 scripts are not C programs :-) and which was not noticed because it works
1304 on Linux.
1305
1306 16. pcretest is supposed to handle any length of pattern and data line (as one
1307 line or as a continued sequence of lines) by extending its input buffer if
1308 necessary. This feature was broken for very long pattern lines, leading to
1309 a string of junk being passed to pcre_compile() if the pattern was longer
1310 than about 50K.
1311
1312 17. I have done a major re-factoring of the way pcre_compile() computes the
1313 amount of memory needed for a compiled pattern. Previously, there was code
1314 that made a preliminary scan of the pattern in order to do this. That was
1315 OK when PCRE was new, but as the facilities have expanded, it has become
1316 harder and harder to keep it in step with the real compile phase, and there
1317 have been a number of bugs (see for example, 4 above). I have now found a
1318 cunning way of running the real compile function in a "fake" mode that
1319 enables it to compute how much memory it would need, while actually only
1320 ever using a few hundred bytes of working memory and without too many
1321 tests of the mode. This should make future maintenance and development
1322 easier. A side effect of this work is that the limit of 200 on the nesting
1323 depth of parentheses has been removed (though this was never a serious
1324 limitation, I suspect). However, there is a downside: pcre_compile() now
1325 runs more slowly than before (30% or more, depending on the pattern). I
1326 hope this isn't a big issue. There is no effect on runtime performance.
1327
1328 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1329 newline (only possible for the last line of a file) and it was a
1330 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1331
1332 19. Added additional timing features to pcretest. (1) The -tm option now times
1333 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1334 separate command line item, by a number that specifies the number of
1335 repeats to use when timing. The default is 50000; this gives better
1336 precision, but takes uncomfortably long for very large patterns.
1337
1338 20. Extended pcre_study() to be more clever in cases where a branch of a
1339 subpattern has no definite first character. For example, (a*|b*)[cd] would
1340 previously give no result from pcre_study(). Now it recognizes that the
1341 first character must be a, b, c, or d.
1342
1343 21. There was an incorrect error "recursive call could loop indefinitely" if
1344 a subpattern (or the entire pattern) that was being tested for matching an
1345 empty string contained only one non-empty item after a nested subpattern.
1346 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1347 incorrectly, because the \d was being skipped in the check.
1348
1349 22. The pcretest program now has a new pattern option /B and a command line
1350 option -b, which is equivalent to adding /B to every pattern. This causes
1351 it to show the compiled bytecode, without the additional information that
1352 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1353 is the same as /B/I).
1354
1355 23. A new optimization is now able automatically to treat some sequences such
1356 as a*b as a*+b. More specifically, if something simple (such as a character
1357 or a simple class like \d) has an unlimited quantifier, and is followed by
1358 something that cannot possibly match the quantified thing, the quantifier
1359 is automatically "possessified".
1360
1361 24. A recursive reference to a subpattern whose number was greater than 39
1362 went wrong under certain circumstances in UTF-8 mode. This bug could also
1363 have affected the operation of pcre_study().
1364
1365 25. Realized that a little bit of performance could be had by replacing
1366 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1367
1368 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1369
1370 27. Possessive quantifiers such as a++ were previously implemented by turning
1371 them into atomic groups such as ($>a+). Now they have their own opcodes,
1372 which improves performance. This includes the automatically created ones
1373 from 23 above.
1374
1375 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1376 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1377 the first matched character to be a colon. This applied both to named and
1378 numbered groups.
1379
1380 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1381
1382 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1383 building PCRE on Windows easier, so I have included it in the distribution.
1384
1385 31. There is now a check in pcretest against a ridiculously large number being
1386 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1387 loop, the loop is abandoned.
1388
1389 32. Forward references to subpatterns in conditions such as (?(2)...) where
1390 subpattern 2 is defined later cause pcre_compile() to search forwards in
1391 the pattern for the relevant set of parentheses. This search went wrong
1392 when there were unescaped parentheses in a character class, parentheses
1393 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1394
1395 33. "Subroutine" calls and backreferences were previously restricted to
1396 referencing subpatterns earlier in the regex. This restriction has now
1397 been removed.
1398
1399 34. Added a number of extra features that are going to be in Perl 5.10. On the
1400 whole, these are just syntactic alternatives for features that PCRE had
1401 previously implemented using the Python syntax or my own invention. The
1402 other formats are all retained for compatibility.
1403
1404 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1405 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1406 also .NET compatible.
1407
1408 (b) A recursion or subroutine call to a named group can now be defined as
1409 (?&name) as well as (?P>name).
1410
1411 (c) A backreference to a named group can now be defined as \k<name> or
1412 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1413 5.10, are also .NET compatible.
1414
1415 (d) A conditional reference to a named group can now use the syntax
1416 (?(<name>) or (?('name') as well as (?(name).
1417
1418 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1419 groups (named and numbered) that are never evaluated inline, but can be
1420 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1421 is always false. There may be only one alternative in such a group.
1422
1423 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1424 as the simple (?(R). The condition is true only if the most recent
1425 recursion is that of the given number or name. It does not search out
1426 through the entire recursion stack.
1427
1428 (g) The escape \gN or \g{N} has been added, where N is a positive or
1429 negative number, specifying an absolute or relative reference.
1430
1431 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1432 some "unreachable code" warnings.
1433
1434 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1435 things, this adds five new scripts.
1436
1437 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1438 There were also incompatibilities regarding the handling of \Q..\E inside
1439 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1440 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1441
1442 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1443 matches an empty string, and forcibly breaks the loop. There were bugs in
1444 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1445 against aaaa the result was just "a" rather than "aaaa", for example. Two
1446 separate and independent bugs (that affected different cases) have been
1447 fixed.
1448
1449 39. Refactored the code to abolish the use of different opcodes for small
1450 capturing bracket numbers. This is a tidy that I avoided doing when I
1451 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1452 The new approach is not only tidier, it makes it possible to reduce the
1453 memory needed to fix the previous bug (38).
1454
1455 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1456 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1457 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1458 mode.
1459
1460 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1461 report.
1462
1463 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1464 copy construction and assignment in the C++ wrapper.
1465
1466 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1467 couple of bugs that could have given wrong results in the "--newline=crlf"
1468 case.
1469
1470 44. Added a number of casts and did some reorganization of signed/unsigned int
1471 variables following suggestions from Dair Grant. Also renamed the variable
1472 "this" as "item" because it is a C++ keyword.
1473
1474 45. Arranged for dftables to add
1475
1476 #include "pcre_internal.h"
1477
1478 to pcre_chartables.c because without it, gcc 4.x may remove the array
1479 definition from the final binary if PCRE is built into a static library and
1480 dead code stripping is activated.
1481
1482 46. For an unanchored pattern, if a match attempt fails at the start of a
1483 newline sequence, and the newline setting is CRLF or ANY, and the next two
1484 characters are CRLF, advance by two characters instead of one.
1485
1486
1487 Version 6.7 04-Jul-06
1488 ---------------------
1489
1490 1. In order to handle tests when input lines are enormously long, pcretest has
1491 been re-factored so that it automatically extends its buffers when
1492 necessary. The code is crude, but this _is_ just a test program. The
1493 default size has been increased from 32K to 50K.
1494
1495 2. The code in pcre_study() was using the value of the re argument before
1496 testing it for NULL. (Of course, in any sensible call of the function, it
1497 won't be NULL.)
1498
1499 3. The memmove() emulation function in pcre_internal.h, which is used on
1500 systems that lack both memmove() and bcopy() - that is, hardly ever -
1501 was missing a "static" storage class specifier.
1502
1503 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1504 containing an extended class (one that cannot be represented by a bitmap
1505 because it contains high-valued characters or Unicode property items, e.g.
1506 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1507 pattern, but PCRE should not loop if you do not (it no longer does).
1508 [Detail: two cases were found: (a) a repeated subpattern containing an
1509 extended class; (b) a recursive reference to a subpattern that followed a
1510 previous extended class. It wasn't skipping over the extended class
1511 correctly when UTF-8 mode was not set.]
1512
1513 5. A negated single-character class was not being recognized as fixed-length
1514 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1515 compile error "lookbehind assertion is not fixed length".
1516
1517 6. The RunPerlTest auxiliary script was showing an unexpected difference
1518 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1519 write a Perl script that can interpret lines of an input file either as
1520 byte characters or as UTF-8, which is what "perltest" was being required to
1521 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1522 can't do is switch easily at run time between having the "use utf8;" pragma
1523 or not. In the end, I fudged it by using the RunPerlTest script to insert
1524 "use utf8;" explicitly for the UTF-8 tests.
1525
1526 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1527 the end of the subject string, contrary to the documentation and to what
1528 Perl does. This was true of both matching functions. Now it matches only at
1529 the start of the subject and immediately after *internal* newlines.
1530
1531 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1532 a pointer to an int instead of a pointer to an unsigned long int. This
1533 caused problems on 64-bit systems.
1534
1535 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1536 instance of the 'standard' template library not being so standard".
1537
1538 10. There was no check on the number of named subpatterns nor the maximum
1539 length of a subpattern name. The product of these values is used to compute
1540 the size of the memory block for a compiled pattern. By supplying a very
1541 long subpattern name and a large number of named subpatterns, the size
1542 computation could be caused to overflow. This is now prevented by limiting
1543 the length of names to 32 characters, and the number of named subpatterns
1544 to 10,000.
1545
1546 11. Subpatterns that are repeated with specific counts have to be replicated in
1547 the compiled pattern. The size of memory for this was computed from the
1548 length of the subpattern and the repeat count. The latter is limited to
1549 65535, but there was no limit on the former, meaning that integer overflow
1550 could in principle occur. The compiled length of a repeated subpattern is
1551 now limited to 30,000 bytes in order to prevent this.
1552
1553 12. Added the optional facility to have named substrings with the same name.
1554
1555 13. Added the ability to use a named substring as a condition, using the
1556 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1557 are numbers (not recommended). Forward references are permitted.
1558
1559 14. Added forward references in named backreferences (if you see what I mean).
1560
1561 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1562 pattern could run off the end of the subject. For example, the pattern
1563 "(?s)(.{1,5})"8 did this with the subject "ab".
1564
1565 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1566 PCRE_CASELESS was set when matching characters that were quantified with ?
1567 or *.
1568
1569 17. A character class other than a single negated character that had a minimum
1570 but no maximum quantifier - for example [ab]{6,} - was not handled
1571 correctly by pce_dfa_exec(). It would match only one character.
1572
1573 18. A valid (though odd) pattern that looked like a POSIX character
1574 class but used an invalid character after [ (for example [[,abc,]]) caused
1575 pcre_compile() to give the error "Failed: internal error: code overflow" or
1576 in some cases to crash with a glibc free() error. This could even happen if
1577 the pattern terminated after [[ but there just happened to be a sequence of
1578 letters, a binary zero, and a closing ] in the memory that followed.
1579
1580 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1581 over the years. Originally (before any Unicode support), just the bottom 8
1582 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1583 output from "man perlunicode" includes this:
1584
1585 The regular expression compiler produces polymorphic opcodes. That
1586 is, the pattern adapts to the data and automatically switches to
1587 the Unicode character scheme when presented with Unicode data--or
1588 instead uses a traditional byte scheme when presented with byte
1589 data.
1590
1591 Sadly, a wide octal escape does not cause a switch, and in a string with
1592 no other multibyte characters, these octal escapes are treated as before.
1593 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1594 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1595 Unicode string.
1596
1597 I have not perpetrated such confusion in PCRE. Up till now, it took just
1598 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1599 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1600 translate to the appropriate multibyte character.
1601
1602 29. Applied some refactoring to reduce the number of warnings from Microsoft
1603 and Borland compilers. This has included removing the fudge introduced
1604 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1605 a warning about an unused variable.
1606
1607 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1608 characters since release 4.0, because Perl (from release 5.004) does not.
1609 [Or at least, is documented not to: some releases seem to be in conflict
1610 with the documentation.] However, when a pattern was studied with
1611 pcre_study() and all its branches started with \s, PCRE still included VT
1612 as a possible starting character. Of course, this did no harm; it just
1613 caused an unnecessary match attempt.
1614
1615 22. Removed a now-redundant internal flag bit that recorded the fact that case
1616 dependency changed within the pattern. This was once needed for "required
1617 byte" processing, but is no longer used. This recovers a now-scarce options
1618 bit. Also moved the least significant internal flag bit to the most-
1619 significant bit of the word, which was not previously used (hangover from
1620 the days when it was an int rather than a uint) to free up another bit for
1621 the future.
1622
1623 23. Added support for CRLF line endings as well as CR and LF. As well as the
1624 default being selectable at build time, it can now be changed at runtime
1625 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1626 specify that it is scanning data with non-default line endings.
1627
1628 24. Changed the definition of CXXLINK to make it agree with the definition of
1629 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1630
1631 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1632 recursions. This makes a big different to stack usage for some patterns.
1633
1634 26. If a subpattern containing a named recursion or subroutine reference such
1635 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1636 the space required for the compiled pattern went wrong and gave too small a
1637 value. Depending on the environment, this could lead to "Failed: internal
1638 error: code overflow at offset 49" or "glibc detected double free or
1639 corruption" errors.
1640
1641 27. Applied patches from Google (a) to support the new newline modes and (b) to
1642 advance over multibyte UTF-8 characters in GlobalReplace.
1643
1644 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1645 difference for some implementation of PCRE in some Windows version.
1646
1647 29. Added some extra testing facilities to pcretest:
1648
1649 \q<number> in a data line sets the "match limit" value
1650 \Q<number> in a data line sets the "match recursion limt" value
1651 -S <number> sets the stack size, where <number> is in megabytes
1652
1653 The -S option isn't available for Windows.
1654
1655
1656 Version 6.6 06-Feb-06
1657 ---------------------
1658
1659 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1660 in pcreposix.h. I have copied the definition from pcre.h.
1661
1662 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1663 because pcre.h is no longer a built file.
1664
1665 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1666 not normally included in the compiled code.
1667
1668
1669 Version 6.5 01-Feb-06
1670 ---------------------
1671
1672 1. When using the partial match feature with pcre_dfa_exec(), it was not
1673 anchoring the second and subsequent partial matches at the new starting
1674 point. This could lead to incorrect results. For example, with the pattern
1675 /1234/, partially matching against "123" and then "a4" gave a match.
1676
1677 2. Changes to pcregrep:
1678
1679 (a) All non-match returns from pcre_exec() were being treated as failures
1680 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1681 error message is output. Some extra information is given for the
1682 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1683 probably the only errors that are likely to be caused by users (by
1684 specifying a regex that has nested indefinite repeats, for instance).
1685 If there are more than 20 of these errors, pcregrep is abandoned.
1686
1687 (b) A binary zero was treated as data while matching, but terminated the
1688 output line if it was written out. This has been fixed: binary zeroes
1689 are now no different to any other data bytes.
1690
1691 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1692 used to set a locale for matching. The --locale=xxxx long option has
1693 been added (no short equivalent) to specify a locale explicitly on the
1694 pcregrep command, overriding the environment variables.
1695
1696 (d) When -B was used with -n, some line numbers in the output were one less
1697 than they should have been.
1698
1699 (e) Added the -o (--only-matching) option.
1700
1701 (f) If -A or -C was used with -c (count only), some lines of context were
1702 accidentally printed for the final match.
1703
1704 (g) Added the -H (--with-filename) option.
1705
1706 (h) The combination of options -rh failed to suppress file names for files
1707 that were found from directory arguments.
1708
1709 (i) Added the -D (--devices) and -d (--directories) options.
1710
1711 (j) Added the -F (--fixed-strings) option.
1712
1713 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1714
1715 (l) Added the --colo(u)r option.
1716
1717 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1718 is not present by default.
1719
1720 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1721 items such as (?R) or (?1), when the recursion could match a number of
1722 alternatives. If it matched one of the alternatives, but subsequently,
1723 outside the recursion, there was a failure, the code tried to back up into
1724 the recursion. However, because of the way PCRE is implemented, this is not
1725 possible, and the result was an incorrect result from the match.
1726
1727 In order to prevent this happening, the specification of recursion has
1728 been changed so that all such subpatterns are automatically treated as
1729 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1730
1731 4. I had overlooked the fact that, in some locales, there are characters for
1732 which isalpha() is true but neither isupper() nor islower() are true. In
1733 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1734 and ordfeminine) are like this. This affected the treatment of \w and \W
1735 when they appeared in character classes, but not when they appeared outside
1736 a character class. The bit map for "word" characters is now created
1737 separately from the results of isalnum() instead of just taking it from the
1738 upper, lower, and digit maps. (Plus the underscore character, of course.)
1739
1740 5. The above bug also affected the handling of POSIX character classes such as
1741 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1742 permanent tables. Instead, the bit maps for such a class were previously
1743 created as the appropriate unions of the upper, lower, and digit bitmaps.
1744 Now they are created by subtraction from the [[:word:]] class, which has
1745 its own bitmap.
1746
1747 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1748 It is created by subtracting the vertical space characters (\x09, \x0a,
1749 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1750 subtraction was done in the overall bitmap for a character class, meaning
1751 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1752 be recognized. This bug has been fixed.
1753
1754 7. Patches from the folks at Google:
1755
1756 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1757 real life, but is still worth protecting against".
1758
1759 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1760 regular expressions".
1761
1762 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1763 have it.
1764
1765 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1766 "configure" and the latter not, in order to fix a problem somebody had
1767 with compiling the Arg class on HP-UX.
1768
1769 (e) Improve the error-handling of the C++ wrapper a little bit.
1770
1771 (f) New tests for checking recursion limiting.
1772
1773 8. The pcre_memmove() function, which is used only if the environment does not
1774 have a standard memmove() function (and is therefore rarely compiled),
1775 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1776 returning a result (though PCRE never actually uses the result).
1777
1778 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1779 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1780 returned instead of calling malloc() with an overflowing number that would
1781 most likely cause subsequent chaos.
1782
1783 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1784
1785 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1786 with this option is matched, the nmatch and pmatch options of regexec() are
1787 ignored.
1788
1789 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1790 provided in case anyone wants to the the POSIX interface with UTF-8
1791 strings.
1792
1793 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1794 C++ linking (needed for some HP-UX environments).
1795
1796 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1797 (unused parameter) and in the pcre_printint() function (omitted "default"
1798 switch label when the default is to do nothing).
1799
1800 15. Added some code to make it possible, when PCRE is compiled as a C++
1801 library, to replace subject pointers for pcre_exec() with a smart pointer
1802 class, thus making it possible to process discontinuous strings.
1803
1804 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1805 much the same function. They were added by different people who were trying
1806 to make PCRE easy to compile on non-Unix systems. It has been suggested
1807 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1808 for compiling on Windows systems. I have therefore replaced it with
1809 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1810 defaults to "extern" for C or "extern C" for C++, which works fine on
1811 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1812 SCOPE with something explicit in config.h. In addition:
1813
1814 (a) pcreposix.h still had just "extern" instead of either of these macros;
1815 I have replaced it with PCRE_DATA_SCOPE.
1816
1817 (b) Functions such as _pcre_xclass(), which are internal to the library,
1818 but external in the C sense, all had PCRE_EXPORT in their definitions.
1819 This is apparently wrong for the Windows case, so I have removed it.
1820 (It makes no difference on Unix-like systems.)
1821
1822 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1823 of recursive calls to match(). This is different to MATCH_LIMIT because
1824 that limits the total number of calls to match(), not all of which increase
1825 the depth of recursion. Limiting the recursion depth limits the amount of
1826 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1827 when PCRE is compiled, and changed at run time. A patch from Google adds
1828 this functionality to the C++ interface.
1829
1830 18. Changes to the handling of Unicode character properties:
1831
1832 (a) Updated the table to Unicode 4.1.0.
1833
1834 (b) Recognize characters that are not in the table as "Cn" (undefined).
1835
1836 (c) I revised the way the table is implemented to a much improved format
1837 which includes recognition of ranges. It now supports the ranges that
1838 are defined in UnicodeData.txt, and it also amalgamates other
1839 characters into ranges. This has reduced the number of entries in the
1840 table from around 16,000 to around 3,000, thus reducing its size
1841 considerably. I realized I did not need to use a tree structure after
1842 all - a binary chop search is just as efficient. Having reduced the
1843 number of entries, I extended their size from 6 bytes to 8 bytes to
1844 allow for more data.
1845
1846 (d) Added support for Unicode script names via properties such as \p{Han}.
1847
1848 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1849 matching that character.
1850
1851 20. When matching a repeated Unicode property with a minimum greater than zero,
1852 (for example \pL{2,}), PCRE could look past the end of the subject if it
1853 reached it while seeking the minimum number of characters. This could
1854 happen only if some of the characters were more than one byte long, because
1855 there is a check for at least the minimum number of bytes.
1856
1857 21. Refactored the implementation of \p and \P so as to be more general, to
1858 allow for more different types of property in future. This has changed the
1859 compiled form incompatibly. Anybody with saved compiled patterns that use
1860 \p or \P will have to recompile them.
1861
1862 22. Added "Any" and "L&" to the supported property types.
1863
1864 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1865 but give a compile time error if the value is greater than 0xff.
1866
1867 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1868 accidentally not being installed or uninstalled.
1869
1870 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1871 made were to insert the current release number. This seemed silly, because
1872 it made things harder for people building PCRE on systems that don't run
1873 "configure". I have turned pcre.h into a distributed file, no longer built
1874 by "configure", with the version identification directly included. There is
1875 no longer a pcre.h.in file.
1876
1877 However, this change necessitated a change to the pcre-config script as
1878 well. It is built from pcre-config.in, and one of the substitutions was the
1879 release number. I have updated configure.ac so that ./configure now finds
1880 the release number by grepping pcre.h.
1881
1882 26. Added the ability to run the tests under valgrind.
1883
1884
1885 Version 6.4 05-Sep-05
1886 ---------------------
1887
1888 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1889 "--" to be printed when multiple files were scanned, even when none of the
1890 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1891 consider it to be a bug, and have restored the previous behaviour.
1892
1893 2. A couple of code tidies to get rid of compiler warnings.
1894
1895 3. The pcretest program used to cheat by referring to symbols in the library
1896 whose names begin with _pcre_. These are internal symbols that are not
1897 really supposed to be visible externally, and in some environments it is
1898 possible to suppress them. The cheating is now confined to including
1899 certain files from the library's source, which is a bit cleaner.
1900
1901 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1902 file's purpose clearer.
1903
1904 5. Reorganized pcre_ucp_findchar().
1905
1906
1907 Version 6.3 15-Aug-05
1908 ---------------------
1909
1910 1. The file libpcre.pc.in did not have general read permission in the tarball.
1911
1912 2. There were some problems when building without C++ support:
1913
1914 (a) If C++ support was not built, "make install" and "make test" still
1915 tried to test it.
1916
1917 (b) There were problems when the value of CXX was explicitly set. Some
1918 changes have been made to try to fix these, and ...
1919
1920 (c) --disable-cpp can now be used to explicitly disable C++ support.
1921
1922 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1923 backslash in a target when C++ was disabled. This confuses some
1924 versions of "make", apparently. Using an intermediate variable solves
1925 this. (Same for CPP_LOBJ.)
1926
1927 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1928 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1929 necessary on certain architectures.
1930
1931 4. Added a setting of -export-symbols-regex to the link command to remove
1932 those symbols that are exported in the C sense, but actually are local
1933 within the library, and not documented. Their names all begin with
1934 "_pcre_". This is not a perfect job, because (a) we have to except some
1935 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1936 available (and never for static libraries). I have made a note to try to
1937 find a way round (a) in the future.
1938
1939
1940 Version 6.2 01-Aug-05
1941 ---------------------
1942
1943 1. There was no test for integer overflow of quantifier values. A construction
1944 such as {1111111111111111} would give undefined results. What is worse, if
1945 a minimum quantifier for a parenthesized subpattern overflowed and became
1946 negative, the calculation of the memory size went wrong. This could have
1947 led to memory overwriting.
1948
1949 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1950
1951 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1952 operating environments where this matters.
1953
1954 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1955 PCRE options from within the C++ wrapper.
1956
1957 5. Named capturing subpatterns were not being correctly counted when a pattern
1958 was compiled. This caused two problems: (a) If there were more than 100
1959 such subpatterns, the calculation of the memory needed for the whole
1960 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1961 back references of the form \12, where the number was greater than 9, were
1962 not recognized as back references, even though there were sufficient
1963 previous subpatterns.
1964
1965 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1966 versions of gcc, e.g. 2.95.4.
1967
1968
1969 Version 6.1 21-Jun-05
1970 ---------------------
1971
1972 1. There was one reference to the variable "posix" in pcretest.c that was not
1973 surrounded by "#if !defined NOPOSIX".
1974
1975 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1976 the cross-check on the old pcre_info() function, for the benefit of the
1977 cut-down version of PCRE that is currently imported into Exim.
1978
1979 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1980 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1981 patterns that start (?i) but I don't think that matters. The use of (?i) is
1982 just an example; this all applies to the other options as well.
1983
1984 4. Since libtool seems to echo the compile commands it is issuing, the output
1985 from "make" can be reduced a bit by putting "@" in front of each libtool
1986 compile command.
1987
1988 5. Patch from the folks at Google for configure.in to be a bit more thorough
1989 in checking for a suitable C++ installation before trying to compile the
1990 C++ stuff. This should fix a reported problem when a compiler was present,
1991 but no suitable headers.
1992
1993 6. The man pages all had just "PCRE" as their title. I have changed them to
1994 be the relevant file name. I have also arranged that these names are
1995 retained in the file doc/pcre.txt, which is a concatenation in text format
1996 of all the man pages except the little individual ones for each function.
1997
1998 7. The NON-UNIX-USE file had not been updated for the different set of source
1999 files that come with release 6. I also added a few comments about the C++
2000 wrapper.
2001
2002
2003 Version 6.0 07-Jun-05
2004 ---------------------
2005
2006 1. Some minor internal re-organization to help with my DFA experiments.
2007
2008 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2009 didn't matter for the library itself when fully configured, but did matter
2010 when compiling without UCP support, or within Exim, where the ucp files are
2011 not imported.
2012
2013 3. Refactoring of the library code to split up the various functions into
2014 different source modules. The addition of the new DFA matching code (see
2015 below) to a single monolithic source would have made it really too
2016 unwieldy, quite apart from causing all the code to be include in a
2017 statically linked application, when only some functions are used. This is
2018 relevant even without the DFA addition now that patterns can be compiled in
2019 one application and matched in another.
2020
2021 The downside of splitting up is that there have to be some external
2022 functions and data tables that are used internally in different modules of
2023 the library but which are not part of the API. These have all had their
2024 names changed to start with "_pcre_" so that they are unlikely to clash
2025 with other external names.
2026
2027 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2028 a different (DFA) algorithm. Although it is slower than the original
2029 function, it does have some advantages for certain types of matching
2030 problem.
2031
2032 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2033 including restarting after a partial match.
2034
2035 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2036 defined when compiling for Windows was sent to me. I have put it into the
2037 code, though I have no means of testing or verifying it.
2038
2039 7. Added the pcre_refcount() auxiliary function.
2040
2041 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2042 match before or at the first newline in the subject string. In pcretest,
2043 the /f option on a pattern can be used to set this.
2044
2045 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2046 would behave wrongly. This has been present in PCRE since release 4.0.
2047
2048 10. A number of changes to the pcregrep command:
2049
2050 (a) Refactored how -x works; insert ^(...)$ instead of setting
2051 PCRE_ANCHORED and checking the length, in preparation for adding
2052 something similar for -w.
2053
2054 (b) Added the -w (match as a word) option.
2055
2056 (c) Refactored the way lines are read and buffered so as to have more
2057 than one at a time available.
2058
2059 (d) Implemented a pcregrep test script.
2060
2061 (e) Added the -M (multiline match) option. This allows patterns to match
2062 over several lines of the subject. The buffering ensures that at least
2063 8K, or the rest of the document (whichever is the shorter) is available
2064 for matching (and similarly the previous 8K for lookbehind assertions).
2065
2066 (f) Changed the --help output so that it now says
2067
2068 -w, --word-regex(p)
2069
2070 instead of two lines, one with "regex" and the other with "regexp"
2071 because that confused at least one person since the short forms are the
2072 same. (This required a bit of code, as the output is generated
2073 automatically from a table. It wasn't just a text change.)
2074
2075 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2076 option but starts with a hyphen. Could be a pattern or a path name
2077 starting with a hyphen, for instance.
2078
2079 (h) "-" can be given as a file name to represent stdin.
2080
2081 (i) When file names are being printed, "(standard input)" is used for
2082 the standard input, for compatibility with GNU grep. Previously
2083 "<stdin>" was used.
2084
2085 (j) The option --label=xxx can be used to supply a name to be used for
2086 stdin when file names are being printed. There is no short form.
2087
2088 (k) Re-factored the options decoding logic because we are going to add
2089 two more options that take data. Such options can now be given in four
2090 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2091
2092 (l) Added the -A, -B, and -C options for requesting that lines of context
2093 around matches be printed.
2094
2095 (m) Added the -L option to print the names of files that do not contain
2096 any matching lines, that is, the complement of -l.
2097
2098 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2099 continue to scan other files.
2100
2101 (o) The -s option was incorrectly implemented. For compatibility with other
2102 greps, it now suppresses the error message for a non-existent or non-
2103 accessible file (but not the return code). There is a new option called
2104 -q that suppresses the output of matching lines, which was what -s was
2105 previously doing.
2106
2107 (p) Added --include and --exclude options to specify files for inclusion
2108 and exclusion when recursing.
2109
2110 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2111 Hopefully, it now does.
2112
2113 12. Missing cast in pcre_study().
2114
2115 13. Added an "uninstall" target to the makefile.
2116
2117 14. Replaced "extern" in the function prototypes in Makefile.in with
2118 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2119 world, but is set differently for Windows.
2120
2121 15. Added a second compiling function called pcre_compile2(). The only
2122 difference is that it has an extra argument, which is a pointer to an
2123 integer error code. When there is a compile-time failure, this is set
2124 non-zero, in addition to the error test pointer being set to point to an
2125 error message. The new argument may be NULL if no error number is required
2126 (but then you may as well call pcre_compile(), which is now just a
2127 wrapper). This facility is provided because some applications need a
2128 numeric error indication, but it has also enabled me to tidy up the way
2129 compile-time errors are handled in the POSIX wrapper.
2130
2131 16. Added VPATH=.libs to the makefile; this should help when building with one
2132 prefix path and installing with another. (Or so I'm told by someone who
2133 knows more about this stuff than I do.)
2134
2135 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2136 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2137 match everything, including newlines. This is not POSIX-compatible, but
2138 somebody wanted the feature. From pcretest it can be activated by using
2139 both the P and the s flags.
2140
2141 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2142
2143 19. libpcre.pc was being incorrectly installed as executable.
2144
2145 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2146 it now also looks for '\r' so that it will work unmodified on Windows.
2147
2148 21. Added Google's contributed C++ wrapper to the distribution.
2149
2150 22. Added some untidy missing memory free() calls in pcretest, to keep
2151 Electric Fence happy when testing.
2152
2153
2154
2155 Version 5.0 13-Sep-04
2156 ---------------------
2157
2158 1. Internal change: literal characters are no longer packed up into items
2159 containing multiple characters in a single byte-string. Each character
2160 is now matched using a separate opcode. However, there may be more than one
2161 byte in the character in UTF-8 mode.
2162
2163 2. The pcre_callout_block structure has two new fields: pattern_position and
2164 next_item_length. These contain the offset in the pattern to the next match
2165 item, and its length, respectively.
2166
2167 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2168 insertion of callouts before each pattern item. Added the /C option to
2169 pcretest to make use of this.
2170
2171 4. On the advice of a Windows user, the lines
2172
2173 #if defined(_WIN32) || defined(WIN32)
2174 _setmode( _fileno( stdout ), 0x8000 );
2175 #endif /* defined(_WIN32) || defined(WIN32) */
2176
2177 have been added to the source of pcretest. This apparently does useful
2178 magic in relation to line terminators.
2179
2180 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2181 for the benefit of those environments where the "b" makes a difference.
2182
2183 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2184 to know about it. I have put a hack into configure.in that adds in code
2185 to set GCC=yes if CC=icc. This seems to end up at a point in the
2186 generated configure script that is early enough to affect the setting of
2187 compiler options, which is what is needed, but I have no means of testing
2188 whether it really works. (The user who reported this had patched the
2189 generated configure script, which of course I cannot do.)
2190
2191 LATER: After change 22 below (new libtool files), the configure script
2192 seems to know about icc (and also ecc). Therefore, I have commented out
2193 this hack in configure.in.
2194
2195 7. Added support for pkg-config (2 patches were sent in).
2196
2197 8. Negated POSIX character classes that used a combination of internal tables
2198 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2199 [[:^ascii]]. Typically, they would match almost any characters. The other
2200 POSIX classes were not broken in this way.
2201
2202 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2203 to find the match, as PCRE was deluded into thinking that the match had to
2204 start at the start point or following a newline. The same bug applied to
2205 patterns with negative forward assertions or any backward assertions
2206 preceding ".*" at the start, unless the pattern required a fixed first
2207 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2208
2209 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2210 starting at the last subject character, bytes beyond the end of the subject
2211 string were read.
2212
2213 11. Renamed the variable "class" as "classbits" to make life easier for C++
2214 users. (Previously there was a macro definition, but it apparently wasn't
2215 enough.)
2216
2217 12. Added the new field "tables" to the extra data so that tables can be passed
2218 in at exec time, or the internal tables can be re-selected. This allows
2219 a compiled regex to be saved and re-used at a later time by a different
2220 program that might have everything at different addresses.
2221
2222 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2223 -R library as well as a -L library.
2224
2225 14. The debugging options of pcretest (-d on the command line or D on a
2226 pattern) showed incorrect output for anything following an extended class
2227 that contained multibyte characters and which was followed by a quantifier.
2228
2229 15. Added optional support for general category Unicode character properties
2230 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2231 support. It adds about 90K to the size of the library. The meanings of the
2232 inbuilt class escapes such as \d and \s have NOT been changed.
2233
2234 16. Updated pcredemo.c to include calls to free() to release the memory for the
2235 compiled pattern.
2236
2237 17. The generated file chartables.c was being created in the source directory
2238 instead of in the building directory. This caused the build to fail if the
2239 source directory was different from the building directory, and was
2240 read-only.
2241
2242 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2243 file. No doubt somebody will tell me if they don't make sense... Also added
2244 Dan Mooney's comments about building on OpenVMS.
2245
2246 19. Added support for partial matching via the PCRE_PARTIAL option for
2247 pcre_exec() and the \P data escape in pcretest.
2248
2249 20. Extended pcretest with 3 new pattern features:
2250
2251 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2252 write the compiled pattern to the file whose name is "rest-of-line".
2253 This is a straight binary dump of the data, with the saved pointer to
2254 the character tables forced to be NULL. The study data, if any, is
2255 written too. After writing, pcretest reads a new pattern.
2256
2257 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2258 compiled pattern from the given file. There must not be any
2259 occurrences of "<" in the file name (pretty unlikely); if there are,
2260 pcretest will instead treat the initial "<" as a pattern delimiter.
2261 After reading in the pattern, pcretest goes on to read data lines as
2262 usual.
2263
2264 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2265 and 16-bit fields in a compiled pattern, to simulate a pattern that
2266 was compiled on a host of opposite endianness.
2267
2268 21. The pcre-exec() function can now cope with patterns that were compiled on
2269 hosts of opposite endianness, with this restriction:
2270
2271 As for any compiled expression that is saved and used later, the tables
2272 pointer field cannot be preserved; the extra_data field in the arguments
2273 to pcre_exec() should be used to pass in a tables address if a value
2274 other than the default internal tables were used at compile time.
2275
2276 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2277 now diagnosed as an error. Previously, most of the time, a negative number
2278 would have been treated as zero, but if in addition "ovector" was passed as
2279 NULL, a crash could occur.
2280
2281 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2282 new versions from the libtool 1.5 distribution (the last one is a copy of
2283 a file called libtool.m4). This seems to have fixed the need to patch
2284 "configure" to support Darwin 1.3 (which I used to do). However, I still
2285 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2286 workstation).
2287
2288 24. Changed the PCRE licence to be the more standard "BSD" licence.
2289
2290
2291 Version 4.5 01-Dec-03
2292 ---------------------
2293
2294 1. There has been some re-arrangement of the code for the match() function so
2295 that it can be compiled in a version that does not call itself recursively.
2296 Instead, it keeps those local variables that need separate instances for
2297 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2298 needs to "recurse". Keeping track of where control must go is done by means
2299 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2300 hide most of the details from the main code, and operates only if
2301 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2302 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2303 operating.
2304
2305 To make it easier for callers to provide specially tailored get/free
2306 functions for this usage, two new functions, pcre_stack_malloc, and
2307 pcre_stack_free, are used. They are always called in strict stacking order,
2308 and the size of block requested is always the same.
2309
2310 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2311 PCRE has been compiled to use the stack or the heap for recursion. The
2312 -C option of pcretest uses this to show which version is compiled.
2313
2314 A new data escape \S, is added to pcretest; it causes the amounts of store
2315 obtained and freed by both kinds of malloc/free at match time to be added
2316 to the output.
2317
2318 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2319 what's available on my current Linux desktop machine.
2320
2321 3. When matching a UTF-8 string, the test for a valid string at the start has
2322 been extended. If start_offset is not zero, PCRE now checks that it points
2323 to a byte that is the start of a UTF-8 character. If not, it returns
2324 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2325 this is necessary because there may be backward assertions in the pattern.
2326 When matching the same subject several times, it may save resources to use
2327 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2328
2329 4. The code for checking the validity of UTF-8 strings has been tightened so
2330 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2331 containing "overlong sequences".
2332
2333 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2334 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2335 should have been "&", but it just so happened that all the cases this let
2336 through by mistake were picked up later in the function.
2337
2338 6. I had used a variable called "isblank" - this is a C99 function, causing
2339 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2340
2341 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2342 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2343 so the version is shown; (c) stop "make test" echoing "./RunTest".
2344
2345 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2346
2347 9. The prototype for memmove() for systems that don't have it was using
2348 size_t, but the inclusion of the header that defines size_t was later. I've
2349 moved the #includes for the C headers earlier to avoid this.
2350
2351 10. Added some adjustments to the code to make it easier to compiler on certain
2352 special systems:
2353
2354 (a) Some "const" qualifiers were missing.
2355 (b) Added the macro EXPORT before all exported functions; by default this
2356 is defined to be empty.
2357 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2358 that it reads its output file name as an argument instead of writing
2359 to the standard output and assuming this can be redirected.
2360
2361 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2362 class containing characters with values greater than 255, PCRE compilation
2363 went into a loop.
2364
2365 12. A recursive reference to a subpattern that was within another subpattern
2366 that had a minimum quantifier of zero caused PCRE to crash. For example,
2367 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2368 recursion. If the recursively-called subpattern itself had a zero repeat,
2369 that was OK.
2370
2371 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2372 buffer into which it was copied (for escape processing) was still set at
2373 1024, so long lines caused crashes.
2374
2375 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2376 "internal error: code overflow...". This applied to any character class
2377 that was followed by a possessive quantifier.
2378
2379 15. Modified the Makefile to add libpcre.la as a prerequisite for
2380 libpcreposix.la because I was told this is needed for a parallel build to
2381 work.
2382
2383 16. If a pattern that contained .* following optional items at the start was
2384 studied, the wrong optimizing data was generated, leading to matching
2385 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2386 matching string must start with a or b or c. The correct conclusion for
2387 this pattern is that a match can start with any character.
2388
2389
2390 Version 4.4 13-Aug-03
2391 ---------------------
2392
2393 1. In UTF-8 mode, a character class containing characters with values between
2394 127 and 255 was not handled correctly if the compiled pattern was studied.
2395 In fixing this, I have also improved the studying algorithm for such
2396 classes (slightly).
2397
2398 2. Three internal functions had redundant arguments passed to them. Removal
2399 might give a very teeny performance improvement.
2400
2401 3. Documentation bug: the value of the capture_top field in a callout is *one
2402 more than* the number of the hightest numbered captured substring.
2403
2404 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2405 in incorrectly linking with a previously installed version. They now link
2406 explicitly with libpcre.la.
2407
2408 5. configure.in no longer needs to recognize Cygwin specially.
2409
2410 6. A problem in pcre.in for Windows platforms is fixed.
2411
2412 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2413 pcretest, it used to include the size of the study block as part of its
2414 output. Unfortunately, the structure contains a field that has a different
2415 size on different hardware architectures. This meant that the tests that
2416 showed this size failed. As the block is currently always of a fixed size,
2417 this information isn't actually particularly useful in pcretest output, so
2418 I have just removed it.
2419
2420 8. Three pre-processor statements accidentally did not start in column 1.
2421 Sadly, there are *still* compilers around that complain, even though
2422 standard C has not required this for well over a decade. Sigh.
2423
2424 9. In pcretest, the code for checking callouts passed small integers in the
2425 callout_data field, which is a void * field. However, some picky compilers
2426 complained about the casts involved for this on 64-bit systems. Now
2427 pcretest passes the address of the small integer instead, which should get
2428 rid of the warnings.
2429
2430 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2431 both compile and run time, and gives an error if an invalid UTF-8 sequence
2432 is found. There is a option for disabling this check in cases where the
2433 string is known to be correct and/or the maximum performance is wanted.
2434
2435 11. In response to a bug report, I changed one line in Makefile.in from
2436
2437 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2438 to
2439 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2440
2441 to look similar to other lines, but I have no way of telling whether this
2442 is the right thing to do, as I do not use Windows. No doubt I'll get told
2443 if it's wrong...
2444
2445
2446 Version 4.3 21-May-03
2447 ---------------------
2448
2449 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2450 Makefile.
2451
2452 2. Some refactoring to improve the quality of the code:
2453
2454 (i) The utf8_table... variables are now declared "const".
2455
2456 (ii) The code for \cx, which used the "case flipping" table to upper case
2457 lower case letters, now just substracts 32. This is ASCII-specific,
2458 but the whole concept of \cx is ASCII-specific, so it seems
2459 reasonable.
2460
2461 (iii) PCRE was using its character types table to recognize decimal and
2462 hexadecimal digits in the pattern. This is silly, because it handles
2463 only 0-9, a-f, and A-F, but the character types table is locale-
2464 specific, which means strange things might happen. A private
2465 table is now used for this - though it costs 256 bytes, a table is
2466 much faster than multiple explicit tests. Of course, the standard
2467 character types table is still used for matching digits in subject
2468 strings against \d.
2469
2470 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2471 ending in _t are). So I've renamed it as ESC_tee.
2472
2473 3. The first argument for regexec() in the POSIX wrapper should have been
2474 defined as "const".
2475
2476 4. Changed pcretest to use malloc() for its buffers so that they can be
2477 Electric Fenced for debugging.
2478
2479 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2480 to read one or more bytes before the start of the subject string. Often this
2481 had no effect on PCRE's behaviour, but in some circumstances it could
2482 provoke a segmentation fault.
2483
2484 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2485 to try to read one or more bytes before the start of the subject string.
2486
2487 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2488 UTF-8 support could misbehave in various ways if the subject string
2489 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2490 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2491 back over UTF-8 characters.)
2492
2493
2494 Version 4.2 14-Apr-03
2495 ---------------------
2496
2497 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2498
2499 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2500 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2501 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2502 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2503 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2504 and BUILD_EXEEXT
2505 Note: automatic setting of the BUILD variables is not yet working
2506 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2507 compile-time but not at link-time
2508 [LINK]: use for linking executables only
2509 make different versions for Windows and non-Windows
2510 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2511 libraries
2512 [LINK_FOR_BUILD]: new variable
2513 [OBJEXT]: use throughout
2514 [EXEEXT]: use throughout
2515 <winshared>: new target
2516 <wininstall>: new target
2517 <dftables.o>: use native compiler
2518 <dftables>: use native linker
2519 <install>: handle Windows platform correctly
2520 <clean>: ditto
2521 <check>: ditto
2522 copy DLL to top builddir before testing
2523
2524 As part of these changes, -no-undefined was removed again. This was reported
2525 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2526 in any case.
2527
2528 3. Some tidies to get rid of compiler warnings:
2529
2530 . In the match_data structure, match_limit was an unsigned long int, whereas
2531 match_call_count was an int. I've made them both unsigned long ints.
2532
2533 . In pcretest the fact that a const uschar * doesn't automatically cast to
2534 a void * provoked a warning.
2535
2536 . Turning on some more compiler warnings threw up some "shadow" variables
2537 and a few more missing casts.
2538
2539 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2540 option, a class that contained a single character with a value between 128
2541 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2542
2543 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2544 option, a class that contained several characters, but with at least one
2545 whose value was between 128 and 255 caused PCRE to crash.
2546
2547
2548 Version 4.1 12-Mar-03
2549 ---------------------
2550
2551 1. Compiling with gcc -pedantic found a couple of places where casts were
2552 needed, and a string in dftables.c that was longer than standard compilers are
2553 required to support.
2554
2555 2. Compiling with Sun's compiler found a few more places where the code could
2556 be tidied up in order to avoid warnings.
2557
2558 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2559 first of these names is deprecated in the latest Autoconf in favour of the name
2560 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2561 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2562 analogy I have changed it to CFLAGS_FOR_BUILD.
2563
2564 4. Added -no-undefined to the linking command in the Makefile, because this is
2565 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2566 linking step for the pcreposix library.
2567
2568 5. PCRE was failing to diagnose the case of two named groups with the same
2569 name.
2570
2571 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2572 literal character that is needed in the subject for a match, and scans along to
2573 ensure that it is present before embarking on the full matching process. This
2574 saves time in cases of nested unlimited repeats that are never going to match.
2575 Problem: the scan can take a lot of time if the subject is very long (e.g.
2576 megabytes), thus penalizing straightforward matches. It is now done only if the
2577 amount of subject to be scanned is less than 1000 bytes.
2578
2579 7. A lesser problem with the same optimization is that it was recording the
2580 first character of an anchored pattern as "needed", thus provoking a search
2581 right along the subject, even when the first match of the pattern was going to
2582 fail. The "needed" character is now not set for anchored patterns, unless it
2583 follows something in the pattern that is of non-fixed length. Thus, it still
2584 fulfils its original purpose of finding quick non-matches in cases of nested
2585 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2586
2587
2588 Version 4.0 17-Feb-03
2589 ---------------------
2590
2591 1. If a comment in an extended regex that started immediately after a meta-item
2592 extended to the end of string, PCRE compiled incorrect data. This could lead to
2593 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2594
2595 2. Moved to autoconf 2.53 and libtool 1.4.2.
2596
2597 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2598 the special perltest8 script is no longer needed - all the tests can be run
2599 from a single perltest script.
2600
2601 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2602 by \s. It has now been removed in PCRE. This means it isn't recognized as
2603 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2604 class [:space:] *does* include VT, thereby creating a mess.
2605
2606 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2607 space and tab.
2608
2609 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2610 its new features into the main test script, reducing the number of scripts.
2611
2612 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2613 were backward compatible, and made the (?i) apply to the whole pattern, as if
2614 /i were given. Now it behaves more logically, and applies the option setting
2615 only to what follows. PCRE has been changed to follow suit. However, if it
2616 finds options settings right at the start of the pattern, it extracts them into
2617 the global options, as before. Thus, they show up in the info data.
2618
2619 8. Added support for the \Q...\E escape sequence. Characters in between are
2620 treated as literals. This is slightly different from Perl in that $ and @ are
2621 also handled as literals inside the quotes. In Perl, they will cause variable
2622 interpolation. Note the following examples:
2623
2624 Pattern PCRE matches Perl matches
2625
2626 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2627 \Qabc\$xyz\E abc\$xyz abc\$xyz
2628 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2629
2630 For compatibility with Perl, \Q...\E sequences are recognized inside character
2631 classes as well as outside them.
2632
2633 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2634 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2635 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2636 signed/unsigned warnings.
2637
2638 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2639 option for pcretest, so I've replaced it by a simple function that does just
2640 that job.
2641
2642 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2643 "pcregrep -".
2644
2645 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2646 Java package. This provides some syntactic sugar for simple cases of what my
2647 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2648 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2649 item, you can use this simplified notation. Note that only makes sense with
2650 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2651 greediness, whatever the setting of the PCRE_UNGREEDY option.
2652
2653 13. A change of greediness default within a pattern was not taking effect at
2654 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2655 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2656 was abstracted outside.
2657
2658 14. PCRE now supports the \G assertion. It is true when the current matching
2659 position is at the start point of the match. This differs from \A when the
2660 starting offset is non-zero. Used with the /g option of pcretest (or similar
2661 code), it works in the same way as it does for Perl's /g option. If all
2662 alternatives of a regex begin with \G, the expression is anchored to the start
2663 match position, and the "anchored" flag is set in the compiled expression.
2664
2665 15. Some bugs concerning the handling of certain option changes within patterns
2666 have been fixed. These applied to options other than (?ims). For example,
2667 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2668 been the other way round. Some of this was related to change 7 above.
2669
2670 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2671 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2672 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2673 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2674
2675 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2676 mode. Unlike ".", it always matches newline, whatever the setting of
2677 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2678 assertions. Perl allows it, but it doesn't (in general) work because it can't
2679 calculate the length of the lookbehind. At least, that's the case for Perl
2680 5.8.0 - I've been told they are going to document that it doesn't work in
2681 future.
2682
2683 18. Added an error diagnosis for escapes that PCRE does not support: these are
2684 \L, \l, \N, \P, \p, \U, \u, and \X.
2685
2686 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2687 reading past the end of the pattern in cases such as /[abcd/.
2688
2689 20. PCRE was getting more memory than necessary for patterns with classes that
2690 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2691
2692 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2693 compiling PCRE for use with Virtual Pascal.
2694
2695 22. Small fix to the Makefile to make it work properly if the build is done
2696 outside the source tree.
2697
2698 23. Added a new extension: a condition to go with recursion. If a conditional
2699 subpattern starts with (?(R) the "true" branch is used if recursion has
2700 happened, whereas the "false" branch is used only at the top level.
2701
2702 24. When there was a very long string of literal characters (over 255 bytes
2703 without UTF support, over 250 bytes with UTF support), the computation of how
2704 much memory was required could be incorrect, leading to segfaults or other
2705 strange effects.
2706
2707 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2708 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2709 there was a subsequent back reference to those brackets. This meant that, for
2710 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2711 possible to check for precisely this case. All we can do is abandon the
2712 optimization if .* occurs inside capturing brackets when there are any back
2713 references whatsoever. (See below for a better fix that came later.)
2714
2715 26. The handling of the optimization for finding the first character of a
2716 non-anchored pattern, and for finding a character that is required later in the
2717 match were failing in some cases. This didn't break the matching; it just
2718 failed to optimize when it could. The way this is done has been re-implemented.
2719
2720 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2721
2722 28. Added a new feature that provides some of the functionality that Perl
2723 provides with (?{...}). The facility is termed a "callout". The way it is done
2724 in PCRE is for the caller to provide an optional function, by setting
2725 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2726 global variable. By default it is unset, which disables all calling out. To get
2727 the function called, the regex must include (?C) at appropriate points. This
2728 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2729 This provides a means of identifying different callout points. When PCRE
2730 reaches such a point in the regex, if pcre_callout has been set, the external
2731 function is called. It is provided with data in a structure called
2732 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2733 matching continues; if it returns a non-zero value, the match at the current
2734 point fails. However, backtracking will occur if possible. [This was changed
2735 later and other features added - see item 49 below.]
2736
2737 29. pcretest is upgraded to test the callout functionality. It provides a
2738 callout function that displays information. By default, it shows the start of
2739 the match and the current position in the text. There are some new data escapes
2740 to vary what happens:
2741
2742 \C+ in addition, show current contents of captured substrings
2743 \C- do not supply a callout function
2744 \C!n return 1 when callout number n is reached
2745 \C!n!m return 1 when callout number n is reached for the mth time
2746
2747 30. If pcregrep was called with the -l option and just a single file name, it
2748 output "<stdin>" if a match was found, instead of the file name.
2749
2750 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2751 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2752 pcre_exec(). This saves a malloc/free per call. The default value of
2753 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2754 when configuring.
2755
2756 32. The default maximum size of a compiled pattern is 64K. There have been a
2757 few cases of people hitting this limit. The code now uses macros to handle the
2758 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2759 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2760 configuring. Tests 2 and 5 work only with 2-byte links because they output
2761 debugging information about compiled patterns.
2762
2763 33. Internal code re-arrangements:
2764
2765 (a) Moved the debugging function for printing out a compiled regex into
2766 its own source file (printint.c) and used #include to pull it into
2767 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2768 separate copies.
2769
2770 (b) Defined the list of op-code names for debugging as a macro in
2771 internal.h so that it is next to the definition of the opcodes.
2772
2773 (c) Defined a table of op-code lengths for simpler skipping along compiled
2774 code. This is again a macro in internal.h so that it is next to the
2775 definition of the opcodes.
2776
2777 34. Added support for recursive calls to individual subpatterns, along the
2778 lines of Robin Houston's patch (but implemented somewhat differently).
2779
2780 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2781 allow it to read and process whole directories in Win32. This code was
2782 contributed by Lionel Fourquaux; it has not been tested by me.
2783
2784 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2785 used to name a group. Names consist of alphanumerics and underscores, and must
2786 be unique. Back references use the syntax (?P=name) and recursive calls use
2787 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2788 numbers. The function pcre_fullinfo() can be used after compilation to extract
2789 a name/number map. There are three relevant calls:
2790
2791 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2792 PCRE_INFO_NAMECOUNT yields the number of entries
2793 PCRE_INFO_NAMETABLE yields a pointer to the map.
2794
2795 The map is a vector of fixed-size entries. The size of each entry depends on
2796 the length of the longest name used. The first two bytes of each entry are the
2797 group number, most significant byte first. There follows the corresponding
2798 name, zero terminated. The names are in alphabetical order.
2799
2800 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2801 case instead of 255. Making it the same both with and without UTF-8 support
2802 means that the same test output works with both.
2803
2804 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2805 calling malloc() with a zero argument.
2806
2807 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2808 optimization. I've improved things by keeping a bitmap of backreferences with
2809 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2810 fact referenced, the optimization can be applied. It is unlikely that a
2811 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2812 the match to follow \n) will appear inside brackets with a number greater than
2813 31, but if it does, any back reference > 31 suppresses the optimization.
2814
2815 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2816 of disabling numbered capturing parentheses. Any opening parenthesis that is
2817 not followed by ? behaves as if it were followed by ?: but named parentheses
2818 can still be used for capturing (and they will acquire numbers in the usual
2819 way).
2820
2821 41. Redesigned the return codes from the match() function into yes/no/error so
2822 that errors can be passed back from deep inside the nested calls. A malloc
2823 failure while inside a recursive subpattern call now causes the
2824 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2825
2826 42. It is now possible to set a limit on the number of times the match()
2827 function is called in a call to pcre_exec(). This facility makes it possible to
2828 limit the amount of recursion and backtracking, though not in a directly
2829 obvious way, because the match() function is used in a number of different
2830 circumstances. The count starts from zero for each position in the subject
2831 string (for non-anchored patterns). The default limit is, for compatibility, a
2832 large number, namely 10 000 000. You can change this in two ways:
2833
2834 (a) When configuring PCRE before making, you can use --with-match-limit=n
2835 to set a default value for the compiled library.
2836
2837 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2838 a different value is set. See 45 below.
2839
2840 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2841
2842 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2843 of things that can be changed at compile time. The first argument specifies
2844 what is wanted and the second points to where the information is to be placed.
2845 The current list of available information is:
2846
2847 PCRE_CONFIG_UTF8
2848
2849 The output is an integer that is set to one if UTF-8 support is available;
2850 otherwise it is set to zero.
2851
2852 PCRE_CONFIG_NEWLINE
2853
2854 The output is an integer that it set to the value of the code that is used for
2855 newline. It is either LF (10) or CR (13).
2856
2857 PCRE_CONFIG_LINK_SIZE
2858
2859 The output is an integer that contains the number of bytes used for internal
2860 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2861
2862 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2863
2864 The output is an integer that contains the threshold above which the POSIX
2865 interface uses malloc() for output vectors. See item 31 above.
2866
2867 PCRE_CONFIG_MATCH_LIMIT
2868
2869 The output is an unsigned integer that contains the default limit of the number
2870 of match() calls in a pcre_exec() execution. See 42 above.
2871
2872 44. pcretest has been upgraded by the addition of the -C option. This causes it
2873 to extract all the available output from the new pcre_config() function, and to
2874 output it. The program then exits immediately.
2875
2876 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2877 order to support additional features. One way would have been to define
2878 pcre_exec2() (for example) with extra arguments, but this would not have been
2879 extensible, and would also have required all calls to the original function to
2880 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2881 is used for passing in "extra" data from pcre_study().
2882
2883 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2884 contains the following fields:
2885
2886 flags a bitmap indicating which of the following fields are set
2887 study_data opaque data from pcre_study()
2888 match_limit a way of specifying a limit on match() calls for a specific
2889 call to pcre_exec()
2890 callout_data data for callouts (see 49 below)
2891
2892 The flag bits are also defined in pcre.h, and are
2893
2894 PCRE_EXTRA_STUDY_DATA
2895 PCRE_EXTRA_MATCH_LIMIT
2896 PCRE_EXTRA_CALLOUT_DATA
2897
2898 The pcre_study() function now returns one of these new pcre_extra blocks, with
2899 the actual study data pointed to by the study_data field, and the
2900 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2901 before. That is, this change is entirely upwards-compatible and requires no
2902 change to existing code.
2903
2904 If you want to pass in additional data to pcre_exec(), you can either place it
2905 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2906 block.
2907
2908 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2909 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2910 times with different match limits, until it finds the minimum value needed for
2911 pcre_exec() to complete. The value is then output. This can be instructive; for
2912 most simple matches the number is quite small, but for pathological cases it
2913 gets very large very quickly.
2914
2915 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2916 returns the size of the data block pointed to by the study_data field in a
2917 pcre_extra block, that is, the value that was passed as the argument to
2918 pcre_malloc() when PCRE was getting memory in which to place the information
2919 created by pcre_study(). The fourth argument should point to a size_t variable.
2920 pcretest has been extended so that this information is shown after a successful
2921 pcre_study() call when information about the compiled regex is being displayed.
2922
2923 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2924 because what follows is always an absolute path. (Later: it turns out that this
2925 is more than cosmetic for MinGW, because it doesn't like empty path
2926 components.)
2927
2928 49. Some changes have been made to the callout feature (see 28 above):
2929
2930 (i) A callout function now has three choices for what it returns:
2931
2932 0 => success, carry on matching
2933 > 0 => failure at this point, but backtrack if possible
2934 < 0 => serious error, return this value from pcre_exec()
2935
2936 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2937 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2938 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2939 use by callout functions. It will never be used by PCRE itself.
2940
2941 (ii) The pcre_extra structure (see 45 above) has a void * field called
2942 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2943 pcre_callout_block structure has a field of the same name. The contents of
2944 the field passed in the pcre_extra structure are passed to the callout
2945 function in the corresponding field in the callout block. This makes it
2946 easier to use the same callout-containing regex from multiple threads. For
2947 testing, the pcretest program has a new data escape
2948
2949 \C*n pass the number n (may be negative) as callout_data
2950
2951 If the callout function in pcretest receives a non-zero value as
2952 callout_data, it returns that value.
2953
2954 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2955 there were some redundant $(CFLAGS) in commands that are now specified as
2956 $(LINK), which already includes $(CFLAGS).
2957
2958 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2959 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2960 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2961 one-byte characters throughout. Note that case-insensitive matching applies
2962 only to characters whose values are less than 256. PCRE doesn't support the
2963 notion of cases for higher-valued characters.
2964
2965 (i) A character class whose characters are all within 0-255 is handled as
2966 a bit map, and the map is inverted for negative classes. Previously, a
2967 character > 255 always failed to match such a class; however it should
2968 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2969
2970 (ii) A negated character class with a single character < 255 is coded as
2971 "not this character" (OP_NOT). This wasn't working properly when the test
2972 character was multibyte, either singly or repeated.
2973
2974 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2975 mode, for example: \x{100}{2,3}.
2976
2977 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2978 singly or repeated) now correctly test multibyte characters. However,
2979 PCRE doesn't recognize any characters with values greater than 255 as
2980 digits, spaces, or word characters. Such characters always match \D, \S,
2981 and \W, and never match \d, \s, or \w.
2982
2983 (v) Classes may now contain characters and character ranges with values
2984 greater than 255. For example: [ab\x{100}-\x{400}].
2985
2986 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2987 PCRE in UTF-8 mode.
2988
2989 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2990 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2991 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2992 value.)
2993
2994 53. The single man page has become too large. I have therefore split it up into
2995 a number of separate man pages. These also give rise to individual HTML pages;
2996 these are now put in a separate directory, and there is an index.html page that
2997 lists them all. Some hyperlinking between the pages has been installed.
2998
2999 54. Added convenience functions for handling named capturing parentheses.
3000
3001 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3002 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3003 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3004 are faulted.
3005
3006 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3007 calling configure. These values are used when compiling the dftables.c program
3008 which is run to generate the source of the default character tables. They
3009 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3010 you will need to set these values.
3011
3012 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3013
3014
3015 Version 3.9 02-Jan-02
3016 ---------------------
3017
3018 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3019
3020 2. If --disable-static was given, the building process failed when trying to
3021 build pcretest and pcregrep. (For some reason it was using libtool to compile
3022 them, which is not right, as they aren't part of the library.)
3023
3024
3025 Version 3.8 18-Dec-01
3026 ---------------------
3027
3028 1. The experimental UTF-8 code was completely screwed up. It was packing the
3029 bytes in the wrong order. How dumb can you get?
3030
3031
3032 Version 3.7 29-Oct-01
3033 ---------------------
3034
3035 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3036 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3037 this didn't happen under Solaris 8, where I normally test things.
3038
3039 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3040 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3041 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3042 this fix an hour or so after the initial 3.7 release.)
3043
3044
3045 Version 3.6 23-Oct-01
3046 ---------------------
3047
3048 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3049 offsets passed as NULL with zero offset count.
3050
3051 2. The config.guess and config.sub files had not been updated when I moved to
3052 the latest autoconf.
3053
3054
3055 Version 3.5 15-Aug-01
3056 ---------------------
3057
3058 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3059 had been forgotten.
3060
3061 2. By using declared but undefined structures, we can avoid using "void"
3062 definitions in pcre.h while keeping the internal definitions of the structures
3063 private.
3064
3065 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3066 user point of view, this means that both static and shared libraries are built
3067 by default, but this can be individually controlled. More of the work of
3068 handling this static/shared cases is now inside libtool instead of PCRE's make
3069 file.
3070
3071 4. The pcretest utility is now installed along with pcregrep because it is
3072 useful for users (to test regexs) and by doing this, it automatically gets
3073 relinked by libtool. The documentation has been turned into a man page, so
3074 there are now .1, .txt, and .html versions in /doc.
3075
3076 5. Upgrades to pcregrep:
3077 (i) Added long-form option names like gnu grep.
3078 (ii) Added --help to list all options with an explanatory phrase.
3079 (iii) Added -r, --recursive to recurse into sub-directories.
3080 (iv) Added -f, --file to read patterns from a file.
3081
3082 6. pcre_exec() was referring to its "code" argument before testing that
3083 argument for NULL (and giving an error if it was NULL).
3084
3085 7. Upgraded Makefile.in to allow for compiling in a different directory from
3086 the source directory.
3087
3088 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3089 options bits, the pointer it was passed was to an int instead of to an unsigned
3090 long int. This mattered only on 64-bit systems.
3091
3092 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3093 generated) instead of pcre.in, which it its source. Also made the same change
3094 in several of the .c files.
3095
3096 10. A new release of gcc defines printf() as a macro, which broke pcretest
3097 because it had an ifdef in the middle of a string argument for printf(). Fixed
3098 by using separate calls to printf().
3099
3100 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3101 script, to force use of CR or LF instead of \n in the source. On non-Unix
3102 systems, the value can be set in config.h.
3103
3104 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3105 absolute limit. Changed the text of the error message to make this clear, and
3106 likewise updated the man page.
3107
3108 13. The limit of 99 on the number of capturing subpatterns has been removed.
3109 The new limit is 65535, which I hope will not be a "real" limit.
3110
3111
3112 Version 3.4 22-Aug-00
3113 ---------------------
3114
3115 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3116
3117 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3118
3119
3120 Version 3.3 01-Aug-00
3121 ---------------------
3122
3123 1. If an octal character was given, but the value was greater than \377, it
3124 was not getting masked to the least significant bits, as documented. This could
3125 lead to crashes in some systems.
3126
3127 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3128 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3129
3130 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3131 These just pass their arguments on to (pcre_free)(), but they are provided
3132 because some uses of PCRE bind it to non-C systems that can call its functions,
3133 but cannot call free() or pcre_free() directly.
3134
3135 4. Add "make test" as a synonym for "make check". Corrected some comments in
3136 the Makefile.
3137
3138 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3139 Makefile.
3140
3141 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3142 command called pgrep for grepping around the active processes.
3143
3144 7. Added the beginnings of support for UTF-8 character strings.
3145
3146 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3147 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3148 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3149 out for the ar command.)
3150
3151
3152 Version 3.2 12-May-00
3153 ---------------------
3154
3155 This is purely a bug fixing release.
3156
3157 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3158 of ZA. This was just one example of several cases that could provoke this bug,
3159 which was introduced by change 9 of version 2.00. The code for breaking
3160 infinite loops after an iteration that matches an empty string was't working
3161 correctly.
3162
3163 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3164 when matched against abbab (for example). After matching an empty string, it
3165 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3166 caused it to match further down the string than it should.
3167
3168 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3169 was there because it doesn't seem to be needed, and it causes trouble on some
3170 systems, as it is not a Standard C header. It has been removed.
3171
3172 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3173 were reported on the Macintosh. The changes were from
3174
3175 while ((c = *(++ptr)) != 0 && c != '\n');
3176 to
3177 while ((c = *(++ptr)) != 0 && c != '\n') ;
3178
3179 Totally extraordinary, but if that's what it takes...
3180
3181 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3182 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3183 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3184 assumes the way PCRE uses memmove() (always moving upwards).
3185
3186 6. PCRE is being used in one environment where strchr() is not available. There
3187 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3188 faster code anyway.
3189
3190
3191 Version 3.1 09-Feb-00
3192 ---------------------
3193
3194 The only change in this release is the fixing of some bugs in Makefile.in for
3195 the "install" target:
3196
3197 (1) It was failing to install pcreposix.h.
3198
3199 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3200
3201
3202 Version 3.0 01-Feb-00
3203 ---------------------
3204
3205 1. Add support for the /+ modifier to perltest (to output $` like it does in
3206 pcretest).
3207
3208 2. Add support for the /g modifier to perltest.
3209
3210 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3211 matches null strings.
3212
3213 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3214 pattern. Perl treats empty patterns specially - it reuses the most recent
3215 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3216 effect.
3217
3218 5. The POSIX interface was broken in that it was just handing over the POSIX
3219 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3220 required a bigger vector, with some working space on the end. This means that
3221 the POSIX wrapper now has to get and free some memory, and copy the results.
3222
3223 6. Added some simple autoconf support, placing the test data and the
3224 documentation in separate directories, re-organizing some of the
3225 information files, and making it build pcre-config (a GNU standard). Also added
3226 libtool support for building PCRE as a shared library, which is now the
3227 default.
3228
3229 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3230 09 are not valid octal constants. Single digits will be used for minor values
3231 less than 10.
3232
3233 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3234 existing programs that set these in the POSIX interface can use PCRE without
3235 modification.
3236
3237 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3238 return all that pcre_info() returns, plus additional data. The pcre_info()
3239 function is retained for compatibility, but is considered to be obsolete.
3240
3241 10. Added experimental recursion feature (?R) to handle one common case that
3242 Perl 5.6 will be able to do with (?p{...}).
3243
3244 11. Added support for POSIX character classes like [:alpha:], which Perl is
3245 adopting.
3246
3247
3248 Version 2.08 31-Aug-99
3249 ----------------------
3250
3251 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3252 trying to match at the startoffset position, but instead was moving forward to
3253 the next newline as if a previous match had failed.
3254
3255 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3256 and could get into a loop if a null string was matched other than at the start
3257 of the subject.
3258
3259 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3260 be distinguished at compile time, and for completeness also added PCRE_DATE.
3261
3262 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3263 in GnuWin32 environments.
3264
3265
3266 Version 2.07 29-Jul-99
3267 ----------------------
3268
3269 1. The documentation is now supplied in plain text form and HTML as well as in
3270 the form of man page sources.
3271
3272 2. C++ compilers don't like assigning (void *) values to other pointer types.
3273 In particular this affects malloc(). Although there is no problem in Standard
3274 C, I've put in casts to keep C++ compilers happy.
3275
3276 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3277 should be (const char *).
3278
3279 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3280 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3281 However, I haven't made this a standard facility. The documentation doesn't
3282 mention it, and the Makefile doesn't support it.
3283
3284 5. The Makefile now contains an "install" target, with editable destinations at
3285 the top of the file. The pcretest program is not installed.
3286
3287 6. pgrep -V now gives the PCRE version number and date.
3288
3289 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3290 causing the entire string to be ignored, instead of just the last character.
3291
3292 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3293 non-matching string, it can take a very, very long time, even for strings of
3294 quite modest length, because of the nested recursion. PCRE now does better in
3295 some of these cases. It does this by remembering the last required literal
3296 character in the pattern, and pre-searching the subject to ensure it is present
3297 before running the real match. In other words, it applies a heuristic to detect
3298 some types of certain failure quickly, and in the above example, if presented
3299 with a string that has no trailing " it gives "no match" very quickly.
3300
3301 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3302 other alternatives are tried instead.
3303
3304
3305 Version 2.06 09-Jun-99
3306 ----------------------
3307
3308 1. Change pcretest's output for amount of store used to show just the code
3309 space, because the remainder (the data block) varies in size between 32-bit and
3310 64-bit systems.
3311
3312 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3313 start matching at. This allows lookbehinds to work when searching for multiple
3314 occurrences in a string.
3315
3316 3. Added additional options to pcretest for testing multiple occurrences:
3317
3318 /+ outputs the rest of the string that follows a match
3319 /g loops for multiple occurrences, using the new startoffset argument
3320 /G loops for multiple occurrences by passing an incremented pointer
3321
3322 4. PCRE wasn't doing the "first character" optimization for patterns starting
3323 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3324 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3325 the letter 'x'. On long subject strings, this gives a significant speed-up.
3326
3327
3328 Version 2.05 21-Apr-99
3329 ----------------------
3330
3331 1. Changed the type of magic_number from int to long int so that it works
3332 properly on 16-bit systems.
3333
3334 2. Fixed a bug which caused patterns starting with .* not to work correctly
3335 when the subject string contained newline characters. PCRE was assuming
3336 anchoring for such patterns in all cases, which is not correct because .* will
3337 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3338 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3339 must be retried after every newline in the subject.
3340
3341
3342 Version 2.04 18-Feb-99
3343 ----------------------
3344
3345 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3346 computation of the store needed to hold the pattern was incorrect (too large).
3347 If such patterns were nested a few deep, this could multiply and become a real
3348 problem.
3349
3350 2. Added /M option to pcretest to show the memory requirement of a specific
3351 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3352
3353 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3354 compiled in such a way that the backtracking after subsequent failure was
3355 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3356 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3357
3358
3359 Version 2.03 02-Feb-99
3360 ----------------------
3361
3362 1. Fixed typo and small mistake in man page.
3363
3364 2. Added 4th condition (GPL supersedes if conflict) and created separate
3365 LICENCE file containing the conditions.
3366
3367 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3368 Perl, that is the internal \ allows the delimiter to be included in the
3369 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3370 the final delimiter, add \ to the end of the pattern (to test the error).
3371
3372 4. Added the convenience functions for extracting substrings after a successful
3373 match. Updated pcretest to make it able to test these functions.
3374
3375
3376 Version 2.02 14-Jan-99
3377 ----------------------
3378
3379 1. Initialized the working variables associated with each extraction so that
3380 their saving and restoring doesn't refer to uninitialized store.
3381
3382 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3383 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3384 fix the problem.
3385
3386 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3387 calls, and wasn't printing the correct value for compiling calls. Increased the
3388 default value of LOOPREPEAT, and the number of significant figures in the
3389 times.
3390
3391 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3392
3393 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3394 a building problem on Windows NT with a FAT file system.
3395
3396
3397 Version 2.01 21-Oct-98
3398 ----------------------
3399
3400 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3401 to character tables built by pcre_maketables() in the current locale. If NULL
3402 is passed, the default tables are used.
3403
3404
3405 Version 2.00 24-Sep-98
3406 ----------------------
3407
3408 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3409 it any more.
3410
3411 2. Allow quantification of (?>) groups, and make it work correctly.
3412
3413 3. The first character computation wasn't working for (?>) groups.
3414
3415 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3416 end of the subject) and add 5.005's \z, which really does match only at the
3417 very end of the subject.
3418
3419 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3420
3421 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3422 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3423 localized options. All options to pcre_study() were also removed.
3424
3425 7. Add other new features from 5.005:
3426
3427 $(?<= positive lookbehind
3428 $(?<! negative lookbehind
3429 (?imsx-imsx) added the unsetting capability
3430 such a setting is global if at outer level; local otherwise
3431 (?imsx-imsx:) non-capturing groups with option setting
3432 (?(cond)re|re) conditional pattern matching
3433
3434 A backreference to itself in a repeated group matches the previous
3435 captured string.
3436
3437 8. General tidying up of studying (both automatic and via "study")
3438 consequential on the addition of new assertions.
3439
3440 9. As in 5.005, unlimited repeated groups that could match an empty substring
3441 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3442 runtime if any iteration does actually match an empty substring.
3443
3444 10. Include the RunTest script in the distribution.
3445
3446 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3447 discrepancies, some of which were old and were also with respect to 5.004. They
3448 have now been fixed.
3449
3450
3451 Version 1.09 28-Apr-98
3452 ----------------------
3453
3454 1. A negated single character class followed by a quantifier with a minimum
3455 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3456 program crashes, or just wrong answers. This did not apply to negated classes
3457 containing more than one character, or to minima other than one.
3458
3459
3460 Version 1.08 27-Mar-98
3461 ----------------------
3462
3463 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3464
3465 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3466 latter must appear before anything that relies on it in the pattern.
3467
3468
3469 Version 1.07 16-Feb-98
3470 ----------------------
3471
3472 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3473 repeat of a potentially empty string).
3474
3475
3476 Version 1.06 23-Jan-98
3477 ----------------------
3478
3479 1. Added Markus Oberhumer's little patches for C++.
3480
3481 2. Literal strings longer than 255 characters were broken.
3482
3483
3484 Version 1.05 23-Dec-97
3485 ----------------------
3486
3487 1. Negated character classes containing more than one character were failing if
3488 PCRE_CASELESS was set at run time.
3489
3490
3491 Version 1.04 19-Dec-97
3492 ----------------------
3493
3494 1. Corrected the man page, where some "const" qualifiers had been omitted.
3495
3496 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3497 input syntax.
3498
3499 3. Fixed memory leak which occurred when a regex with back references was
3500 matched with an offsets vector that wasn't big enough. The temporary memory
3501 that is used in this case wasn't being freed if the match failed.
3502
3503 4. Tidied pcretest to ensure it frees memory that it gets.
3504
3505 5. Temporary memory was being obtained in the case where the passed offsets
3506 vector was exactly big enough.
3507
3508 6. Corrected definition of offsetof() from change 5 below.
3509
3510 7. I had screwed up change 6 below and broken the rules for the use of
3511 setjmp(). Now fixed.
3512
3513
3514 Version 1.03 18-Dec-97
3515 ----------------------
3516
3517 1. A erroneous regex with a missing opening parenthesis was correctly
3518 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3519 on some systems.
3520
3521 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3522 it was reported that one broken compiler failed on the former because "code" is
3523 also an independent variable.
3524
3525 3. The erroneous regex a[]b caused an array overrun reference.
3526
3527 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3528 fail on data ending with that character. (It was going on too far, and checking
3529 the next character, typically a binary zero.) This was specific to the
3530 optimized code for single-character negative classes.
3531
3532 5. Added a contributed patch from the TIN world which does the following:
3533
3534 + Add an undef for memmove, in case the the system defines a macro for it.
3535
3536 + Add a definition of offsetof(), in case there isn't one. (I don't know
3537 the reason behind this - offsetof() is part of the ANSI standard - but
3538 it does no harm).
3539
3540 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3541 most of the places where whitespace preceded '#'. I have given up and
3542 allowed the remaining 2 cases to be at the margin.
3543
3544 + Rename some variables in pcre to eliminate shadowing. This seems very
3545 pedantic, but does no harm, of course.
3546
3547 6. Moved the call to setjmp() into its own function, to get rid of warnings
3548 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3549
3550 7. Constructs such as \d{8,} were compiling into the equivalent of
3551 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3552 outcome, but in this particular case used more store than had been allocated,
3553 which caused the bug to be discovered because it threw up an internal error.
3554
3555 8. The debugging code in both pcre and pcretest for outputting the compiled
3556 form of a regex was going wrong in the case of back references followed by
3557 curly-bracketed repeats.
3558
3559
3560 Version 1.02 12-Dec-97
3561 ----------------------
3562
3563 1. Typos in pcre.3 and comments in the source fixed.
3564
3565 2. Applied a contributed patch to get rid of places where it used to remove
3566 'const' from variables, and fixed some signed/unsigned and uninitialized
3567 variable warnings.
3568
3569 3. Added the "runtest" target to Makefile.
3570
3571 4. Set default compiler flag to -O2 rather than just -O.
3572
3573
3574 Version 1.01 19-Nov-97
3575 ----------------------
3576
3577 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3578 like /([ab]*)*/, that is, for classes with more than one character in them.
3579
3580 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3581 as /((?>a*))*/ (a PCRE_EXTRA facility).
3582
3583
3584 Version 1.00 18-Nov-97
3585 ----------------------
3586
3587 1. Added compile-time macros to support systems such as SunOS4 which don't have
3588 memmove() or strerror() but have other things that can be used instead.
3589
3590 2. Arranged that "make clean" removes the executables.
3591
3592
3593 Version 0.99 27-Oct-97
3594 ----------------------
3595
3596 1. Fixed bug in code for optimizing classes with only one character. It was
3597 initializing a 32-byte map regardless, which could cause it to run off the end
3598 of the memory it had got.
3599
3600 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3601
3602
3603 Version 0.98 22-Oct-97
3604 ----------------------
3605
3606 1. Fixed bug in code for handling temporary memory usage when there are more
3607 back references than supplied space in the ovector. This could cause segfaults.
3608
3609
3610 Version 0.97 21-Oct-97
3611 ----------------------
3612
3613 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3614
3615 2. Optimized negated single characters not to use a bit map.
3616
3617 3. Brought error texts together as macro definitions; clarified some of them;
3618 fixed one that was wrong - it said "range out of order" when it meant "invalid
3619 escape sequence".
3620
3621 4. Changed some char * arguments to const char *.
3622
3623 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3624
3625 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3626 pcretest.
3627
3628
3629 Version 0.96 16-Oct-97
3630 ----------------------
3631
3632 1. Added a simple "pgrep" utility to the distribution.
3633
3634 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3635 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3636 where "ddd" means "one or more decimal digits".
3637
3638 3. Fixed serious bug. If a pattern had a back reference, but the call to
3639 pcre_exec() didn't supply a large enough ovector to record the related
3640 identifying subpattern, the match always failed. PCRE now remembers the number
3641 of the largest back reference, and gets some temporary memory in which to save
3642 the offsets during matching if necessary, in order to ensure that
3643 backreferences always work.
3644
3645 4. Increased the compatibility with Perl in a number of ways:
3646
3647 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3648 to request this handling. The option can be set at compile or exec time.
3649
3650 (b) $ matches before a terminating newline by default; an option
3651 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3652 mode). The option can be set at compile or exec time.
3653
3654 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3655 the same as Perl's. If the decimal number it represents is less than 10
3656 or there aren't that many previous left capturing parentheses, an octal
3657 escape is read. Inside a character class, it's always an octal escape,
3658 even if it is a single digit.
3659
3660 (d) An escaped but undefined alphabetic character is taken as a literal,
3661 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3662 escapes.
3663
3664 (e) {0} is now permitted. (The previous item is removed from the compiled
3665 pattern).
3666
3667 5. Changed all the names of code files so that the basic parts are no longer
3668 than 10 characters, and abolished the teeny "globals.c" file.
3669
3670 6. Changed the handling of character classes; they are now done with a 32-byte
3671 bit map always.
3672
3673 7. Added the -d and /D options to pcretest to make it possible to look at the
3674 internals of compilation without having to recompile pcre.
3675
3676
3677 Version 0.95 23-Sep-97
3678 ----------------------
3679
3680 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3681 \x20 at the start of a run of normal characters. These were being treated as
3682 real characters, instead of the source characters being re-checked.
3683
3684
3685 Version 0.94 18-Sep-97
3686 ----------------------
3687
3688 1. The functions are now thread-safe, with the caveat that the global variables
3689 containing pointers to malloc() and free() or alternative functions are the
3690 same for all threads.
3691
3692 2. Get pcre_study() to generate a bitmap of initial characters for non-
3693 anchored patterns when this is possible, and use it if passed to pcre_exec().
3694
3695
3696 Version 0.93 15-Sep-97
3697 ----------------------
3698
3699 1. /(b)|(:+)/ was computing an incorrect first character.
3700
3701 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3702 but not actually doing anything yet.
3703
3704 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3705 as Perl does (e.g. [-az] or [az-]).
3706
3707 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3708 all possible positions.
3709
3710 5. Split up into different modules to avoid including unneeded functions in a
3711 compiled binary. However, compile and exec are still in one module. The "study"
3712 function is split off.
3713
3714 6. The character tables are now in a separate module whose source is generated
3715 by an auxiliary program - but can then be edited by hand if required. There are
3716 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3717 toupper() in the code.
3718
3719 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3720 make them global. Abolish the function for setting them, as the caller can now
3721 set them directly.
3722
3723
3724 Version 0.92 11-Sep-97
3725 ----------------------
3726
3727 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3728 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3729
3730 2. Caseless matching was not working in character classes if the characters in
3731 the pattern were in upper case.
3732
3733 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3734
3735 4. Make PCRE_ANCHORED public and accept as a compile option.
3736
3737 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3738 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3739 pass them.
3740
3741 6. Give an error if bad option bits passed at compile or run time.
3742
3743 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3744 pcretest to cause it to pass that flag.
3745
3746 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3747 options, and the first character, if set.
3748
3749 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3750
3751
3752 Version 0.91 10-Sep-97
3753 ----------------------
3754
3755 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3756 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3757
3758 2. PCRE was looping on encountering an indefinitely repeated back reference to
3759 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3760 Perl does - treats the match as successful.
3761
3762 ****

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12