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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 475 - (show annotations) (download)
Sat Jan 2 18:21:30 2010 UTC (4 years, 8 months ago) by ph10
File size: 177419 byte(s)
Tidies to allow easier embedded compilation; avoid (double) where possible.

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

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