/[pcre]/code/tags/pcre-8.01/ChangeLog
ViewVC logotype

Contents of /code/tags/pcre-8.01/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 469 - (show annotations) (download)
Mon Oct 19 14:38:48 2009 UTC (4 years, 11 months ago) by ph10
Original Path: code/trunk/ChangeLog
File size: 174788 byte(s)
Final doc and source tidies for 8.00

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8 was out-of-date, and there was no check on the pcre_compile() error code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

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