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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 461 - (show annotations) (download)
Mon Oct 5 10:59:35 2009 UTC (4 years, 9 months ago) by ph10
File size: 174821 byte(s)
Tidy up, remove trailing spaces, etc. for 8.00-RC1.

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8 was out-of-date, and there was no check on the pcre_compile() error code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

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