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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 473 - (show annotations) (download)
Sat Jan 2 12:40:07 2010 UTC (4 years, 11 months ago) by ph10
File size: 176103 byte(s)
Fix bugs relating to the use of (*SKIP) etc in assertions.

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35 was out-of-date, and there was no check on the pcre_compile() error code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

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