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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 484 - (show annotations) (download)
Mon Jan 4 17:11:54 2010 UTC (4 years, 10 months ago) by ph10
File size: 180124 byte(s)
Update autogen.sh to try several names for libtoolize.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.01 11-Dec-09
5 ----------------------
6
7 1. If a pattern contained a conditional subpattern with only one branch (in
8 particular, this includes all (DEFINE) patterns), a call to pcre_study()
9 computed the wrong minimum data length (which is of course zero for such
10 subpatterns).
11
12 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
13 the pattern is reset in the first branch, pcre_compile() failed with
14 "internal error: code overflow at offset...". This happened only when
15 the reset was to the original external option setting. (An optimization
16 abstracts leading options settings into an external setting, which was the
17 cause of this.)
18
19 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
20 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
21 assertion pattern did not match (meaning that the assertion was true), it
22 was incorrectly treated as false if the SKIP had been reached during the
23 matching. This also applied to assertions used as conditions.
24
25 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
26 assertion subpattern, including such a pattern used as a condition,
27 unpredictable results occurred, instead of the error return
28 PCRE_ERROR_DFA_UITEM.
29
30 5. The C++ GlobalReplace function was not working like Perl for the special
31 situation when an empty string is matched. It now does the fancy magic
32 stuff that is necessary.
33
34 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
35 removed. (These were left over from very, very early versions of PCRE.)
36
37 7. Some cosmetic changes to the code to make life easier when compiling it
38 as part of something else:
39
40 (a) Change DEBUG to PCRE_DEBUG.
41
42 (b) In pcre_compile(), rename the member of the "branch_chain" structure
43 called "current" as "current_branch", to prevent a collision with the
44 Linux macro when compiled as a kernel module.
45
46 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
47 prevent a collision with the Linux macro when compiled as a kernel
48 module.
49
50 8. In pcre_compile() there are some checks for integer overflows that used to
51 cast potentially large values to (double). This has been changed to that
52 when building, a check for int64_t is made, and if it is found, it is used
53 instead, thus avoiding the use of floating point arithmetic. (There is no
54 other use of FP in PCRE.) If int64_t is not found, the fallback is to
55 double.
56
57 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
58 2005 (difference between two addresses compared to an unsigned value).
59
60 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
61 custom one, because of the following reported problem in Windows:
62
63 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
64 under Win32.
65 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
66 therefore missing the function definition.
67 - The compiler thus generates a "C" signature for the test function.
68 - The linker fails to find the "C" function.
69 - PCRE fails to configure if asked to do so against libbz2.
70
71 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
72 messages were output:
73
74 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
75 rerunning libtoolize, to keep the correct libtool macros in-tree.
76 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
77
78 I have done both of these things.
79
80 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
81 most of the time, it *can* run out if it is given a pattern that contains a
82 runaway infinite recursion. I updated the discussion in the pcrestack man
83 page.
84
85 13. Now that we have gone to the x.xx style of version numbers, the minor
86 version may start with zero. Using 08 or 09 is a bad idea because users
87 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
88 interpreted as invalid octal numbers. I've updated the previous comment in
89 configure.ac, and also added a check that gives an error if 08 or 09 are
90 used.
91
92 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
93 causing partial matching to fail when the end of the subject matched \W
94 in a UTF-8 pattern where \W was quantified with a minimum of 3.
95
96 15. There were some discrepancies between the declarations in pcre_internal.h
97 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
98 their definitions. The declarations used "const uschar *" and the
99 definitions used USPTR. Even though USPTR is normally defined as "const
100 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
101 reported that: "This difference in casting confuses some C++ compilers, for
102 example, SunCC recognizes above declarations as different functions and
103 generates broken code for hbpcre." I have changed the declarations to use
104 USPTR.
105
106 16. GNU libtool is named differently on some systems. The autogen.sh script now
107 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
108 (FreeBSD).
109
110
111
112 Version 8.00 19-Oct-09
113 ----------------------
114
115 1. The table for translating pcre_compile() error codes into POSIX error codes
116 was out-of-date, and there was no check on the pcre_compile() error code
117 being within the table. This could lead to an OK return being given in
118 error.
119
120 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
121 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
122 in a Windows environment.
123
124 3. The pcregrep --count option prints the count for each file even when it is
125 zero, as does GNU grep. However, pcregrep was also printing all files when
126 --files-with-matches was added. Now, when both options are given, it prints
127 counts only for those files that have at least one match. (GNU grep just
128 prints the file name in this circumstance, but including the count seems
129 more useful - otherwise, why use --count?) Also ensured that the
130 combination -clh just lists non-zero counts, with no names.
131
132 4. The long form of the pcregrep -F option was incorrectly implemented as
133 --fixed_strings instead of --fixed-strings. This is an incompatible change,
134 but it seems right to fix it, and I didn't think it was worth preserving
135 the old behaviour.
136
137 5. The command line items --regex=pattern and --regexp=pattern were not
138 recognized by pcregrep, which required --regex pattern or --regexp pattern
139 (with a space rather than an '='). The man page documented the '=' forms,
140 which are compatible with GNU grep; these now work.
141
142 6. No libpcreposix.pc file was created for pkg-config; there was just
143 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
144
145 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
146 when UCP support is not needed, by modifying the Python script that
147 generates it from Unicode data files. This should not matter if the module
148 is correctly used as a library, but I received one complaint about 50K of
149 unwanted data. My guess is that the person linked everything into his
150 program rather than using a library. Anyway, it does no harm.
151
152 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
153 was a minimum greater than 1 for a wide character in a possessive
154 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
155 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
156 character. Chaos in the form of incorrect output or a compiling loop could
157 result.
158
159 9. The restrictions on what a pattern can contain when partial matching is
160 requested for pcre_exec() have been removed. All patterns can now be
161 partially matched by this function. In addition, if there are at least two
162 slots in the offset vector, the offset of the earliest inspected character
163 for the match and the offset of the end of the subject are set in them when
164 PCRE_ERROR_PARTIAL is returned.
165
166 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
167 synonymous with PCRE_PARTIAL, for backwards compatibility, and
168 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
169 and may be more useful for multi-segment matching.
170
171 11. Partial matching with pcre_exec() is now more intuitive. A partial match
172 used to be given if ever the end of the subject was reached; now it is
173 given only if matching could not proceed because another character was
174 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
175 string "Z", which now yields "no match" instead of "partial match". In the
176 case of pcre_dfa_exec(), "no match" is given if every matching path for the
177 final character ended with (*FAIL).
178
179 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
180 if the pattern had a "must contain" character that was already found in the
181 earlier partial match, unless partial matching was again requested. For
182 example, with the pattern /dog.(body)?/, the "must contain" character is
183 "g". If the first part-match was for the string "dog", restarting with
184 "sbody" failed. This bug has been fixed.
185
186 13. The string returned by pcre_dfa_exec() after a partial match has been
187 changed so that it starts at the first inspected character rather than the
188 first character of the match. This makes a difference only if the pattern
189 starts with a lookbehind assertion or \b or \B (\K is not supported by
190 pcre_dfa_exec()). It's an incompatible change, but it makes the two
191 matching functions compatible, and I think it's the right thing to do.
192
193 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
194 so that the demonstration program is easily available in environments where
195 PCRE has not been installed from source.
196
197 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
198 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
199 library.
200
201 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
202 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
203 is not the first non-POSIX option to be added. Clearly some people find
204 these options useful.
205
206 17. If a caller to the POSIX matching function regexec() passes a non-zero
207 value for nmatch with a NULL value for pmatch, the value of
208 nmatch is forced to zero.
209
210 18. RunGrepTest did not have a test for the availability of the -u option of
211 the diff command, as RunTest does. It now checks in the same way as
212 RunTest, and also checks for the -b option.
213
214 19. If an odd number of negated classes containing just a single character
215 interposed, within parentheses, between a forward reference to a named
216 subpattern and the definition of the subpattern, compilation crashed with
217 an internal error, complaining that it could not find the referenced
218 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
219 [The bug was that it was starting one character too far in when skipping
220 over the character class, thus treating the ] as data rather than
221 terminating the class. This meant it could skip too much.]
222
223 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
224 /g option in pcretest when the pattern contains \K, which makes it possible
225 to have an empty string match not at the start, even when the pattern is
226 anchored. Updated pcretest and pcredemo to use this option.
227
228 21. If the maximum number of capturing subpatterns in a recursion was greater
229 than the maximum at the outer level, the higher number was returned, but
230 with unset values at the outer level. The correct (outer level) value is
231 now given.
232
233 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
234 PCRE did not set those parentheses (unlike Perl). I have now found a way to
235 make it do so. The string so far is captured, making this feature
236 compatible with Perl.
237
238 23. The tests have been re-organized, adding tests 11 and 12, to make it
239 possible to check the Perl 5.10 features against Perl 5.10.
240
241 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
242 pattern matches a fixed length string. PCRE did not allow this; now it
243 does. Neither allows recursion.
244
245 25. I finally figured out how to implement a request to provide the minimum
246 length of subject string that was needed in order to match a given pattern.
247 (It was back references and recursion that I had previously got hung up
248 on.) This code has now been added to pcre_study(); it finds a lower bound
249 to the length of subject needed. It is not necessarily the greatest lower
250 bound, but using it to avoid searching strings that are too short does give
251 some useful speed-ups. The value is available to calling programs via
252 pcre_fullinfo().
253
254 26. While implementing 25, I discovered to my embarrassment that pcretest had
255 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
256 study optimizations had never been tested with that matching function.
257 Oops. What is worse, even when it was passed study data, there was a bug in
258 pcre_dfa_exec() that meant it never actually used it. Double oops. There
259 were also very few tests of studied patterns with pcre_dfa_exec().
260
261 27. If (?| is used to create subpatterns with duplicate numbers, they are now
262 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
263 on the other side of the coin, they are no longer allowed to have different
264 names, because these cannot be distinguished in PCRE, and this has caused
265 confusion. (This is a difference from Perl.)
266
267 28. When duplicate subpattern names are present (necessarily with different
268 numbers, as required by 27 above), and a test is made by name in a
269 conditional pattern, either for a subpattern having been matched, or for
270 recursion in such a pattern, all the associated numbered subpatterns are
271 tested, and the overall condition is true if the condition is true for any
272 one of them. This is the way Perl works, and is also more like the way
273 testing by number works.
274
275
276 Version 7.9 11-Apr-09
277 ---------------------
278
279 1. When building with support for bzlib/zlib (pcregrep) and/or readline
280 (pcretest), all targets were linked against these libraries. This included
281 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
282 libraries. This caused unwanted dependencies to be created. This problem
283 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
284 pcretest is linked with readline.
285
286 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
287 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
288 moved outside it again, because FALSE and TRUE are already defined in AIX,
289 but BOOL is not.
290
291 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
292 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
293
294 4. The pcregrep documentation said spaces were inserted as well as colons (or
295 hyphens) following file names and line numbers when outputting matching
296 lines. This is not true; no spaces are inserted. I have also clarified the
297 wording for the --colour (or --color) option.
298
299 5. In pcregrep, when --colour was used with -o, the list of matching strings
300 was not coloured; this is different to GNU grep, so I have changed it to be
301 the same.
302
303 6. When --colo(u)r was used in pcregrep, only the first matching substring in
304 each matching line was coloured. Now it goes on to look for further matches
305 of any of the test patterns, which is the same behaviour as GNU grep.
306
307 7. A pattern that could match an empty string could cause pcregrep to loop; it
308 doesn't make sense to accept an empty string match in pcregrep, so I have
309 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
310 seems to be how GNU grep behaves.
311
312 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
313 start or after a newline", because the conditional assertion was not being
314 correctly handled. The rule now is that both the assertion and what follows
315 in the first alternative must satisfy the test.
316
317 9. If auto-callout was enabled in a pattern with a conditional group whose
318 condition was an assertion, PCRE could crash during matching, both with
319 pcre_exec() and pcre_dfa_exec().
320
321 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
322 used for matching.
323
324 11. Unicode property support in character classes was not working for
325 characters (bytes) greater than 127 when not in UTF-8 mode.
326
327 12. Added the -M command line option to pcretest.
328
329 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
330
331 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
332
333 16. Added comments and documentation about mis-use of no_arg in the C++
334 wrapper.
335
336 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
337 from Martin Jerabek that uses macro names for all relevant character and
338 string constants.
339
340 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
341 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
342 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
343 these, but not everybody uses configure.
344
345 19. A conditional group that had only one branch was not being correctly
346 recognized as an item that could match an empty string. This meant that an
347 enclosing group might also not be so recognized, causing infinite looping
348 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
349 with the subject "ab", where knowledge that the repeated group can match
350 nothing is needed in order to break the loop.
351
352 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
353 exec(), but without supplying a callout function, matching went wrong.
354
355 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
356 leak if the size of the offset vector was greater than 30. When the vector
357 is smaller, the saved offsets during recursion go onto a local stack
358 vector, but for larger vectors malloc() is used. It was failing to free
359 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
360 error, in fact).
361
362 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
363 heapframe that is used only when UTF-8 support is enabled. This caused no
364 problem, but was untidy.
365
366 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
367 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
368 included within another project.
369
370 24. Steven Van Ingelgem's patches to add more options to the CMake support,
371 slightly modified by me:
372
373 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
374 not building pcregrep.
375
376 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
377 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
378
379 25. Forward references, both numeric and by name, in patterns that made use of
380 duplicate group numbers, could behave incorrectly or give incorrect errors,
381 because when scanning forward to find the reference group, PCRE was not
382 taking into account the duplicate group numbers. A pattern such as
383 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
384
385 26. Changed a few more instances of "const unsigned char *" to USPTR, making
386 the feature of a custom pointer more persuasive (as requested by a user).
387
388 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
389 pcretest.c, inside #ifndefs, because it seems they are sometimes already
390 pre-defined.
391
392 28. Added support for (*UTF8) at the start of a pattern.
393
394 29. Arrange for flags added by the "release type" setting in CMake to be shown
395 in the configuration summary.
396
397
398 Version 7.8 05-Sep-08
399 ---------------------
400
401 1. Replaced UCP searching code with optimized version as implemented for Ad
402 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
403 stage table and inline lookup instead of a function, giving speed ups of 2
404 to 5 times on some simple patterns that I tested. Permission was given to
405 distribute the MultiStage2.py script that generates the tables (it's not in
406 the tarball, but is in the Subversion repository).
407
408 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
409 scripts.
410
411 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
412 a group with a zero qualifier. The result of the study could be incorrect,
413 or the function might crash, depending on the pattern.
414
415 4. Caseless matching was not working for non-ASCII characters in back
416 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
417 It now works when Unicode Property Support is available.
418
419 5. In pcretest, an escape such as \x{de} in the data was always generating
420 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
421 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
422 truncation.
423
424 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
425
426 7. Added two (int) casts to pcregrep when printing the difference of two
427 pointers, in case they are 64-bit values.
428
429 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
430 test 2 if it fails.
431
432 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
433 and a #define of that name to empty if it is not externally set. This is to
434 allow users of MSVC to set it if necessary.
435
436 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
437 the convenience functions in the pcre_get.c source file.
438
439 11. An option change at the start of a pattern that had top-level alternatives
440 could cause overwriting and/or a crash. This command provoked a crash in
441 some environments:
442
443 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
444
445 This potential security problem was recorded as CVE-2008-2371.
446
447 12. For a pattern where the match had to start at the beginning or immediately
448 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
449 pcre_dfa_exec() could read past the end of the passed subject if there was
450 no match. To help with detecting such bugs (e.g. with valgrind), I modified
451 pcretest so that it places the subject at the end of its malloc-ed buffer.
452
453 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
454 exec() might read past the end of the data buffer in UTF-8 mode.
455
456 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
457 the data contained the byte 0x85 as part of a UTF-8 character within its
458 first line. This applied both to normal and DFA matching.
459
460 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
461 /^[^d]*?$/8 failed to match "abc".
462
463 16. Added a missing copyright notice to pcrecpp_internal.h.
464
465 17. Make it more clear in the documentation that values returned from
466 pcre_exec() in ovector are byte offsets, not character counts.
467
468 18. Tidied a few places to stop certain compilers from issuing warnings.
469
470 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
471 supplied by Stefan Weber. I made a further small update for 7.8 because
472 there is a change of source arrangements: the pcre_searchfuncs.c module is
473 replaced by pcre_ucd.c.
474
475
476 Version 7.7 07-May-08
477 ---------------------
478
479 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
480 a string to a long long, pretend we don't even have a long long." This is
481 done by checking for the strtoq, strtoll, and _strtoi64 functions.
482
483 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
484 pre-7.6 versions, which defined a global no_arg variable instead of putting
485 it in the RE class. (See also #8 below.)
486
487 3. Remove a line of dead code, identified by coverity and reported by Nuno
488 Lopes.
489
490 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
491
492 (1) The include/exclude patterns were being applied to the whole pathnames
493 of files, instead of just to the final components.
494
495 (2) If there was more than one level of directory, the subdirectories were
496 skipped unless they satisfied the include/exclude conditions. This is
497 inconsistent with GNU grep (and could even be seen as contrary to the
498 pcregrep specification - which I improved to make it absolutely clear).
499 The action now is always to scan all levels of directory, and just
500 apply the include/exclude patterns to regular files.
501
502 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
503 --exclude_dir in the tests to avoid scanning .svn directories.
504
505 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
506 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
507 doesn't support NULs in patterns.
508
509 7. Added some missing "const"s to declarations of static tables in
510 pcre_compile.c and pcre_dfa_exec.c.
511
512 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
513 caused by fix #2 above. (Subsequently also a second patch to fix the
514 first patch. And a third patch - this was a messy problem.)
515
516 9. Applied Craig's patch to remove the use of push_back().
517
518 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
519 matching function regexec().
520
521 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
522 which, however, unlike Perl's \g{...}, are subroutine calls, not back
523 references. PCRE supports relative numbers with this syntax (I don't think
524 Oniguruma does).
525
526 12. Previously, a group with a zero repeat such as (...){0} was completely
527 omitted from the compiled regex. However, this means that if the group
528 was called as a subroutine from elsewhere in the pattern, things went wrong
529 (an internal error was given). Such groups are now left in the compiled
530 pattern, with a new opcode that causes them to be skipped at execution
531 time.
532
533 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
534 to the way PCRE behaves:
535
536 (a) A lone ] character is dis-allowed (Perl treats it as data).
537
538 (b) A back reference to an unmatched subpattern matches an empty string
539 (Perl fails the current match path).
540
541 (c) A data ] in a character class must be notated as \] because if the
542 first data character in a class is ], it defines an empty class. (In
543 Perl it is not possible to have an empty class.) The empty class []
544 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
545 The negative empty class [^] matches any one character, independently
546 of the DOTALL setting.
547
548 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
549 non-existent subpattern following a character class starting with ']' and
550 containing () gave an internal compiling error instead of "reference to
551 non-existent subpattern". Fortunately, when the pattern did exist, the
552 compiled code was correct. (When scanning forwards to check for the
553 existencd of the subpattern, it was treating the data ']' as terminating
554 the class, so got the count wrong. When actually compiling, the reference
555 was subsequently set up correctly.)
556
557 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
558 it was being rejected as not supported by pcre_dfa_exec(), even though
559 other assertions are supported. I have made pcre_dfa_exec() support
560 (*FAIL).
561
562 16. The implementation of 13c above involved the invention of a new opcode,
563 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
564 cannot be changed at match time, I realized I could make a small
565 improvement to matching performance by compiling OP_ALLANY instead of
566 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
567 on the OP_ANY path.
568
569 17. Compiling pcretest on Windows with readline support failed without the
570 following two fixes: (1) Make the unistd.h include conditional on
571 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
572
573 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
574 ncurses library to be included for pcretest when ReadLine support is
575 requested, but also to allow for it to be overridden. This patch came from
576 Daniel Bergström.
577
578 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
579 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
580 any errors with the current Unicode tables. Thanks to Peter Kankowski for
581 spotting this.
582
583
584 Version 7.6 28-Jan-08
585 ---------------------
586
587 1. A character class containing a very large number of characters with
588 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
589 overflow.
590
591 2. Patch to cut out the "long long" test in pcrecpp_unittest when
592 HAVE_LONG_LONG is not defined.
593
594 3. Applied Christian Ehrlicher's patch to update the CMake build files to
595 bring them up to date and include new features. This patch includes:
596
597 - Fixed PH's badly added libz and libbz2 support.
598 - Fixed a problem with static linking.
599 - Added pcredemo. [But later removed - see 7 below.]
600 - Fixed dftables problem and added an option.
601 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
602 HAVE_LONG_LONG.
603 - Added readline support for pcretest.
604 - Added an listing of the option settings after cmake has run.
605
606 4. A user submitted a patch to Makefile that makes it easy to create
607 "pcre.dll" under mingw when using Configure/Make. I added stuff to
608 Makefile.am that cause it to include this special target, without
609 affecting anything else. Note that the same mingw target plus all
610 the other distribution libraries and programs are now supported
611 when configuring with CMake (see 6 below) instead of with
612 Configure/Make.
613
614 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
615 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
616 exported in the Windows port". It has not yet been confirmed that the patch
617 solves the problem, but it does no harm.
618
619 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
620 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
621 with CMake, and also correct the comment about stack recursion.
622
623 7. Remove the automatic building of pcredemo from the ./configure system and
624 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
625 of a program that users should build themselves after PCRE is installed, so
626 building it automatically is not really right. What is more, it gave
627 trouble in some build environments.
628
629 8. Further tidies to CMakeLists.txt from Sheri and Christian.
630
631
632 Version 7.5 10-Jan-08
633 ---------------------
634
635 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
636 values in parens when parsing an RE using the C++ wrapper."
637
638 2. Negative specials like \S did not work in character classes in UTF-8 mode.
639 Characters greater than 255 were excluded from the class instead of being
640 included.
641
642 3. The same bug as (2) above applied to negated POSIX classes such as
643 [:^space:].
644
645 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
646 defined or documented. It seems to have been a typo for PCRE_STATIC, so
647 I have changed it.
648
649 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
650 first named subpattern) and a construct such as (?&a) would reference the
651 first named subpattern whose name started with "a" (in other words, the
652 length check was missing). Both these problems are fixed. "Subpattern name
653 expected" is now given for (?&) (a zero-length name), and this patch also
654 makes it give the same error for \k'' (previously it complained that that
655 was a reference to a non-existent subpattern).
656
657 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
658 this is right because (?- can be followed by option settings as well as by
659 digits. I have, however, made the messages clearer.
660
661 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
662 than the number used in the conditional) now cause a compile-time error.
663 This is actually not compatible with Perl, which accepts such patterns, but
664 treats the conditional as always being FALSE (as PCRE used to), but it
665 seems to me that giving a diagnostic is better.
666
667 8. Change "alphameric" to the more common word "alphanumeric" in comments
668 and messages.
669
670 9. Fix two occurrences of "backslash" in comments that should have been
671 "backspace".
672
673 10. Remove two redundant lines of code that can never be obeyed (their function
674 was moved elsewhere).
675
676 11. The program that makes PCRE's Unicode character property table had a bug
677 which caused it to generate incorrect table entries for sequences of
678 characters that have the same character type, but are in different scripts.
679 It amalgamated them into a single range, with the script of the first of
680 them. In other words, some characters were in the wrong script. There were
681 thirteen such cases, affecting characters in the following ranges:
682
683 U+002b0 - U+002c1
684 U+0060c - U+0060d
685 U+0061e - U+00612
686 U+0064b - U+0065e
687 U+0074d - U+0076d
688 U+01800 - U+01805
689 U+01d00 - U+01d77
690 U+01d9b - U+01dbf
691 U+0200b - U+0200f
692 U+030fc - U+030fe
693 U+03260 - U+0327f
694 U+0fb46 - U+0fbb1
695 U+10450 - U+1049d
696
697 12. The -o option (show only the matching part of a line) for pcregrep was not
698 compatible with GNU grep in that, if there was more than one match in a
699 line, it showed only the first of them. It now behaves in the same way as
700 GNU grep.
701
702 13. If the -o and -v options were combined for pcregrep, it printed a blank
703 line for every non-matching line. GNU grep prints nothing, and pcregrep now
704 does the same. The return code can be used to tell if there were any
705 non-matching lines.
706
707 14. Added --file-offsets and --line-offsets to pcregrep.
708
709 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
710 infinitely looping recursion. The bug was that positive lookaheads were not
711 being skipped when checking for a possible empty match (negative lookaheads
712 and both kinds of lookbehind were skipped).
713
714 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
715 inclusion of <windows.h> to before rather than after the definition of
716 INVALID_FILE_ATTRIBUTES (patch from David Byron).
717
718 17. Specifying a possessive quantifier with a specific limit for a Unicode
719 character property caused pcre_compile() to compile bad code, which led at
720 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
721 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
722 caused the error; without that there was no problem.
723
724 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
725
726 19. Added --enable-pcretest-libreadline.
727
728 20. In pcrecpp.cc, the variable 'count' was incremented twice in
729 RE::GlobalReplace(). As a result, the number of replacements returned was
730 double what it should be. I removed one of the increments, but Craig sent a
731 later patch that removed the other one (the right fix) and added unit tests
732 that check the return values (which was not done before).
733
734 21. Several CMake things:
735
736 (1) Arranged that, when cmake is used on Unix, the libraries end up with
737 the names libpcre and libpcreposix, not just pcre and pcreposix.
738
739 (2) The above change means that pcretest and pcregrep are now correctly
740 linked with the newly-built libraries, not previously installed ones.
741
742 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
743
744 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
745 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
746 UTF-8 newline character). The key issue is that the pattern starts .*;
747 this means that the match must be either at the beginning, or after a
748 newline. The bug was in the code for advancing after a failed match and
749 checking that the new position followed a newline. It was not taking
750 account of UTF-8 characters correctly.
751
752 23. PCRE was behaving differently from Perl in the way it recognized POSIX
753 character classes. PCRE was not treating the sequence [:...:] as a
754 character class unless the ... were all letters. Perl, however, seems to
755 allow any characters between [: and :], though of course it rejects as
756 unknown any "names" that contain non-letters, because all the known class
757 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
758 for example, whereas PCRE did not - it did not recognize a POSIX character
759 class. This seemed a bit dangerous, so the code has been changed to be
760 closer to Perl. The behaviour is not identical to Perl, because PCRE will
761 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
762 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
763 Perl does, and where it didn't before.
764
765 24. Rewrite so as to remove the single use of %n from pcregrep because in some
766 Windows environments %n is disabled by default.
767
768
769 Version 7.4 21-Sep-07
770 ---------------------
771
772 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
773 means that a class such as [\s] counted as "explicit reference to CR or
774 LF". That isn't really right - the whole point of the change was to try to
775 help when there was an actual mention of one of the two characters. So now
776 the change happens only if \r or \n (or a literal CR or LF) character is
777 encountered.
778
779 2. The 32-bit options word was also used for 6 internal flags, but the numbers
780 of both had grown to the point where there were only 3 bits left.
781 Fortunately, there was spare space in the data structure, and so I have
782 moved the internal flags into a new 16-bit field to free up more option
783 bits.
784
785 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
786 but did not set the internal JCHANGED flag - either of these is enough to
787 control the way the "get" function works - but the PCRE_INFO_JCHANGED
788 facility is supposed to tell if (?J) was ever used, so now (?J) at the
789 start sets both bits.
790
791 4. Added options (at build time, compile time, exec time) to change \R from
792 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
793
794 5. doc/pcresyntax.html was missing from the distribution.
795
796 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
797 compatibility, even though it is no longer used.
798
799 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
800 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
801 windows.h file is present (where different names are used). [This was
802 reversed later after testing - see 16 below.]
803
804 8. Changed all #include <config.h> to #include "config.h". There were also
805 some further <pcre.h> cases that I changed to "pcre.h".
806
807 9. When pcregrep was used with the --colour option, it missed the line ending
808 sequence off the lines that it output.
809
810 10. It was pointed out to me that arrays of string pointers cause lots of
811 relocations when a shared library is dynamically loaded. A technique of
812 using a single long string with a table of offsets can drastically reduce
813 these. I have refactored PCRE in four places to do this. The result is
814 dramatic:
815
816 Originally: 290
817 After changing UCP table: 187
818 After changing error message table: 43
819 After changing table of "verbs" 36
820 After changing table of Posix names 22
821
822 Thanks to the folks working on Gregex for glib for this insight.
823
824 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
825 unicode-properties was also set.
826
827 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
828
829 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
830 checked only for CRLF.
831
832 14. Added casts to pcretest.c to avoid compiler warnings.
833
834 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
835
836 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
837 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
838 entirely. This removes changes made in 7 above.
839
840 17. The CMake files have been updated, and there is now more information about
841 building with CMake in the NON-UNIX-USE document.
842
843
844 Version 7.3 28-Aug-07
845 ---------------------
846
847 1. In the rejigging of the build system that eventually resulted in 7.1, the
848 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
849 brackets there is not right, since it causes compilers to look for an
850 installed pcre.h, not the version that is in the source that is being
851 compiled (which of course may be different). I have changed it back to:
852
853 #include "pcre.h"
854
855 I have a vague recollection that the change was concerned with compiling in
856 different directories, but in the new build system, that is taken care of
857 by the VPATH setting the Makefile.
858
859 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
860 when the subject happened to end in the byte 0x85 (e.g. if the last
861 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
862 characters but of course it shouldn't be taken as a newline when it is part
863 of another character. The bug was that, for an unlimited repeat of . in
864 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
865 characters when looking for a newline.
866
867 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
868
869 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
870 in debug output.
871
872 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
873 long printing in the pcrecpp unittest when running under MinGW.
874
875 6. ESC_K was left out of the EBCDIC table.
876
877 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
878 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
879 limit also applies to "virtual nesting" when a pattern is recursive, and in
880 this case 1000 isn't so big. I have been able to remove this limit at the
881 expense of backing off one optimization in certain circumstances. Normally,
882 when pcre_exec() would call its internal match() function recursively and
883 immediately return the result unconditionally, it uses a "tail recursion"
884 feature to save stack. However, when a subpattern that can match an empty
885 string has an unlimited repetition quantifier, it no longer makes this
886 optimization. That gives it a stack frame in which to save the data for
887 checking that an empty string has been matched. Previously this was taken
888 from the 1000-entry workspace that had been reserved. So now there is no
889 explicit limit, but more stack is used.
890
891 8. Applied Daniel's patches to solve problems with the import/export magic
892 syntax that is required for Windows, and which was going wrong for the
893 pcreposix and pcrecpp parts of the library. These were overlooked when this
894 problem was solved for the main library.
895
896 9. There were some crude static tests to avoid integer overflow when computing
897 the size of patterns that contain repeated groups with explicit upper
898 limits. As the maximum quantifier is 65535, the maximum group length was
899 set at 30,000 so that the product of these two numbers did not overflow a
900 32-bit integer. However, it turns out that people want to use groups that
901 are longer than 30,000 bytes (though not repeat them that many times).
902 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
903 made it possible to implement the integer overflow checks in a much more
904 dynamic way, which I have now done. The artificial limitation on group
905 length has been removed - we now have only the limit on the total length of
906 the compiled pattern, which depends on the LINK_SIZE setting.
907
908 10. Fixed a bug in the documentation for get/copy named substring when
909 duplicate names are permitted. If none of the named substrings are set, the
910 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
911 empty string.
912
913 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
914 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
915 because the ] is interpreted as the first data character and the
916 terminating ] is not found. PCRE has been made compatible with Perl in this
917 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
918 cause memory overwriting.
919
920 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
921 string has been matched (to stop an infinite loop). It was not recognizing
922 a conditional subpattern that could match an empty string if that
923 subpattern was within another subpattern. For example, it looped when
924 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
925 condition was not nested. This bug has been fixed.
926
927 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
928 past the start of the subject in the presence of bytes with the top bit
929 set, for example "\x8aBCD".
930
931 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
932 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
933
934 14. Optimized (?!) to (*FAIL).
935
936 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
937 This restricts code points to be within the range 0 to 0x10FFFF, excluding
938 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
939 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
940 does: it's just the validity check that is more restrictive.
941
942 16. Inserted checks for integer overflows during escape sequence (backslash)
943 processing, and also fixed erroneous offset values for syntax errors during
944 backslash processing.
945
946 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
947 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
948
949 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
950 caused an overrun.
951
952 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
953 something other than just ASCII characters) inside a group that had an
954 unlimited repeat caused a loop at compile time (while checking to see
955 whether the group could match an empty string).
956
957 20. Debugging a pattern containing \p or \P could cause a crash. For example,
958 [\P{Any}] did so. (Error in the code for printing property names.)
959
960 21. An orphan \E inside a character class could cause a crash.
961
962 22. A repeated capturing bracket such as (A)? could cause a wild memory
963 reference during compilation.
964
965 23. There are several functions in pcre_compile() that scan along a compiled
966 expression for various reasons (e.g. to see if it's fixed length for look
967 behind). There were bugs in these functions when a repeated \p or \P was
968 present in the pattern. These operators have additional parameters compared
969 with \d, etc, and these were not being taken into account when moving along
970 the compiled data. Specifically:
971
972 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
973 length.
974
975 (b) An item such as \pL+ within a repeated group could cause crashes or
976 loops.
977
978 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
979 "reference to non-existent subpattern" error.
980
981 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
982
983 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
984 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
985
986 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
987
988 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
989 character were causing crashes (broken optimization).
990
991 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
992 \p or \P) caused a compile-time loop.
993
994 28. More problems have arisen in unanchored patterns when CRLF is a valid line
995 break. For example, the unstudied pattern [\r\n]A does not match the string
996 "\r\nA" because change 7.0/46 below moves the current point on by two
997 characters after failing to match at the start. However, the pattern \nA
998 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
999 the same is true. There doesn't seem any very clean way out of this, but
1000 what I have chosen to do makes the common cases work: PCRE now takes note
1001 of whether there can be an explicit match for \r or \n anywhere in the
1002 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1003 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1004 pattern has explicit CR or LF references.
1005
1006 29. Added (*CR) etc for changing newline setting at start of pattern.
1007
1008
1009 Version 7.2 19-Jun-07
1010 ---------------------
1011
1012 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1013 which is apparently normally available under Windows.
1014
1015 2. Re-jig the pcregrep tests with different newline settings in an attempt
1016 to make them independent of the local environment's newline setting.
1017
1018 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1019
1020 4. Some of the "internals" tests were previously cut out when the link size
1021 was not 2, because the output contained actual offsets. The recent new
1022 "Z" feature of pcretest means that these can be cut out, making the tests
1023 usable with all link sizes.
1024
1025 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1026 stack recursion. This gives a massive performance boost under BSD, but just
1027 a small improvement under Linux. However, it saves one field in the frame
1028 in all cases.
1029
1030 6. Added more features from the forthcoming Perl 5.10:
1031
1032 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1033 recursion call. It refers to the nth most recently opened parentheses.
1034
1035 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1036 to be opened parentheses.
1037
1038 (c) Conditions that refer to capturing parentheses can be specified
1039 relatively, for example, (?(-2)... or (?(+3)...
1040
1041 (d) \K resets the start of the current match so that everything before
1042 is not part of it.
1043
1044 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1045
1046 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1047 reference syntax.
1048
1049 (g) (?| introduces a group in which the numbering of parentheses in each
1050 alternative starts with the same number.
1051
1052 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1053
1054 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1055 PCRE_INFO_JCHANGED.
1056
1057 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1058 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1059 for detecting groups that can match an empty string.
1060
1061 9. A pattern with a very large number of alternatives (more than several
1062 hundred) was running out of internal workspace during the pre-compile
1063 phase, where pcre_compile() figures out how much memory will be needed. A
1064 bit of new cunning has reduced the workspace needed for groups with
1065 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1066 workspace instead of running out of the 4096 that are available.
1067
1068 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1069
1070 11. Applied patch from Google to remove an optimization that didn't quite work.
1071 The report of the bug said:
1072
1073 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1074 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1075 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1076
1077 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1078 it matched the wrong number of bytes.
1079
1080
1081 Version 7.1 24-Apr-07
1082 ---------------------
1083
1084 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1085 that is more "standard", making use of automake and other Autotools. There
1086 is some re-arrangement of the files and adjustment of comments consequent
1087 on this.
1088
1089 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1090 for recursive directory scanning broke on some systems because the files
1091 are not scanned in any specific order and on different systems the order
1092 was different. A call to "sort" has been inserted into RunGrepTest for the
1093 approprate test as a short-term fix. In the longer term there may be an
1094 alternative.
1095
1096 3. I had an email from Eric Raymond about problems translating some of PCRE's
1097 man pages to HTML (despite the fact that I distribute HTML pages, some
1098 people do their own conversions for various reasons). The problems
1099 concerned the use of low-level troff macros .br and .in. I have therefore
1100 removed all such uses from the man pages (some were redundant, some could
1101 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1102 HTML has been updated to handle .nf/.fi and to complain if it encounters
1103 .br or .in.
1104
1105 4. Updated comments in configure.ac that get placed in config.h.in and also
1106 arranged for config.h to be included in the distribution, with the name
1107 config.h.generic, for the benefit of those who have to compile without
1108 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1109
1110 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1111 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1112 makevp.bat for the current PCRE, using the additional files
1113 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1114
1115 6. A Windows user reported a minor discrepancy with test 2, which turned out
1116 to be caused by a trailing space on an input line that had got lost in his
1117 copy. The trailing space was an accident, so I've just removed it.
1118
1119 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1120 that is needed.
1121
1122 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1123 as "const" (a) because they are and (b) because it helps the PHP
1124 maintainers who have recently made a script to detect big data structures
1125 in the php code that should be moved to the .rodata section. I remembered
1126 to update Builducptable as well, so it won't revert if ucptable.h is ever
1127 re-created.
1128
1129 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1130 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1131 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1132 support is not required. This saves 1.5-2K of code, which is important in
1133 some applications.
1134
1135 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1136 so as not to refer to the tables, even though these functions will never be
1137 called when UTF-8 support is disabled. Otherwise there are problems with a
1138 shared library.
1139
1140 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1141
1142 (a) It was defining its arguments as char * instead of void *.
1143
1144 (b) It was assuming that all moves were upwards in memory; this was true
1145 a long time ago when I wrote it, but is no longer the case.
1146
1147 The emulated memove() is provided for those environments that have neither
1148 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1149 is clearly not the case, as these two bugs were recently reported.
1150
1151 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1152 and Detrail to create the HTML documentation, the .txt form of the man
1153 pages, and it removes trailing spaces from listed files. It also creates
1154 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1155 case, it wraps all the #defines with #ifndefs. This script should be run
1156 before "make dist".
1157
1158 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1159 with Unicode property support.
1160
1161 (a) For a maximizing quantifier, if the two different cases of the
1162 character were of different lengths in their UTF-8 codings (there are
1163 some cases like this - I found 11), and the matching function had to
1164 back up over a mixture of the two cases, it incorrectly assumed they
1165 were both the same length.
1166
1167 (b) When PCRE was configured to use the heap rather than the stack for
1168 recursion during matching, it was not correctly preserving the data for
1169 the other case of a UTF-8 character when checking ahead for a match
1170 while processing a minimizing repeat. If the check also involved
1171 matching a wide character, but failed, corruption could cause an
1172 erroneous result when trying to check for a repeat of the original
1173 character.
1174
1175 13. Some tidying changes to the testing mechanism:
1176
1177 (a) The RunTest script now detects the internal link size and whether there
1178 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1179 values substituted by "configure". (The RunGrepTest script already did
1180 this for UTF-8.) The configure.ac script no longer substitutes the
1181 relevant variables.
1182
1183 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1184 with length and offset values. This means that the output is different
1185 for different internal link sizes. Test 2 is skipped for link sizes
1186 other than 2 because of this, bypassing the problem. Unfortunately,
1187 there was also a test in test 3 (the locale tests) that used /B and
1188 failed for link sizes other than 2. Rather than cut the whole test out,
1189 I have added a new /Z option to pcretest that replaces the length and
1190 offset values with spaces. This is now used to make test 3 independent
1191 of link size. (Test 2 will be tidied up later.)
1192
1193 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1194 segmentation fault instead of returning the appropriate error message.
1195
1196 15. In multiline mode when the newline sequence was set to "any", the pattern
1197 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1198 This doesn't seem right; it now treats the CRLF combination as the line
1199 ending, and so does not match in that case. It's only a pattern such as ^$
1200 that would hit this one: something like ^ABC$ would have failed after \r
1201 and then tried again after \r\n.
1202
1203 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1204 in an attempt to make files that differ only in their line terminators
1205 compare equal. This works on Linux.
1206
1207 17. Under certain error circumstances pcregrep might try to free random memory
1208 as it exited. This is now fixed, thanks to valgrind.
1209
1210 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1211 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1212 was because its rules for how to advance for /g after matching an empty
1213 string at the end of a line did not allow for this case. They now check for
1214 it specially.
1215
1216 20. pcretest is supposed to handle patterns and data of any length, by
1217 extending its buffers when necessary. It was getting this wrong when the
1218 buffer for a data line had to be extended.
1219
1220 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1221 CRLF as a newline sequence.
1222
1223 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1224 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1225 I have nevertheless tidied it up.
1226
1227 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1228
1229 24. Added a man page for pcre-config.
1230
1231
1232 Version 7.0 19-Dec-06
1233 ---------------------
1234
1235 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1236 moving to gcc 4.1.1.
1237
1238 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1239 sys/time.h, which is documented as needed for this function. It doesn't
1240 seem to matter on Linux, but it showed up on some releases of OS X.
1241
1242 3. It seems that there are systems where bytes whose values are greater than
1243 127 match isprint() in the "C" locale. The "C" locale should be the
1244 default when a C program starts up. In most systems, only ASCII printing
1245 characters match isprint(). This difference caused the output from pcretest
1246 to vary, making some of the tests fail. I have changed pcretest so that:
1247
1248 (a) When it is outputting text in the compiled version of a pattern, bytes
1249 other than 32-126 are always shown as hex escapes.
1250
1251 (b) When it is outputting text that is a matched part of a subject string,
1252 it does the same, unless a different locale has been set for the match
1253 (using the /L modifier). In this case, it uses isprint() to decide.
1254
1255 4. Fixed a major bug that caused incorrect computation of the amount of memory
1256 required for a compiled pattern when options that changed within the
1257 pattern affected the logic of the preliminary scan that determines the
1258 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1259 that the computed length was too small. The symptoms of this bug were
1260 either the PCRE error "internal error: code overflow" from pcre_compile(),
1261 or a glibc crash with a message such as "pcretest: free(): invalid next
1262 size (fast)". Examples of patterns that provoked this bug (shown in
1263 pcretest format) are:
1264
1265 /(?-x: )/x
1266 /(?x)(?-x: \s*#\s*)/
1267 /((?i)[\x{c0}])/8
1268 /(?i:[\x{c0}])/8
1269
1270 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1271 is now done differently.
1272
1273 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1274 wrapper classes; (b) implement a new function in the C++ scanner that is
1275 more efficient than the old way of doing things because it avoids levels of
1276 recursion in the regex matching; (c) add a paragraph to the documentation
1277 for the FullMatch() function.
1278
1279 6. The escape sequence \n was being treated as whatever was defined as
1280 "newline". Not only was this contrary to the documentation, which states
1281 that \n is character 10 (hex 0A), but it also went horribly wrong when
1282 "newline" was defined as CRLF. This has been fixed.
1283
1284 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1285 was being set to -1 for the "end of line" case (supposedly a value that no
1286 character can have). Though this value is never used (the check for end of
1287 line is "zero bytes in current character"), it caused compiler complaints.
1288 I've changed it to 0xffffffff.
1289
1290 8. In pcre_version.c, the version string was being built by a sequence of
1291 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1292 string (as it is for production releases) called a macro with an empty
1293 argument. The C standard says the result of this is undefined. The gcc
1294 compiler treats it as an empty string (which was what was wanted) but it is
1295 reported that Visual C gives an error. The source has been hacked around to
1296 avoid this problem.
1297
1298 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1299 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1300 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1301 of them did).
1302
1303 10. Originally, pcretest opened its input and output without "b"; then I was
1304 told that "b" was needed in some environments, so it was added for release
1305 5.0 to both the input and output. (It makes no difference on Unix-like
1306 systems.) Later I was told that it is wrong for the input on Windows. I've
1307 now abstracted the modes into two macros, to make it easier to fiddle with
1308 them, and removed "b" from the input mode under Windows.
1309
1310 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1311
1312 12. Added -help and --help to pcretest as an official way of being reminded
1313 of the options.
1314
1315 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1316 and pcrecpp.cc because they annoy compilers at high warning levels.
1317
1318 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1319
1320 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1321 scripts are not C programs :-) and which was not noticed because it works
1322 on Linux.
1323
1324 16. pcretest is supposed to handle any length of pattern and data line (as one
1325 line or as a continued sequence of lines) by extending its input buffer if
1326 necessary. This feature was broken for very long pattern lines, leading to
1327 a string of junk being passed to pcre_compile() if the pattern was longer
1328 than about 50K.
1329
1330 17. I have done a major re-factoring of the way pcre_compile() computes the
1331 amount of memory needed for a compiled pattern. Previously, there was code
1332 that made a preliminary scan of the pattern in order to do this. That was
1333 OK when PCRE was new, but as the facilities have expanded, it has become
1334 harder and harder to keep it in step with the real compile phase, and there
1335 have been a number of bugs (see for example, 4 above). I have now found a
1336 cunning way of running the real compile function in a "fake" mode that
1337 enables it to compute how much memory it would need, while actually only
1338 ever using a few hundred bytes of working memory and without too many
1339 tests of the mode. This should make future maintenance and development
1340 easier. A side effect of this work is that the limit of 200 on the nesting
1341 depth of parentheses has been removed (though this was never a serious
1342 limitation, I suspect). However, there is a downside: pcre_compile() now
1343 runs more slowly than before (30% or more, depending on the pattern). I
1344 hope this isn't a big issue. There is no effect on runtime performance.
1345
1346 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1347 newline (only possible for the last line of a file) and it was a
1348 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1349
1350 19. Added additional timing features to pcretest. (1) The -tm option now times
1351 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1352 separate command line item, by a number that specifies the number of
1353 repeats to use when timing. The default is 50000; this gives better
1354 precision, but takes uncomfortably long for very large patterns.
1355
1356 20. Extended pcre_study() to be more clever in cases where a branch of a
1357 subpattern has no definite first character. For example, (a*|b*)[cd] would
1358 previously give no result from pcre_study(). Now it recognizes that the
1359 first character must be a, b, c, or d.
1360
1361 21. There was an incorrect error "recursive call could loop indefinitely" if
1362 a subpattern (or the entire pattern) that was being tested for matching an
1363 empty string contained only one non-empty item after a nested subpattern.
1364 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1365 incorrectly, because the \d was being skipped in the check.
1366
1367 22. The pcretest program now has a new pattern option /B and a command line
1368 option -b, which is equivalent to adding /B to every pattern. This causes
1369 it to show the compiled bytecode, without the additional information that
1370 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1371 is the same as /B/I).
1372
1373 23. A new optimization is now able automatically to treat some sequences such
1374 as a*b as a*+b. More specifically, if something simple (such as a character
1375 or a simple class like \d) has an unlimited quantifier, and is followed by
1376 something that cannot possibly match the quantified thing, the quantifier
1377 is automatically "possessified".
1378
1379 24. A recursive reference to a subpattern whose number was greater than 39
1380 went wrong under certain circumstances in UTF-8 mode. This bug could also
1381 have affected the operation of pcre_study().
1382
1383 25. Realized that a little bit of performance could be had by replacing
1384 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1385
1386 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1387
1388 27. Possessive quantifiers such as a++ were previously implemented by turning
1389 them into atomic groups such as ($>a+). Now they have their own opcodes,
1390 which improves performance. This includes the automatically created ones
1391 from 23 above.
1392
1393 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1394 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1395 the first matched character to be a colon. This applied both to named and
1396 numbered groups.
1397
1398 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1399
1400 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1401 building PCRE on Windows easier, so I have included it in the distribution.
1402
1403 31. There is now a check in pcretest against a ridiculously large number being
1404 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1405 loop, the loop is abandoned.
1406
1407 32. Forward references to subpatterns in conditions such as (?(2)...) where
1408 subpattern 2 is defined later cause pcre_compile() to search forwards in
1409 the pattern for the relevant set of parentheses. This search went wrong
1410 when there were unescaped parentheses in a character class, parentheses
1411 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1412
1413 33. "Subroutine" calls and backreferences were previously restricted to
1414 referencing subpatterns earlier in the regex. This restriction has now
1415 been removed.
1416
1417 34. Added a number of extra features that are going to be in Perl 5.10. On the
1418 whole, these are just syntactic alternatives for features that PCRE had
1419 previously implemented using the Python syntax or my own invention. The
1420 other formats are all retained for compatibility.
1421
1422 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1423 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1424 also .NET compatible.
1425
1426 (b) A recursion or subroutine call to a named group can now be defined as
1427 (?&name) as well as (?P>name).
1428
1429 (c) A backreference to a named group can now be defined as \k<name> or
1430 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1431 5.10, are also .NET compatible.
1432
1433 (d) A conditional reference to a named group can now use the syntax
1434 (?(<name>) or (?('name') as well as (?(name).
1435
1436 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1437 groups (named and numbered) that are never evaluated inline, but can be
1438 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1439 is always false. There may be only one alternative in such a group.
1440
1441 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1442 as the simple (?(R). The condition is true only if the most recent
1443 recursion is that of the given number or name. It does not search out
1444 through the entire recursion stack.
1445
1446 (g) The escape \gN or \g{N} has been added, where N is a positive or
1447 negative number, specifying an absolute or relative reference.
1448
1449 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1450 some "unreachable code" warnings.
1451
1452 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1453 things, this adds five new scripts.
1454
1455 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1456 There were also incompatibilities regarding the handling of \Q..\E inside
1457 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1458 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1459
1460 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1461 matches an empty string, and forcibly breaks the loop. There were bugs in
1462 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1463 against aaaa the result was just "a" rather than "aaaa", for example. Two
1464 separate and independent bugs (that affected different cases) have been
1465 fixed.
1466
1467 39. Refactored the code to abolish the use of different opcodes for small
1468 capturing bracket numbers. This is a tidy that I avoided doing when I
1469 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1470 The new approach is not only tidier, it makes it possible to reduce the
1471 memory needed to fix the previous bug (38).
1472
1473 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1474 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1475 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1476 mode.
1477
1478 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1479 report.
1480
1481 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1482 copy construction and assignment in the C++ wrapper.
1483
1484 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1485 couple of bugs that could have given wrong results in the "--newline=crlf"
1486 case.
1487
1488 44. Added a number of casts and did some reorganization of signed/unsigned int
1489 variables following suggestions from Dair Grant. Also renamed the variable
1490 "this" as "item" because it is a C++ keyword.
1491
1492 45. Arranged for dftables to add
1493
1494 #include "pcre_internal.h"
1495
1496 to pcre_chartables.c because without it, gcc 4.x may remove the array
1497 definition from the final binary if PCRE is built into a static library and
1498 dead code stripping is activated.
1499
1500 46. For an unanchored pattern, if a match attempt fails at the start of a
1501 newline sequence, and the newline setting is CRLF or ANY, and the next two
1502 characters are CRLF, advance by two characters instead of one.
1503
1504
1505 Version 6.7 04-Jul-06
1506 ---------------------
1507
1508 1. In order to handle tests when input lines are enormously long, pcretest has
1509 been re-factored so that it automatically extends its buffers when
1510 necessary. The code is crude, but this _is_ just a test program. The
1511 default size has been increased from 32K to 50K.
1512
1513 2. The code in pcre_study() was using the value of the re argument before
1514 testing it for NULL. (Of course, in any sensible call of the function, it
1515 won't be NULL.)
1516
1517 3. The memmove() emulation function in pcre_internal.h, which is used on
1518 systems that lack both memmove() and bcopy() - that is, hardly ever -
1519 was missing a "static" storage class specifier.
1520
1521 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1522 containing an extended class (one that cannot be represented by a bitmap
1523 because it contains high-valued characters or Unicode property items, e.g.
1524 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1525 pattern, but PCRE should not loop if you do not (it no longer does).
1526 [Detail: two cases were found: (a) a repeated subpattern containing an
1527 extended class; (b) a recursive reference to a subpattern that followed a
1528 previous extended class. It wasn't skipping over the extended class
1529 correctly when UTF-8 mode was not set.]
1530
1531 5. A negated single-character class was not being recognized as fixed-length
1532 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1533 compile error "lookbehind assertion is not fixed length".
1534
1535 6. The RunPerlTest auxiliary script was showing an unexpected difference
1536 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1537 write a Perl script that can interpret lines of an input file either as
1538 byte characters or as UTF-8, which is what "perltest" was being required to
1539 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1540 can't do is switch easily at run time between having the "use utf8;" pragma
1541 or not. In the end, I fudged it by using the RunPerlTest script to insert
1542 "use utf8;" explicitly for the UTF-8 tests.
1543
1544 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1545 the end of the subject string, contrary to the documentation and to what
1546 Perl does. This was true of both matching functions. Now it matches only at
1547 the start of the subject and immediately after *internal* newlines.
1548
1549 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1550 a pointer to an int instead of a pointer to an unsigned long int. This
1551 caused problems on 64-bit systems.
1552
1553 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1554 instance of the 'standard' template library not being so standard".
1555
1556 10. There was no check on the number of named subpatterns nor the maximum
1557 length of a subpattern name. The product of these values is used to compute
1558 the size of the memory block for a compiled pattern. By supplying a very
1559 long subpattern name and a large number of named subpatterns, the size
1560 computation could be caused to overflow. This is now prevented by limiting
1561 the length of names to 32 characters, and the number of named subpatterns
1562 to 10,000.
1563
1564 11. Subpatterns that are repeated with specific counts have to be replicated in
1565 the compiled pattern. The size of memory for this was computed from the
1566 length of the subpattern and the repeat count. The latter is limited to
1567 65535, but there was no limit on the former, meaning that integer overflow
1568 could in principle occur. The compiled length of a repeated subpattern is
1569 now limited to 30,000 bytes in order to prevent this.
1570
1571 12. Added the optional facility to have named substrings with the same name.
1572
1573 13. Added the ability to use a named substring as a condition, using the
1574 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1575 are numbers (not recommended). Forward references are permitted.
1576
1577 14. Added forward references in named backreferences (if you see what I mean).
1578
1579 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1580 pattern could run off the end of the subject. For example, the pattern
1581 "(?s)(.{1,5})"8 did this with the subject "ab".
1582
1583 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1584 PCRE_CASELESS was set when matching characters that were quantified with ?
1585 or *.
1586
1587 17. A character class other than a single negated character that had a minimum
1588 but no maximum quantifier - for example [ab]{6,} - was not handled
1589 correctly by pce_dfa_exec(). It would match only one character.
1590
1591 18. A valid (though odd) pattern that looked like a POSIX character
1592 class but used an invalid character after [ (for example [[,abc,]]) caused
1593 pcre_compile() to give the error "Failed: internal error: code overflow" or
1594 in some cases to crash with a glibc free() error. This could even happen if
1595 the pattern terminated after [[ but there just happened to be a sequence of
1596 letters, a binary zero, and a closing ] in the memory that followed.
1597
1598 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1599 over the years. Originally (before any Unicode support), just the bottom 8
1600 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1601 output from "man perlunicode" includes this:
1602
1603 The regular expression compiler produces polymorphic opcodes. That
1604 is, the pattern adapts to the data and automatically switches to
1605 the Unicode character scheme when presented with Unicode data--or
1606 instead uses a traditional byte scheme when presented with byte
1607 data.
1608
1609 Sadly, a wide octal escape does not cause a switch, and in a string with
1610 no other multibyte characters, these octal escapes are treated as before.
1611 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1612 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1613 Unicode string.
1614
1615 I have not perpetrated such confusion in PCRE. Up till now, it took just
1616 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1617 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1618 translate to the appropriate multibyte character.
1619
1620 29. Applied some refactoring to reduce the number of warnings from Microsoft
1621 and Borland compilers. This has included removing the fudge introduced
1622 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1623 a warning about an unused variable.
1624
1625 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1626 characters since release 4.0, because Perl (from release 5.004) does not.
1627 [Or at least, is documented not to: some releases seem to be in conflict
1628 with the documentation.] However, when a pattern was studied with
1629 pcre_study() and all its branches started with \s, PCRE still included VT
1630 as a possible starting character. Of course, this did no harm; it just
1631 caused an unnecessary match attempt.
1632
1633 22. Removed a now-redundant internal flag bit that recorded the fact that case
1634 dependency changed within the pattern. This was once needed for "required
1635 byte" processing, but is no longer used. This recovers a now-scarce options
1636 bit. Also moved the least significant internal flag bit to the most-
1637 significant bit of the word, which was not previously used (hangover from
1638 the days when it was an int rather than a uint) to free up another bit for
1639 the future.
1640
1641 23. Added support for CRLF line endings as well as CR and LF. As well as the
1642 default being selectable at build time, it can now be changed at runtime
1643 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1644 specify that it is scanning data with non-default line endings.
1645
1646 24. Changed the definition of CXXLINK to make it agree with the definition of
1647 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1648
1649 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1650 recursions. This makes a big different to stack usage for some patterns.
1651
1652 26. If a subpattern containing a named recursion or subroutine reference such
1653 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1654 the space required for the compiled pattern went wrong and gave too small a
1655 value. Depending on the environment, this could lead to "Failed: internal
1656 error: code overflow at offset 49" or "glibc detected double free or
1657 corruption" errors.
1658
1659 27. Applied patches from Google (a) to support the new newline modes and (b) to
1660 advance over multibyte UTF-8 characters in GlobalReplace.
1661
1662 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1663 difference for some implementation of PCRE in some Windows version.
1664
1665 29. Added some extra testing facilities to pcretest:
1666
1667 \q<number> in a data line sets the "match limit" value
1668 \Q<number> in a data line sets the "match recursion limt" value
1669 -S <number> sets the stack size, where <number> is in megabytes
1670
1671 The -S option isn't available for Windows.
1672
1673
1674 Version 6.6 06-Feb-06
1675 ---------------------
1676
1677 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1678 in pcreposix.h. I have copied the definition from pcre.h.
1679
1680 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1681 because pcre.h is no longer a built file.
1682
1683 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1684 not normally included in the compiled code.
1685
1686
1687 Version 6.5 01-Feb-06
1688 ---------------------
1689
1690 1. When using the partial match feature with pcre_dfa_exec(), it was not
1691 anchoring the second and subsequent partial matches at the new starting
1692 point. This could lead to incorrect results. For example, with the pattern
1693 /1234/, partially matching against "123" and then "a4" gave a match.
1694
1695 2. Changes to pcregrep:
1696
1697 (a) All non-match returns from pcre_exec() were being treated as failures
1698 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1699 error message is output. Some extra information is given for the
1700 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1701 probably the only errors that are likely to be caused by users (by
1702 specifying a regex that has nested indefinite repeats, for instance).
1703 If there are more than 20 of these errors, pcregrep is abandoned.
1704
1705 (b) A binary zero was treated as data while matching, but terminated the
1706 output line if it was written out. This has been fixed: binary zeroes
1707 are now no different to any other data bytes.
1708
1709 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1710 used to set a locale for matching. The --locale=xxxx long option has
1711 been added (no short equivalent) to specify a locale explicitly on the
1712 pcregrep command, overriding the environment variables.
1713
1714 (d) When -B was used with -n, some line numbers in the output were one less
1715 than they should have been.
1716
1717 (e) Added the -o (--only-matching) option.
1718
1719 (f) If -A or -C was used with -c (count only), some lines of context were
1720 accidentally printed for the final match.
1721
1722 (g) Added the -H (--with-filename) option.
1723
1724 (h) The combination of options -rh failed to suppress file names for files
1725 that were found from directory arguments.
1726
1727 (i) Added the -D (--devices) and -d (--directories) options.
1728
1729 (j) Added the -F (--fixed-strings) option.
1730
1731 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1732
1733 (l) Added the --colo(u)r option.
1734
1735 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1736 is not present by default.
1737
1738 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1739 items such as (?R) or (?1), when the recursion could match a number of
1740 alternatives. If it matched one of the alternatives, but subsequently,
1741 outside the recursion, there was a failure, the code tried to back up into
1742 the recursion. However, because of the way PCRE is implemented, this is not
1743 possible, and the result was an incorrect result from the match.
1744
1745 In order to prevent this happening, the specification of recursion has
1746 been changed so that all such subpatterns are automatically treated as
1747 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1748
1749 4. I had overlooked the fact that, in some locales, there are characters for
1750 which isalpha() is true but neither isupper() nor islower() are true. In
1751 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1752 and ordfeminine) are like this. This affected the treatment of \w and \W
1753 when they appeared in character classes, but not when they appeared outside
1754 a character class. The bit map for "word" characters is now created
1755 separately from the results of isalnum() instead of just taking it from the
1756 upper, lower, and digit maps. (Plus the underscore character, of course.)
1757
1758 5. The above bug also affected the handling of POSIX character classes such as
1759 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1760 permanent tables. Instead, the bit maps for such a class were previously
1761 created as the appropriate unions of the upper, lower, and digit bitmaps.
1762 Now they are created by subtraction from the [[:word:]] class, which has
1763 its own bitmap.
1764
1765 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1766 It is created by subtracting the vertical space characters (\x09, \x0a,
1767 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1768 subtraction was done in the overall bitmap for a character class, meaning
1769 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1770 be recognized. This bug has been fixed.
1771
1772 7. Patches from the folks at Google:
1773
1774 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1775 real life, but is still worth protecting against".
1776
1777 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1778 regular expressions".
1779
1780 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1781 have it.
1782
1783 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1784 "configure" and the latter not, in order to fix a problem somebody had
1785 with compiling the Arg class on HP-UX.
1786
1787 (e) Improve the error-handling of the C++ wrapper a little bit.
1788
1789 (f) New tests for checking recursion limiting.
1790
1791 8. The pcre_memmove() function, which is used only if the environment does not
1792 have a standard memmove() function (and is therefore rarely compiled),
1793 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1794 returning a result (though PCRE never actually uses the result).
1795
1796 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1797 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1798 returned instead of calling malloc() with an overflowing number that would
1799 most likely cause subsequent chaos.
1800
1801 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1802
1803 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1804 with this option is matched, the nmatch and pmatch options of regexec() are
1805 ignored.
1806
1807 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1808 provided in case anyone wants to the the POSIX interface with UTF-8
1809 strings.
1810
1811 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1812 C++ linking (needed for some HP-UX environments).
1813
1814 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1815 (unused parameter) and in the pcre_printint() function (omitted "default"
1816 switch label when the default is to do nothing).
1817
1818 15. Added some code to make it possible, when PCRE is compiled as a C++
1819 library, to replace subject pointers for pcre_exec() with a smart pointer
1820 class, thus making it possible to process discontinuous strings.
1821
1822 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1823 much the same function. They were added by different people who were trying
1824 to make PCRE easy to compile on non-Unix systems. It has been suggested
1825 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1826 for compiling on Windows systems. I have therefore replaced it with
1827 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1828 defaults to "extern" for C or "extern C" for C++, which works fine on
1829 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1830 SCOPE with something explicit in config.h. In addition:
1831
1832 (a) pcreposix.h still had just "extern" instead of either of these macros;
1833 I have replaced it with PCRE_DATA_SCOPE.
1834
1835 (b) Functions such as _pcre_xclass(), which are internal to the library,
1836 but external in the C sense, all had PCRE_EXPORT in their definitions.
1837 This is apparently wrong for the Windows case, so I have removed it.
1838 (It makes no difference on Unix-like systems.)
1839
1840 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1841 of recursive calls to match(). This is different to MATCH_LIMIT because
1842 that limits the total number of calls to match(), not all of which increase
1843 the depth of recursion. Limiting the recursion depth limits the amount of
1844 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1845 when PCRE is compiled, and changed at run time. A patch from Google adds
1846 this functionality to the C++ interface.
1847
1848 18. Changes to the handling of Unicode character properties:
1849
1850 (a) Updated the table to Unicode 4.1.0.
1851
1852 (b) Recognize characters that are not in the table as "Cn" (undefined).
1853
1854 (c) I revised the way the table is implemented to a much improved format
1855 which includes recognition of ranges. It now supports the ranges that
1856 are defined in UnicodeData.txt, and it also amalgamates other
1857 characters into ranges. This has reduced the number of entries in the
1858 table from around 16,000 to around 3,000, thus reducing its size
1859 considerably. I realized I did not need to use a tree structure after
1860 all - a binary chop search is just as efficient. Having reduced the
1861 number of entries, I extended their size from 6 bytes to 8 bytes to
1862 allow for more data.
1863
1864 (d) Added support for Unicode script names via properties such as \p{Han}.
1865
1866 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1867 matching that character.
1868
1869 20. When matching a repeated Unicode property with a minimum greater than zero,
1870 (for example \pL{2,}), PCRE could look past the end of the subject if it
1871 reached it while seeking the minimum number of characters. This could
1872 happen only if some of the characters were more than one byte long, because
1873 there is a check for at least the minimum number of bytes.
1874
1875 21. Refactored the implementation of \p and \P so as to be more general, to
1876 allow for more different types of property in future. This has changed the
1877 compiled form incompatibly. Anybody with saved compiled patterns that use
1878 \p or \P will have to recompile them.
1879
1880 22. Added "Any" and "L&" to the supported property types.
1881
1882 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1883 but give a compile time error if the value is greater than 0xff.
1884
1885 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1886 accidentally not being installed or uninstalled.
1887
1888 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1889 made were to insert the current release number. This seemed silly, because
1890 it made things harder for people building PCRE on systems that don't run
1891 "configure". I have turned pcre.h into a distributed file, no longer built
1892 by "configure", with the version identification directly included. There is
1893 no longer a pcre.h.in file.
1894
1895 However, this change necessitated a change to the pcre-config script as
1896 well. It is built from pcre-config.in, and one of the substitutions was the
1897 release number. I have updated configure.ac so that ./configure now finds
1898 the release number by grepping pcre.h.
1899
1900 26. Added the ability to run the tests under valgrind.
1901
1902
1903 Version 6.4 05-Sep-05
1904 ---------------------
1905
1906 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1907 "--" to be printed when multiple files were scanned, even when none of the
1908 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1909 consider it to be a bug, and have restored the previous behaviour.
1910
1911 2. A couple of code tidies to get rid of compiler warnings.
1912
1913 3. The pcretest program used to cheat by referring to symbols in the library
1914 whose names begin with _pcre_. These are internal symbols that are not
1915 really supposed to be visible externally, and in some environments it is
1916 possible to suppress them. The cheating is now confined to including
1917 certain files from the library's source, which is a bit cleaner.
1918
1919 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1920 file's purpose clearer.
1921
1922 5. Reorganized pcre_ucp_findchar().
1923
1924
1925 Version 6.3 15-Aug-05
1926 ---------------------
1927
1928 1. The file libpcre.pc.in did not have general read permission in the tarball.
1929
1930 2. There were some problems when building without C++ support:
1931
1932 (a) If C++ support was not built, "make install" and "make test" still
1933 tried to test it.
1934
1935 (b) There were problems when the value of CXX was explicitly set. Some
1936 changes have been made to try to fix these, and ...
1937
1938 (c) --disable-cpp can now be used to explicitly disable C++ support.
1939
1940 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1941 backslash in a target when C++ was disabled. This confuses some
1942 versions of "make", apparently. Using an intermediate variable solves
1943 this. (Same for CPP_LOBJ.)
1944
1945 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1946 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1947 necessary on certain architectures.
1948
1949 4. Added a setting of -export-symbols-regex to the link command to remove
1950 those symbols that are exported in the C sense, but actually are local
1951 within the library, and not documented. Their names all begin with
1952 "_pcre_". This is not a perfect job, because (a) we have to except some
1953 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1954 available (and never for static libraries). I have made a note to try to
1955 find a way round (a) in the future.
1956
1957
1958 Version 6.2 01-Aug-05
1959 ---------------------
1960
1961 1. There was no test for integer overflow of quantifier values. A construction
1962 such as {1111111111111111} would give undefined results. What is worse, if
1963 a minimum quantifier for a parenthesized subpattern overflowed and became
1964 negative, the calculation of the memory size went wrong. This could have
1965 led to memory overwriting.
1966
1967 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1968
1969 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1970 operating environments where this matters.
1971
1972 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1973 PCRE options from within the C++ wrapper.
1974
1975 5. Named capturing subpatterns were not being correctly counted when a pattern
1976 was compiled. This caused two problems: (a) If there were more than 100
1977 such subpatterns, the calculation of the memory needed for the whole
1978 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1979 back references of the form \12, where the number was greater than 9, were
1980 not recognized as back references, even though there were sufficient
1981 previous subpatterns.
1982
1983 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1984 versions of gcc, e.g. 2.95.4.
1985
1986
1987 Version 6.1 21-Jun-05
1988 ---------------------
1989
1990 1. There was one reference to the variable "posix" in pcretest.c that was not
1991 surrounded by "#if !defined NOPOSIX".
1992
1993 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1994 the cross-check on the old pcre_info() function, for the benefit of the
1995 cut-down version of PCRE that is currently imported into Exim.
1996
1997 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1998 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1999 patterns that start (?i) but I don't think that matters. The use of (?i) is
2000 just an example; this all applies to the other options as well.
2001
2002 4. Since libtool seems to echo the compile commands it is issuing, the output
2003 from "make" can be reduced a bit by putting "@" in front of each libtool
2004 compile command.
2005
2006 5. Patch from the folks at Google for configure.in to be a bit more thorough
2007 in checking for a suitable C++ installation before trying to compile the
2008 C++ stuff. This should fix a reported problem when a compiler was present,
2009 but no suitable headers.
2010
2011 6. The man pages all had just "PCRE" as their title. I have changed them to
2012 be the relevant file name. I have also arranged that these names are
2013 retained in the file doc/pcre.txt, which is a concatenation in text format
2014 of all the man pages except the little individual ones for each function.
2015
2016 7. The NON-UNIX-USE file had not been updated for the different set of source
2017 files that come with release 6. I also added a few comments about the C++
2018 wrapper.
2019
2020
2021 Version 6.0 07-Jun-05
2022 ---------------------
2023
2024 1. Some minor internal re-organization to help with my DFA experiments.
2025
2026 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2027 didn't matter for the library itself when fully configured, but did matter
2028 when compiling without UCP support, or within Exim, where the ucp files are
2029 not imported.
2030
2031 3. Refactoring of the library code to split up the various functions into
2032 different source modules. The addition of the new DFA matching code (see
2033 below) to a single monolithic source would have made it really too
2034 unwieldy, quite apart from causing all the code to be include in a
2035 statically linked application, when only some functions are used. This is
2036 relevant even without the DFA addition now that patterns can be compiled in
2037 one application and matched in another.
2038
2039 The downside of splitting up is that there have to be some external
2040 functions and data tables that are used internally in different modules of
2041 the library but which are not part of the API. These have all had their
2042 names changed to start with "_pcre_" so that they are unlikely to clash
2043 with other external names.
2044
2045 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2046 a different (DFA) algorithm. Although it is slower than the original
2047 function, it does have some advantages for certain types of matching
2048 problem.
2049
2050 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2051 including restarting after a partial match.
2052
2053 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2054 defined when compiling for Windows was sent to me. I have put it into the
2055 code, though I have no means of testing or verifying it.
2056
2057 7. Added the pcre_refcount() auxiliary function.
2058
2059 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2060 match before or at the first newline in the subject string. In pcretest,
2061 the /f option on a pattern can be used to set this.
2062
2063 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2064 would behave wrongly. This has been present in PCRE since release 4.0.
2065
2066 10. A number of changes to the pcregrep command:
2067
2068 (a) Refactored how -x works; insert ^(...)$ instead of setting
2069 PCRE_ANCHORED and checking the length, in preparation for adding
2070 something similar for -w.
2071
2072 (b) Added the -w (match as a word) option.
2073
2074 (c) Refactored the way lines are read and buffered so as to have more
2075 than one at a time available.
2076
2077 (d) Implemented a pcregrep test script.
2078
2079 (e) Added the -M (multiline match) option. This allows patterns to match
2080 over several lines of the subject. The buffering ensures that at least
2081 8K, or the rest of the document (whichever is the shorter) is available
2082 for matching (and similarly the previous 8K for lookbehind assertions).
2083
2084 (f) Changed the --help output so that it now says
2085
2086 -w, --word-regex(p)
2087
2088 instead of two lines, one with "regex" and the other with "regexp"
2089 because that confused at least one person since the short forms are the
2090 same. (This required a bit of code, as the output is generated
2091 automatically from a table. It wasn't just a text change.)
2092
2093 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2094 option but starts with a hyphen. Could be a pattern or a path name
2095 starting with a hyphen, for instance.
2096
2097 (h) "-" can be given as a file name to represent stdin.
2098
2099 (i) When file names are being printed, "(standard input)" is used for
2100 the standard input, for compatibility with GNU grep. Previously
2101 "<stdin>" was used.
2102
2103 (j) The option --label=xxx can be used to supply a name to be used for
2104 stdin when file names are being printed. There is no short form.
2105
2106 (k) Re-factored the options decoding logic because we are going to add
2107 two more options that take data. Such options can now be given in four
2108 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2109
2110 (l) Added the -A, -B, and -C options for requesting that lines of context
2111 around matches be printed.
2112
2113 (m) Added the -L option to print the names of files that do not contain
2114 any matching lines, that is, the complement of -l.
2115
2116 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2117 continue to scan other files.
2118
2119 (o) The -s option was incorrectly implemented. For compatibility with other
2120 greps, it now suppresses the error message for a non-existent or non-
2121 accessible file (but not the return code). There is a new option called
2122 -q that suppresses the output of matching lines, which was what -s was
2123 previously doing.
2124
2125 (p) Added --include and --exclude options to specify files for inclusion
2126 and exclusion when recursing.
2127
2128 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2129 Hopefully, it now does.
2130
2131 12. Missing cast in pcre_study().
2132
2133 13. Added an "uninstall" target to the makefile.
2134
2135 14. Replaced "extern" in the function prototypes in Makefile.in with
2136 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2137 world, but is set differently for Windows.
2138
2139 15. Added a second compiling function called pcre_compile2(). The only
2140 difference is that it has an extra argument, which is a pointer to an
2141 integer error code. When there is a compile-time failure, this is set
2142 non-zero, in addition to the error test pointer being set to point to an
2143 error message. The new argument may be NULL if no error number is required
2144 (but then you may as well call pcre_compile(), which is now just a
2145 wrapper). This facility is provided because some applications need a
2146 numeric error indication, but it has also enabled me to tidy up the way
2147 compile-time errors are handled in the POSIX wrapper.
2148
2149 16. Added VPATH=.libs to the makefile; this should help when building with one
2150 prefix path and installing with another. (Or so I'm told by someone who
2151 knows more about this stuff than I do.)
2152
2153 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2154 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2155 match everything, including newlines. This is not POSIX-compatible, but
2156 somebody wanted the feature. From pcretest it can be activated by using
2157 both the P and the s flags.
2158
2159 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2160
2161 19. libpcre.pc was being incorrectly installed as executable.
2162
2163 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2164 it now also looks for '\r' so that it will work unmodified on Windows.
2165
2166 21. Added Google's contributed C++ wrapper to the distribution.
2167
2168 22. Added some untidy missing memory free() calls in pcretest, to keep
2169 Electric Fence happy when testing.
2170
2171
2172
2173 Version 5.0 13-Sep-04
2174 ---------------------
2175
2176 1. Internal change: literal characters are no longer packed up into items
2177 containing multiple characters in a single byte-string. Each character
2178 is now matched using a separate opcode. However, there may be more than one
2179 byte in the character in UTF-8 mode.
2180
2181 2. The pcre_callout_block structure has two new fields: pattern_position and
2182 next_item_length. These contain the offset in the pattern to the next match
2183 item, and its length, respectively.
2184
2185 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2186 insertion of callouts before each pattern item. Added the /C option to
2187 pcretest to make use of this.
2188
2189 4. On the advice of a Windows user, the lines
2190
2191 #if defined(_WIN32) || defined(WIN32)
2192 _setmode( _fileno( stdout ), 0x8000 );
2193 #endif /* defined(_WIN32) || defined(WIN32) */
2194
2195 have been added to the source of pcretest. This apparently does useful
2196 magic in relation to line terminators.
2197
2198 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2199 for the benefit of those environments where the "b" makes a difference.
2200
2201 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2202 to know about it. I have put a hack into configure.in that adds in code
2203 to set GCC=yes if CC=icc. This seems to end up at a point in the
2204 generated configure script that is early enough to affect the setting of
2205 compiler options, which is what is needed, but I have no means of testing
2206 whether it really works. (The user who reported this had patched the
2207 generated configure script, which of course I cannot do.)
2208
2209 LATER: After change 22 below (new libtool files), the configure script
2210 seems to know about icc (and also ecc). Therefore, I have commented out
2211 this hack in configure.in.
2212
2213 7. Added support for pkg-config (2 patches were sent in).
2214
2215 8. Negated POSIX character classes that used a combination of internal tables
2216 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2217 [[:^ascii]]. Typically, they would match almost any characters. The other
2218 POSIX classes were not broken in this way.
2219
2220 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2221 to find the match, as PCRE was deluded into thinking that the match had to
2222 start at the start point or following a newline. The same bug applied to
2223 patterns with negative forward assertions or any backward assertions
2224 preceding ".*" at the start, unless the pattern required a fixed first
2225 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2226
2227 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2228 starting at the last subject character, bytes beyond the end of the subject
2229 string were read.
2230
2231 11. Renamed the variable "class" as "classbits" to make life easier for C++
2232 users. (Previously there was a macro definition, but it apparently wasn't
2233 enough.)
2234
2235 12. Added the new field "tables" to the extra data so that tables can be passed
2236 in at exec time, or the internal tables can be re-selected. This allows
2237 a compiled regex to be saved and re-used at a later time by a different
2238 program that might have everything at different addresses.
2239
2240 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2241 -R library as well as a -L library.
2242
2243 14. The debugging options of pcretest (-d on the command line or D on a
2244 pattern) showed incorrect output for anything following an extended class
2245 that contained multibyte characters and which was followed by a quantifier.
2246
2247 15. Added optional support for general category Unicode character properties
2248 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2249 support. It adds about 90K to the size of the library. The meanings of the
2250 inbuilt class escapes such as \d and \s have NOT been changed.
2251
2252 16. Updated pcredemo.c to include calls to free() to release the memory for the
2253 compiled pattern.
2254
2255 17. The generated file chartables.c was being created in the source directory
2256 instead of in the building directory. This caused the build to fail if the
2257 source directory was different from the building directory, and was
2258 read-only.
2259
2260 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2261 file. No doubt somebody will tell me if they don't make sense... Also added
2262 Dan Mooney's comments about building on OpenVMS.
2263
2264 19. Added support for partial matching via the PCRE_PARTIAL option for
2265 pcre_exec() and the \P data escape in pcretest.
2266
2267 20. Extended pcretest with 3 new pattern features:
2268
2269 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2270 write the compiled pattern to the file whose name is "rest-of-line".
2271 This is a straight binary dump of the data, with the saved pointer to
2272 the character tables forced to be NULL. The study data, if any, is
2273 written too. After writing, pcretest reads a new pattern.
2274
2275 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2276 compiled pattern from the given file. There must not be any
2277 occurrences of "<" in the file name (pretty unlikely); if there are,
2278 pcretest will instead treat the initial "<" as a pattern delimiter.
2279 After reading in the pattern, pcretest goes on to read data lines as
2280 usual.
2281
2282 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2283 and 16-bit fields in a compiled pattern, to simulate a pattern that
2284 was compiled on a host of opposite endianness.
2285
2286 21. The pcre-exec() function can now cope with patterns that were compiled on
2287 hosts of opposite endianness, with this restriction:
2288
2289 As for any compiled expression that is saved and used later, the tables
2290 pointer field cannot be preserved; the extra_data field in the arguments
2291 to pcre_exec() should be used to pass in a tables address if a value
2292 other than the default internal tables were used at compile time.
2293
2294 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2295 now diagnosed as an error. Previously, most of the time, a negative number
2296 would have been treated as zero, but if in addition "ovector" was passed as
2297 NULL, a crash could occur.
2298
2299 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2300 new versions from the libtool 1.5 distribution (the last one is a copy of
2301 a file called libtool.m4). This seems to have fixed the need to patch
2302 "configure" to support Darwin 1.3 (which I used to do). However, I still
2303 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2304 workstation).
2305
2306 24. Changed the PCRE licence to be the more standard "BSD" licence.
2307
2308
2309 Version 4.5 01-Dec-03
2310 ---------------------
2311
2312 1. There has been some re-arrangement of the code for the match() function so
2313 that it can be compiled in a version that does not call itself recursively.
2314 Instead, it keeps those local variables that need separate instances for
2315 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2316 needs to "recurse". Keeping track of where control must go is done by means
2317 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2318 hide most of the details from the main code, and operates only if
2319 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2320 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2321 operating.
2322
2323 To make it easier for callers to provide specially tailored get/free
2324 functions for this usage, two new functions, pcre_stack_malloc, and
2325 pcre_stack_free, are used. They are always called in strict stacking order,
2326 and the size of block requested is always the same.
2327
2328 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2329 PCRE has been compiled to use the stack or the heap for recursion. The
2330 -C option of pcretest uses this to show which version is compiled.
2331
2332 A new data escape \S, is added to pcretest; it causes the amounts of store
2333 obtained and freed by both kinds of malloc/free at match time to be added
2334 to the output.
2335
2336 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2337 what's available on my current Linux desktop machine.
2338
2339 3. When matching a UTF-8 string, the test for a valid string at the start has
2340 been extended. If start_offset is not zero, PCRE now checks that it points
2341 to a byte that is the start of a UTF-8 character. If not, it returns
2342 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2343 this is necessary because there may be backward assertions in the pattern.
2344 When matching the same subject several times, it may save resources to use
2345 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2346
2347 4. The code for checking the validity of UTF-8 strings has been tightened so
2348 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2349 containing "overlong sequences".
2350
2351 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2352 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2353 should have been "&", but it just so happened that all the cases this let
2354 through by mistake were picked up later in the function.
2355
2356 6. I had used a variable called "isblank" - this is a C99 function, causing
2357 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2358
2359 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2360 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2361 so the version is shown; (c) stop "make test" echoing "./RunTest".
2362
2363 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2364
2365 9. The prototype for memmove() for systems that don't have it was using
2366 size_t, but the inclusion of the header that defines size_t was later. I've
2367 moved the #includes for the C headers earlier to avoid this.
2368
2369 10. Added some adjustments to the code to make it easier to compiler on certain
2370 special systems:
2371
2372 (a) Some "const" qualifiers were missing.
2373 (b) Added the macro EXPORT before all exported functions; by default this
2374 is defined to be empty.
2375 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2376 that it reads its output file name as an argument instead of writing
2377 to the standard output and assuming this can be redirected.
2378
2379 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2380 class containing characters with values greater than 255, PCRE compilation
2381 went into a loop.
2382
2383 12. A recursive reference to a subpattern that was within another subpattern
2384 that had a minimum quantifier of zero caused PCRE to crash. For example,
2385 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2386 recursion. If the recursively-called subpattern itself had a zero repeat,
2387 that was OK.
2388
2389 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2390 buffer into which it was copied (for escape processing) was still set at
2391 1024, so long lines caused crashes.
2392
2393 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2394 "internal error: code overflow...". This applied to any character class
2395 that was followed by a possessive quantifier.
2396
2397 15. Modified the Makefile to add libpcre.la as a prerequisite for
2398 libpcreposix.la because I was told this is needed for a parallel build to
2399 work.
2400
2401 16. If a pattern that contained .* following optional items at the start was
2402 studied, the wrong optimizing data was generated, leading to matching
2403 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2404 matching string must start with a or b or c. The correct conclusion for
2405 this pattern is that a match can start with any character.
2406
2407
2408 Version 4.4 13-Aug-03
2409 ---------------------
2410
2411 1. In UTF-8 mode, a character class containing characters with values between
2412 127 and 255 was not handled correctly if the compiled pattern was studied.
2413 In fixing this, I have also improved the studying algorithm for such
2414 classes (slightly).
2415
2416 2. Three internal functions had redundant arguments passed to them. Removal
2417 might give a very teeny performance improvement.
2418
2419 3. Documentation bug: the value of the capture_top field in a callout is *one
2420 more than* the number of the hightest numbered captured substring.
2421
2422 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2423 in incorrectly linking with a previously installed version. They now link
2424 explicitly with libpcre.la.
2425
2426 5. configure.in no longer needs to recognize Cygwin specially.
2427
2428 6. A problem in pcre.in for Windows platforms is fixed.
2429
2430 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2431 pcretest, it used to include the size of the study block as part of its
2432 output. Unfortunately, the structure contains a field that has a different
2433 size on different hardware architectures. This meant that the tests that
2434 showed this size failed. As the block is currently always of a fixed size,
2435 this information isn't actually particularly useful in pcretest output, so
2436 I have just removed it.
2437
2438 8. Three pre-processor statements accidentally did not start in column 1.
2439 Sadly, there are *still* compilers around that complain, even though
2440 standard C has not required this for well over a decade. Sigh.
2441
2442 9. In pcretest, the code for checking callouts passed small integers in the
2443 callout_data field, which is a void * field. However, some picky compilers
2444 complained about the casts involved for this on 64-bit systems. Now
2445 pcretest passes the address of the small integer instead, which should get
2446 rid of the warnings.
2447
2448 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2449 both compile and run time, and gives an error if an invalid UTF-8 sequence
2450 is found. There is a option for disabling this check in cases where the
2451 string is known to be correct and/or the maximum performance is wanted.
2452
2453 11. In response to a bug report, I changed one line in Makefile.in from
2454
2455 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2456 to
2457 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2458
2459 to look similar to other lines, but I have no way of telling whether this
2460 is the right thing to do, as I do not use Windows. No doubt I'll get told
2461 if it's wrong...
2462
2463
2464 Version 4.3 21-May-03
2465 ---------------------
2466
2467 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2468 Makefile.
2469
2470 2. Some refactoring to improve the quality of the code:
2471
2472 (i) The utf8_table... variables are now declared "const".
2473
2474 (ii) The code for \cx, which used the "case flipping" table to upper case
2475 lower case letters, now just substracts 32. This is ASCII-specific,
2476 but the whole concept of \cx is ASCII-specific, so it seems
2477 reasonable.
2478
2479 (iii) PCRE was using its character types table to recognize decimal and
2480 hexadecimal digits in the pattern. This is silly, because it handles
2481 only 0-9, a-f, and A-F, but the character types table is locale-
2482 specific, which means strange things might happen. A private
2483 table is now used for this - though it costs 256 bytes, a table is
2484 much faster than multiple explicit tests. Of course, the standard
2485 character types table is still used for matching digits in subject
2486 strings against \d.
2487
2488 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2489 ending in _t are). So I've renamed it as ESC_tee.
2490
2491 3. The first argument for regexec() in the POSIX wrapper should have been
2492 defined as "const".
2493
2494 4. Changed pcretest to use malloc() for its buffers so that they can be
2495 Electric Fenced for debugging.
2496
2497 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2498 to read one or more bytes before the start of the subject string. Often this
2499 had no effect on PCRE's behaviour, but in some circumstances it could
2500 provoke a segmentation fault.
2501
2502 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2503 to try to read one or more bytes before the start of the subject string.
2504
2505 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2506 UTF-8 support could misbehave in various ways if the subject string
2507 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2508 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2509 back over UTF-8 characters.)
2510
2511
2512 Version 4.2 14-Apr-03
2513 ---------------------
2514
2515 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2516
2517 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2518 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2519 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2520 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2521 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2522 and BUILD_EXEEXT
2523 Note: automatic setting of the BUILD variables is not yet working
2524 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2525 compile-time but not at link-time
2526 [LINK]: use for linking executables only
2527 make different versions for Windows and non-Windows
2528 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2529 libraries
2530 [LINK_FOR_BUILD]: new variable
2531 [OBJEXT]: use throughout
2532 [EXEEXT]: use throughout
2533 <winshared>: new target
2534 <wininstall>: new target
2535 <dftables.o>: use native compiler
2536 <dftables>: use native linker
2537 <install>: handle Windows platform correctly
2538 <clean>: ditto
2539 <check>: ditto
2540 copy DLL to top builddir before testing
2541
2542 As part of these changes, -no-undefined was removed again. This was reported
2543 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2544 in any case.
2545
2546 3. Some tidies to get rid of compiler warnings:
2547
2548 . In the match_data structure, match_limit was an unsigned long int, whereas
2549 match_call_count was an int. I've made them both unsigned long ints.
2550
2551 . In pcretest the fact that a const uschar * doesn't automatically cast to
2552 a void * provoked a warning.
2553
2554 . Turning on some more compiler warnings threw up some "shadow" variables
2555 and a few more missing casts.
2556
2557 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2558 option, a class that contained a single character with a value between 128
2559 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2560
2561 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2562 option, a class that contained several characters, but with at least one
2563 whose value was between 128 and 255 caused PCRE to crash.
2564
2565
2566 Version 4.1 12-Mar-03
2567 ---------------------
2568
2569 1. Compiling with gcc -pedantic found a couple of places where casts were
2570 needed, and a string in dftables.c that was longer than standard compilers are
2571 required to support.
2572
2573 2. Compiling with Sun's compiler found a few more places where the code could
2574 be tidied up in order to avoid warnings.
2575
2576 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2577 first of these names is deprecated in the latest Autoconf in favour of the name
2578 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2579 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2580 analogy I have changed it to CFLAGS_FOR_BUILD.
2581
2582 4. Added -no-undefined to the linking command in the Makefile, because this is
2583 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2584 linking step for the pcreposix library.
2585
2586 5. PCRE was failing to diagnose the case of two named groups with the same
2587 name.
2588
2589 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2590 literal character that is needed in the subject for a match, and scans along to
2591 ensure that it is present before embarking on the full matching process. This
2592 saves time in cases of nested unlimited repeats that are never going to match.
2593 Problem: the scan can take a lot of time if the subject is very long (e.g.
2594 megabytes), thus penalizing straightforward matches. It is now done only if the
2595 amount of subject to be scanned is less than 1000 bytes.
2596
2597 7. A lesser problem with the same optimization is that it was recording the
2598 first character of an anchored pattern as "needed", thus provoking a search
2599 right along the subject, even when the first match of the pattern was going to
2600 fail. The "needed" character is now not set for anchored patterns, unless it
2601 follows something in the pattern that is of non-fixed length. Thus, it still
2602 fulfils its original purpose of finding quick non-matches in cases of nested
2603 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2604
2605
2606 Version 4.0 17-Feb-03
2607 ---------------------
2608
2609 1. If a comment in an extended regex that started immediately after a meta-item
2610 extended to the end of string, PCRE compiled incorrect data. This could lead to
2611 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2612
2613 2. Moved to autoconf 2.53 and libtool 1.4.2.
2614
2615 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2616 the special perltest8 script is no longer needed - all the tests can be run
2617 from a single perltest script.
2618
2619 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2620 by \s. It has now been removed in PCRE. This means it isn't recognized as
2621 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2622 class [:space:] *does* include VT, thereby creating a mess.
2623
2624 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2625 space and tab.
2626
2627 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2628 its new features into the main test script, reducing the number of scripts.
2629
2630 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2631 were backward compatible, and made the (?i) apply to the whole pattern, as if
2632 /i were given. Now it behaves more logically, and applies the option setting
2633 only to what follows. PCRE has been changed to follow suit. However, if it
2634 finds options settings right at the start of the pattern, it extracts them into
2635 the global options, as before. Thus, they show up in the info data.
2636
2637 8. Added support for the \Q...\E escape sequence. Characters in between are
2638 treated as literals. This is slightly different from Perl in that $ and @ are
2639 also handled as literals inside the quotes. In Perl, they will cause variable
2640 interpolation. Note the following examples:
2641
2642 Pattern PCRE matches Perl matches
2643
2644 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2645 \Qabc\$xyz\E abc\$xyz abc\$xyz
2646 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2647
2648 For compatibility with Perl, \Q...\E sequences are recognized inside character
2649 classes as well as outside them.
2650
2651 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2652 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2653 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2654 signed/unsigned warnings.
2655
2656 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2657 option for pcretest, so I've replaced it by a simple function that does just
2658 that job.
2659
2660 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2661 "pcregrep -".
2662
2663 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2664 Java package. This provides some syntactic sugar for simple cases of what my
2665 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2666 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2667 item, you can use this simplified notation. Note that only makes sense with
2668 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2669 greediness, whatever the setting of the PCRE_UNGREEDY option.
2670
2671 13. A change of greediness default within a pattern was not taking effect at
2672 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2673 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2674 was abstracted outside.
2675
2676 14. PCRE now supports the \G assertion. It is true when the current matching
2677 position is at the start point of the match. This differs from \A when the
2678 starting offset is non-zero. Used with the /g option of pcretest (or similar
2679 code), it works in the same way as it does for Perl's /g option. If all
2680 alternatives of a regex begin with \G, the expression is anchored to the start
2681 match position, and the "anchored" flag is set in the compiled expression.
2682
2683 15. Some bugs concerning the handling of certain option changes within patterns
2684 have been fixed. These applied to options other than (?ims). For example,
2685 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2686 been the other way round. Some of this was related to change 7 above.
2687
2688 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2689 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2690 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2691 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2692
2693 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2694 mode. Unlike ".", it always matches newline, whatever the setting of
2695 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2696 assertions. Perl allows it, but it doesn't (in general) work because it can't
2697 calculate the length of the lookbehind. At least, that's the case for Perl
2698 5.8.0 - I've been told they are going to document that it doesn't work in
2699 future.
2700
2701 18. Added an error diagnosis for escapes that PCRE does not support: these are
2702 \L, \l, \N, \P, \p, \U, \u, and \X.
2703
2704 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2705 reading past the end of the pattern in cases such as /[abcd/.
2706
2707 20. PCRE was getting more memory than necessary for patterns with classes that
2708 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2709
2710 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2711 compiling PCRE for use with Virtual Pascal.
2712
2713 22. Small fix to the Makefile to make it work properly if the build is done
2714 outside the source tree.
2715
2716 23. Added a new extension: a condition to go with recursion. If a conditional
2717 subpattern starts with (?(R) the "true" branch is used if recursion has
2718 happened, whereas the "false" branch is used only at the top level.
2719
2720 24. When there was a very long string of literal characters (over 255 bytes
2721 without UTF support, over 250 bytes with UTF support), the computation of how
2722 much memory was required could be incorrect, leading to segfaults or other
2723 strange effects.
2724
2725 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2726 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2727 there was a subsequent back reference to those brackets. This meant that, for
2728 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2729 possible to check for precisely this case. All we can do is abandon the
2730 optimization if .* occurs inside capturing brackets when there are any back
2731 references whatsoever. (See below for a better fix that came later.)
2732
2733 26. The handling of the optimization for finding the first character of a
2734 non-anchored pattern, and for finding a character that is required later in the
2735 match were failing in some cases. This didn't break the matching; it just
2736 failed to optimize when it could. The way this is done has been re-implemented.
2737
2738 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2739
2740 28. Added a new feature that provides some of the functionality that Perl
2741 provides with (?{...}). The facility is termed a "callout". The way it is done
2742 in PCRE is for the caller to provide an optional function, by setting
2743 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2744 global variable. By default it is unset, which disables all calling out. To get
2745 the function called, the regex must include (?C) at appropriate points. This
2746 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2747 This provides a means of identifying different callout points. When PCRE
2748 reaches such a point in the regex, if pcre_callout has been set, the external
2749 function is called. It is provided with data in a structure called
2750 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2751 matching continues; if it returns a non-zero value, the match at the current
2752 point fails. However, backtracking will occur if possible. [This was changed
2753 later and other features added - see item 49 below.]
2754
2755 29. pcretest is upgraded to test the callout functionality. It provides a
2756 callout function that displays information. By default, it shows the start of
2757 the match and the current position in the text. There are some new data escapes
2758 to vary what happens:
2759
2760 \C+ in addition, show current contents of captured substrings
2761 \C- do not supply a callout function
2762 \C!n return 1 when callout number n is reached
2763 \C!n!m return 1 when callout number n is reached for the mth time
2764
2765 30. If pcregrep was called with the -l option and just a single file name, it
2766 output "<stdin>" if a match was found, instead of the file name.
2767
2768 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2769 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2770 pcre_exec(). This saves a malloc/free per call. The default value of
2771 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2772 when configuring.
2773
2774 32. The default maximum size of a compiled pattern is 64K. There have been a
2775 few cases of people hitting this limit. The code now uses macros to handle the
2776 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2777 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2778 configuring. Tests 2 and 5 work only with 2-byte links because they output
2779 debugging information about compiled patterns.
2780
2781 33. Internal code re-arrangements:
2782
2783 (a) Moved the debugging function for printing out a compiled regex into
2784 its own source file (printint.c) and used #include to pull it into
2785 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2786 separate copies.
2787
2788 (b) Defined the list of op-code names for debugging as a macro in
2789 internal.h so that it is next to the definition of the opcodes.
2790
2791 (c) Defined a table of op-code lengths for simpler skipping along compiled
2792 code. This is again a macro in internal.h so that it is next to the
2793 definition of the opcodes.
2794
2795 34. Added support for recursive calls to individual subpatterns, along the
2796 lines of Robin Houston's patch (but implemented somewhat differently).
2797
2798 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2799 allow it to read and process whole directories in Win32. This code was
2800 contributed by Lionel Fourquaux; it has not been tested by me.
2801
2802 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2803 used to name a group. Names consist of alphanumerics and underscores, and must
2804 be unique. Back references use the syntax (?P=name) and recursive calls use
2805 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2806 numbers. The function pcre_fullinfo() can be used after compilation to extract
2807 a name/number map. There are three relevant calls:
2808
2809 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2810 PCRE_INFO_NAMECOUNT yields the number of entries
2811 PCRE_INFO_NAMETABLE yields a pointer to the map.
2812
2813 The map is a vector of fixed-size entries. The size of each entry depends on
2814 the length of the longest name used. The first two bytes of each entry are the
2815 group number, most significant byte first. There follows the corresponding
2816 name, zero terminated. The names are in alphabetical order.
2817
2818 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2819 case instead of 255. Making it the same both with and without UTF-8 support
2820 means that the same test output works with both.
2821
2822 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2823 calling malloc() with a zero argument.
2824
2825 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2826 optimization. I've improved things by keeping a bitmap of backreferences with
2827 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2828 fact referenced, the optimization can be applied. It is unlikely that a
2829 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2830 the match to follow \n) will appear inside brackets with a number greater than
2831 31, but if it does, any back reference > 31 suppresses the optimization.
2832
2833 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2834 of disabling numbered capturing parentheses. Any opening parenthesis that is
2835 not followed by ? behaves as if it were followed by ?: but named parentheses
2836 can still be used for capturing (and they will acquire numbers in the usual
2837 way).
2838
2839 41. Redesigned the return codes from the match() function into yes/no/error so
2840 that errors can be passed back from deep inside the nested calls. A malloc
2841 failure while inside a recursive subpattern call now causes the
2842 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2843
2844 42. It is now possible to set a limit on the number of times the match()
2845 function is called in a call to pcre_exec(). This facility makes it possible to
2846 limit the amount of recursion and backtracking, though not in a directly
2847 obvious way, because the match() function is used in a number of different
2848 circumstances. The count starts from zero for each position in the subject
2849 string (for non-anchored patterns). The default limit is, for compatibility, a
2850 large number, namely 10 000 000. You can change this in two ways:
2851
2852 (a) When configuring PCRE before making, you can use --with-match-limit=n
2853 to set a default value for the compiled library.
2854
2855 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2856 a different value is set. See 45 below.
2857
2858 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2859
2860 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2861 of things that can be changed at compile time. The first argument specifies
2862 what is wanted and the second points to where the information is to be placed.
2863 The current list of available information is:
2864
2865 PCRE_CONFIG_UTF8
2866
2867 The output is an integer that is set to one if UTF-8 support is available;
2868 otherwise it is set to zero.
2869
2870 PCRE_CONFIG_NEWLINE
2871
2872 The output is an integer that it set to the value of the code that is used for
2873 newline. It is either LF (10) or CR (13).
2874
2875 PCRE_CONFIG_LINK_SIZE
2876
2877 The output is an integer that contains the number of bytes used for internal
2878 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2879
2880 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2881
2882 The output is an integer that contains the threshold above which the POSIX
2883 interface uses malloc() for output vectors. See item 31 above.
2884
2885 PCRE_CONFIG_MATCH_LIMIT
2886
2887 The output is an unsigned integer that contains the default limit of the number
2888 of match() calls in a pcre_exec() execution. See 42 above.
2889
2890 44. pcretest has been upgraded by the addition of the -C option. This causes it
2891 to extract all the available output from the new pcre_config() function, and to
2892 output it. The program then exits immediately.
2893
2894 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2895 order to support additional features. One way would have been to define
2896 pcre_exec2() (for example) with extra arguments, but this would not have been
2897 extensible, and would also have required all calls to the original function to
2898 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2899 is used for passing in "extra" data from pcre_study().
2900
2901 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2902 contains the following fields:
2903
2904 flags a bitmap indicating which of the following fields are set
2905 study_data opaque data from pcre_study()
2906 match_limit a way of specifying a limit on match() calls for a specific
2907 call to pcre_exec()
2908 callout_data data for callouts (see 49 below)
2909
2910 The flag bits are also defined in pcre.h, and are
2911
2912 PCRE_EXTRA_STUDY_DATA
2913 PCRE_EXTRA_MATCH_LIMIT
2914 PCRE_EXTRA_CALLOUT_DATA
2915
2916 The pcre_study() function now returns one of these new pcre_extra blocks, with
2917 the actual study data pointed to by the study_data field, and the
2918 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2919 before. That is, this change is entirely upwards-compatible and requires no
2920 change to existing code.
2921
2922 If you want to pass in additional data to pcre_exec(), you can either place it
2923 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2924 block.
2925
2926 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2927 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2928 times with different match limits, until it finds the minimum value needed for
2929 pcre_exec() to complete. The value is then output. This can be instructive; for
2930 most simple matches the number is quite small, but for pathological cases it
2931 gets very large very quickly.
2932
2933 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2934 returns the size of the data block pointed to by the study_data field in a
2935 pcre_extra block, that is, the value that was passed as the argument to
2936 pcre_malloc() when PCRE was getting memory in which to place the information
2937 created by pcre_study(). The fourth argument should point to a size_t variable.
2938 pcretest has been extended so that this information is shown after a successful
2939 pcre_study() call when information about the compiled regex is being displayed.
2940
2941 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2942 because what follows is always an absolute path. (Later: it turns out that this
2943 is more than cosmetic for MinGW, because it doesn't like empty path
2944 components.)
2945
2946 49. Some changes have been made to the callout feature (see 28 above):
2947
2948 (i) A callout function now has three choices for what it returns:
2949
2950 0 => success, carry on matching
2951 > 0 => failure at this point, but backtrack if possible
2952 < 0 => serious error, return this value from pcre_exec()
2953
2954 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2955 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2956 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2957 use by callout functions. It will never be used by PCRE itself.
2958
2959 (ii) The pcre_extra structure (see 45 above) has a void * field called
2960 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2961 pcre_callout_block structure has a field of the same name. The contents of
2962 the field passed in the pcre_extra structure are passed to the callout
2963 function in the corresponding field in the callout block. This makes it
2964 easier to use the same callout-containing regex from multiple threads. For
2965 testing, the pcretest program has a new data escape
2966
2967 \C*n pass the number n (may be negative) as callout_data
2968
2969 If the callout function in pcretest receives a non-zero value as
2970 callout_data, it returns that value.
2971
2972 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2973 there were some redundant $(CFLAGS) in commands that are now specified as
2974 $(LINK), which already includes $(CFLAGS).
2975
2976 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2977 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2978 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2979 one-byte characters throughout. Note that case-insensitive matching applies
2980 only to characters whose values are less than 256. PCRE doesn't support the
2981 notion of cases for higher-valued characters.
2982
2983 (i) A character class whose characters are all within 0-255 is handled as
2984 a bit map, and the map is inverted for negative classes. Previously, a
2985 character > 255 always failed to match such a class; however it should
2986 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2987
2988 (ii) A negated character class with a single character < 255 is coded as
2989 "not this character" (OP_NOT). This wasn't working properly when the test
2990 character was multibyte, either singly or repeated.
2991
2992 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2993 mode, for example: \x{100}{2,3}.
2994
2995 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2996 singly or repeated) now correctly test multibyte characters. However,
2997 PCRE doesn't recognize any characters with values greater than 255 as
2998 digits, spaces, or word characters. Such characters always match \D, \S,
2999 and \W, and never match \d, \s, or \w.
3000
3001 (v) Classes may now contain characters and character ranges with values
3002 greater than 255. For example: [ab\x{100}-\x{400}].
3003
3004 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3005 PCRE in UTF-8 mode.
3006
3007 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3008 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3009 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3010 value.)
3011
3012 53. The single man page has become too large. I have therefore split it up into
3013 a number of separate man pages. These also give rise to individual HTML pages;
3014 these are now put in a separate directory, and there is an index.html page that
3015 lists them all. Some hyperlinking between the pages has been installed.
3016
3017 54. Added convenience functions for handling named capturing parentheses.
3018
3019 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3020 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3021 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3022 are faulted.
3023
3024 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3025 calling configure. These values are used when compiling the dftables.c program
3026 which is run to generate the source of the default character tables. They
3027 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3028 you will need to set these values.
3029
3030 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3031
3032
3033 Version 3.9 02-Jan-02
3034 ---------------------
3035
3036 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3037
3038 2. If --disable-static was given, the building process failed when trying to
3039 build pcretest and pcregrep. (For some reason it was using libtool to compile
3040 them, which is not right, as they aren't part of the library.)
3041
3042
3043 Version 3.8 18-Dec-01
3044 ---------------------
3045
3046 1. The experimental UTF-8 code was completely screwed up. It was packing the
3047 bytes in the wrong order. How dumb can you get?
3048
3049
3050 Version 3.7 29-Oct-01
3051 ---------------------
3052
3053 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3054 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3055 this didn't happen under Solaris 8, where I normally test things.
3056
3057 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3058 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3059 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3060 this fix an hour or so after the initial 3.7 release.)
3061
3062
3063 Version 3.6 23-Oct-01
3064 ---------------------
3065
3066 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3067 offsets passed as NULL with zero offset count.
3068
3069 2. The config.guess and config.sub files had not been updated when I moved to
3070 the latest autoconf.
3071
3072
3073 Version 3.5 15-Aug-01
3074 ---------------------
3075
3076 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3077 had been forgotten.
3078
3079 2. By using declared but undefined structures, we can avoid using "void"
3080 definitions in pcre.h while keeping the internal definitions of the structures
3081 private.
3082
3083 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3084 user point of view, this means that both static and shared libraries are built
3085 by default, but this can be individually controlled. More of the work of
3086 handling this static/shared cases is now inside libtool instead of PCRE's make
3087 file.
3088
3089 4. The pcretest utility is now installed along with pcregrep because it is
3090 useful for users (to test regexs) and by doing this, it automatically gets
3091 relinked by libtool. The documentation has been turned into a man page, so
3092 there are now .1, .txt, and .html versions in /doc.
3093
3094 5. Upgrades to pcregrep:
3095 (i) Added long-form option names like gnu grep.
3096 (ii) Added --help to list all options with an explanatory phrase.
3097 (iii) Added -r, --recursive to recurse into sub-directories.
3098 (iv) Added -f, --file to read patterns from a file.
3099
3100 6. pcre_exec() was referring to its "code" argument before testing that
3101 argument for NULL (and giving an error if it was NULL).
3102
3103 7. Upgraded Makefile.in to allow for compiling in a different directory from
3104 the source directory.
3105
3106 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3107 options bits, the pointer it was passed was to an int instead of to an unsigned
3108 long int. This mattered only on 64-bit systems.
3109
3110 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3111 generated) instead of pcre.in, which it its source. Also made the same change
3112 in several of the .c files.
3113
3114 10. A new release of gcc defines printf() as a macro, which broke pcretest
3115 because it had an ifdef in the middle of a string argument for printf(). Fixed
3116 by using separate calls to printf().
3117
3118 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3119 script, to force use of CR or LF instead of \n in the source. On non-Unix
3120 systems, the value can be set in config.h.
3121
3122 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3123 absolute limit. Changed the text of the error message to make this clear, and
3124 likewise updated the man page.
3125
3126 13. The limit of 99 on the number of capturing subpatterns has been removed.
3127 The new limit is 65535, which I hope will not be a "real" limit.
3128
3129
3130 Version 3.4 22-Aug-00
3131 ---------------------
3132
3133 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3134
3135 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3136
3137
3138 Version 3.3 01-Aug-00
3139 ---------------------
3140
3141 1. If an octal character was given, but the value was greater than \377, it
3142 was not getting masked to the least significant bits, as documented. This could
3143 lead to crashes in some systems.
3144
3145 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3146 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3147
3148 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3149 These just pass their arguments on to (pcre_free)(), but they are provided
3150 because some uses of PCRE bind it to non-C systems that can call its functions,
3151 but cannot call free() or pcre_free() directly.
3152
3153 4. Add "make test" as a synonym for "make check". Corrected some comments in
3154 the Makefile.
3155
3156 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3157 Makefile.
3158
3159 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3160 command called pgrep for grepping around the active processes.
3161
3162 7. Added the beginnings of support for UTF-8 character strings.
3163
3164 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3165 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3166 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3167 out for the ar command.)
3168
3169
3170 Version 3.2 12-May-00
3171 ---------------------
3172
3173 This is purely a bug fixing release.
3174
3175 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3176 of ZA. This was just one example of several cases that could provoke this bug,
3177 which was introduced by change 9 of version 2.00. The code for breaking
3178 infinite loops after an iteration that matches an empty string was't working
3179 correctly.
3180
3181 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3182 when matched against abbab (for example). After matching an empty string, it
3183 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3184 caused it to match further down the string than it should.
3185
3186 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3187 was there because it doesn't seem to be needed, and it causes trouble on some
3188 systems, as it is not a Standard C header. It has been removed.
3189
3190 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3191 were reported on the Macintosh. The changes were from
3192
3193 while ((c = *(++ptr)) != 0 && c != '\n');
3194 to
3195 while ((c = *(++ptr)) != 0 && c != '\n') ;
3196
3197 Totally extraordinary, but if that's what it takes...
3198
3199 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3200 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3201 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3202 assumes the way PCRE uses memmove() (always moving upwards).
3203
3204 6. PCRE is being used in one environment where strchr() is not available. There
3205 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3206 faster code anyway.
3207
3208
3209 Version 3.1 09-Feb-00
3210 ---------------------
3211
3212 The only change in this release is the fixing of some bugs in Makefile.in for
3213 the "install" target:
3214
3215 (1) It was failing to install pcreposix.h.
3216
3217 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3218
3219
3220 Version 3.0 01-Feb-00
3221 ---------------------
3222
3223 1. Add support for the /+ modifier to perltest (to output $` like it does in
3224 pcretest).
3225
3226 2. Add support for the /g modifier to perltest.
3227
3228 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3229 matches null strings.
3230
3231 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3232 pattern. Perl treats empty patterns specially - it reuses the most recent
3233 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3234 effect.
3235
3236 5. The POSIX interface was broken in that it was just handing over the POSIX
3237 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3238 required a bigger vector, with some working space on the end. This means that
3239 the POSIX wrapper now has to get and free some memory, and copy the results.
3240
3241 6. Added some simple autoconf support, placing the test data and the
3242 documentation in separate directories, re-organizing some of the
3243 information files, and making it build pcre-config (a GNU standard). Also added
3244 libtool support for building PCRE as a shared library, which is now the
3245 default.
3246
3247 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3248 09 are not valid octal constants. Single digits will be used for minor values
3249 less than 10.
3250
3251 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3252 existing programs that set these in the POSIX interface can use PCRE without
3253 modification.
3254
3255 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3256 return all that pcre_info() returns, plus additional data. The pcre_info()
3257 function is retained for compatibility, but is considered to be obsolete.
3258
3259 10. Added experimental recursion feature (?R) to handle one common case that
3260 Perl 5.6 will be able to do with (?p{...}).
3261
3262 11. Added support for POSIX character classes like [:alpha:], which Perl is
3263 adopting.
3264
3265
3266 Version 2.08 31-Aug-99
3267 ----------------------
3268
3269 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3270 trying to match at the startoffset position, but instead was moving forward to
3271 the next newline as if a previous match had failed.
3272
3273 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3274 and could get into a loop if a null string was matched other than at the start
3275 of the subject.
3276
3277 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3278 be distinguished at compile time, and for completeness also added PCRE_DATE.
3279
3280 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3281 in GnuWin32 environments.
3282
3283
3284 Version 2.07 29-Jul-99
3285 ----------------------
3286
3287 1. The documentation is now supplied in plain text form and HTML as well as in
3288 the form of man page sources.
3289
3290 2. C++ compilers don't like assigning (void *) values to other pointer types.
3291 In particular this affects malloc(). Although there is no problem in Standard
3292 C, I've put in casts to keep C++ compilers happy.
3293
3294 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3295 should be (const char *).
3296
3297 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3298 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3299 However, I haven't made this a standard facility. The documentation doesn't
3300 mention it, and the Makefile doesn't support it.
3301
3302 5. The Makefile now contains an "install" target, with editable destinations at
3303 the top of the file. The pcretest program is not installed.
3304
3305 6. pgrep -V now gives the PCRE version number and date.
3306
3307 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3308 causing the entire string to be ignored, instead of just the last character.
3309
3310 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3311 non-matching string, it can take a very, very long time, even for strings of
3312 quite modest length, because of the nested recursion. PCRE now does better in
3313 some of these cases. It does this by remembering the last required literal
3314 character in the pattern, and pre-searching the subject to ensure it is present
3315 before running the real match. In other words, it applies a heuristic to detect
3316 some types of certain failure quickly, and in the above example, if presented
3317 with a string that has no trailing " it gives "no match" very quickly.
3318
3319 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3320 other alternatives are tried instead.
3321
3322
3323 Version 2.06 09-Jun-99
3324 ----------------------
3325
3326 1. Change pcretest's output for amount of store used to show just the code
3327 space, because the remainder (the data block) varies in size between 32-bit and
3328 64-bit systems.
3329
3330 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3331 start matching at. This allows lookbehinds to work when searching for multiple
3332 occurrences in a string.
3333
3334 3. Added additional options to pcretest for testing multiple occurrences:
3335
3336 /+ outputs the rest of the string that follows a match
3337 /g loops for multiple occurrences, using the new startoffset argument
3338 /G loops for multiple occurrences by passing an incremented pointer
3339
3340 4. PCRE wasn't doing the "first character" optimization for patterns starting
3341 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3342 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3343 the letter 'x'. On long subject strings, this gives a significant speed-up.
3344
3345
3346 Version 2.05 21-Apr-99
3347 ----------------------
3348
3349 1. Changed the type of magic_number from int to long int so that it works
3350 properly on 16-bit systems.
3351
3352 2. Fixed a bug which caused patterns starting with .* not to work correctly
3353 when the subject string contained newline characters. PCRE was assuming
3354 anchoring for such patterns in all cases, which is not correct because .* will
3355 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3356 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3357 must be retried after every newline in the subject.
3358
3359
3360 Version 2.04 18-Feb-99
3361 ----------------------
3362
3363 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3364 computation of the store needed to hold the pattern was incorrect (too large).
3365 If such patterns were nested a few deep, this could multiply and become a real
3366 problem.
3367
3368 2. Added /M option to pcretest to show the memory requirement of a specific
3369 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3370
3371 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3372 compiled in such a way that the backtracking after subsequent failure was
3373 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3374 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3375
3376
3377 Version 2.03 02-Feb-99
3378 ----------------------
3379
3380 1. Fixed typo and small mistake in man page.
3381
3382 2. Added 4th condition (GPL supersedes if conflict) and created separate
3383 LICENCE file containing the conditions.
3384
3385 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3386 Perl, that is the internal \ allows the delimiter to be included in the
3387 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3388 the final delimiter, add \ to the end of the pattern (to test the error).
3389
3390 4. Added the convenience functions for extracting substrings after a successful
3391 match. Updated pcretest to make it able to test these functions.
3392
3393
3394 Version 2.02 14-Jan-99
3395 ----------------------
3396
3397 1. Initialized the working variables associated with each extraction so that
3398 their saving and restoring doesn't refer to uninitialized store.
3399
3400 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3401 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3402 fix the problem.
3403
3404 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3405 calls, and wasn't printing the correct value for compiling calls. Increased the
3406 default value of LOOPREPEAT, and the number of significant figures in the
3407 times.
3408
3409 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3410
3411 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3412 a building problem on Windows NT with a FAT file system.
3413
3414
3415 Version 2.01 21-Oct-98
3416 ----------------------
3417
3418 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3419 to character tables built by pcre_maketables() in the current locale. If NULL
3420 is passed, the default tables are used.
3421
3422
3423 Version 2.00 24-Sep-98
3424 ----------------------
3425
3426 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3427 it any more.
3428
3429 2. Allow quantification of (?>) groups, and make it work correctly.
3430
3431 3. The first character computation wasn't working for (?>) groups.
3432
3433 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3434 end of the subject) and add 5.005's \z, which really does match only at the
3435 very end of the subject.
3436
3437 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3438
3439 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3440 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3441 localized options. All options to pcre_study() were also removed.
3442
3443 7. Add other new features from 5.005:
3444
3445 $(?<= positive lookbehind
3446 $(?<! negative lookbehind
3447 (?imsx-imsx) added the unsetting capability
3448 such a setting is global if at outer level; local otherwise
3449 (?imsx-imsx:) non-capturing groups with option setting
3450 (?(cond)re|re) conditional pattern matching
3451
3452 A backreference to itself in a repeated group matches the previous
3453 captured string.
3454
3455 8. General tidying up of studying (both automatic and via "study")
3456 consequential on the addition of new assertions.
3457
3458 9. As in 5.005, unlimited repeated groups that could match an empty substring
3459 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3460 runtime if any iteration does actually match an empty substring.
3461
3462 10. Include the RunTest script in the distribution.
3463
3464 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3465 discrepancies, some of which were old and were also with respect to 5.004. They
3466 have now been fixed.
3467
3468
3469 Version 1.09 28-Apr-98
3470 ----------------------
3471
3472 1. A negated single character class followed by a quantifier with a minimum
3473 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3474 program crashes, or just wrong answers. This did not apply to negated classes
3475 containing more than one character, or to minima other than one.
3476
3477
3478 Version 1.08 27-Mar-98
3479 ----------------------
3480
3481 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3482
3483 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3484 latter must appear before anything that relies on it in the pattern.
3485
3486
3487 Version 1.07 16-Feb-98
3488 ----------------------
3489
3490 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3491 repeat of a potentially empty string).
3492
3493
3494 Version 1.06 23-Jan-98
3495 ----------------------
3496
3497 1. Added Markus Oberhumer's little patches for C++.
3498
3499 2. Literal strings longer than 255 characters were broken.
3500
3501
3502 Version 1.05 23-Dec-97
3503 ----------------------
3504
3505 1. Negated character classes containing more than one character were failing if
3506 PCRE_CASELESS was set at run time.
3507
3508
3509 Version 1.04 19-Dec-97
3510 ----------------------
3511
3512 1. Corrected the man page, where some "const" qualifiers had been omitted.
3513
3514 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3515 input syntax.
3516
3517 3. Fixed memory leak which occurred when a regex with back references was
3518 matched with an offsets vector that wasn't big enough. The temporary memory
3519 that is used in this case wasn't being freed if the match failed.
3520
3521 4. Tidied pcretest to ensure it frees memory that it gets.
3522
3523 5. Temporary memory was being obtained in the case where the passed offsets
3524 vector was exactly big enough.
3525
3526 6. Corrected definition of offsetof() from change 5 below.
3527
3528 7. I had screwed up change 6 below and broken the rules for the use of
3529 setjmp(). Now fixed.
3530
3531
3532 Version 1.03 18-Dec-97
3533 ----------------------
3534
3535 1. A erroneous regex with a missing opening parenthesis was correctly
3536 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3537 on some systems.
3538
3539 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3540 it was reported that one broken compiler failed on the former because "code" is
3541 also an independent variable.
3542
3543 3. The erroneous regex a[]b caused an array overrun reference.
3544
3545 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3546 fail on data ending with that character. (It was going on too far, and checking
3547 the next character, typically a binary zero.) This was specific to the
3548 optimized code for single-character negative classes.
3549
3550 5. Added a contributed patch from the TIN world which does the following:
3551
3552 + Add an undef for memmove, in case the the system defines a macro for it.
3553
3554 + Add a definition of offsetof(), in case there isn't one. (I don't know
3555 the reason behind this - offsetof() is part of the ANSI standard - but
3556 it does no harm).
3557
3558 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3559 most of the places where whitespace preceded '#'. I have given up and
3560 allowed the remaining 2 cases to be at the margin.
3561
3562 + Rename some variables in pcre to eliminate shadowing. This seems very
3563 pedantic, but does no harm, of course.
3564
3565 6. Moved the call to setjmp() into its own function, to get rid of warnings
3566 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3567
3568 7. Constructs such as \d{8,} were compiling into the equivalent of
3569 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3570 outcome, but in this particular case used more store than had been allocated,
3571 which caused the bug to be discovered because it threw up an internal error.
3572
3573 8. The debugging code in both pcre and pcretest for outputting the compiled
3574 form of a regex was going wrong in the case of back references followed by
3575 curly-bracketed repeats.
3576
3577
3578 Version 1.02 12-Dec-97
3579 ----------------------
3580
3581 1. Typos in pcre.3 and comments in the source fixed.
3582
3583 2. Applied a contributed patch to get rid of places where it used to remove
3584 'const' from variables, and fixed some signed/unsigned and uninitialized
3585 variable warnings.
3586
3587 3. Added the "runtest" target to Makefile.
3588
3589 4. Set default compiler flag to -O2 rather than just -O.
3590
3591
3592 Version 1.01 19-Nov-97
3593 ----------------------
3594
3595 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3596 like /([ab]*)*/, that is, for classes with more than one character in them.
3597
3598 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3599 as /((?>a*))*/ (a PCRE_EXTRA facility).
3600
3601
3602 Version 1.00 18-Nov-97
3603 ----------------------
3604
3605 1. Added compile-time macros to support systems such as SunOS4 which don't have
3606 memmove() or strerror() but have other things that can be used instead.
3607
3608 2. Arranged that "make clean" removes the executables.
3609
3610
3611 Version 0.99 27-Oct-97
3612 ----------------------
3613
3614 1. Fixed bug in code for optimizing classes with only one character. It was
3615 initializing a 32-byte map regardless, which could cause it to run off the end
3616 of the memory it had got.
3617
3618 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3619
3620
3621 Version 0.98 22-Oct-97
3622 ----------------------
3623
3624 1. Fixed bug in code for handling temporary memory usage when there are more
3625 back references than supplied space in the ovector. This could cause segfaults.
3626
3627
3628 Version 0.97 21-Oct-97
3629 ----------------------
3630
3631 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3632
3633 2. Optimized negated single characters not to use a bit map.
3634
3635 3. Brought error texts together as macro definitions; clarified some of them;
3636 fixed one that was wrong - it said "range out of order" when it meant "invalid
3637 escape sequence".
3638
3639 4. Changed some char * arguments to const char *.
3640
3641 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3642
3643 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3644 pcretest.
3645
3646
3647 Version 0.96 16-Oct-97
3648 ----------------------
3649
3650 1. Added a simple "pgrep" utility to the distribution.
3651
3652 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3653 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3654 where "ddd" means "one or more decimal digits".
3655
3656 3. Fixed serious bug. If a pattern had a back reference, but the call to
3657 pcre_exec() didn't supply a large enough ovector to record the related
3658 identifying subpattern, the match always failed. PCRE now remembers the number
3659 of the largest back reference, and gets some temporary memory in which to save
3660 the offsets during matching if necessary, in order to ensure that
3661 backreferences always work.
3662
3663 4. Increased the compatibility with Perl in a number of ways:
3664
3665 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3666 to request this handling. The option can be set at compile or exec time.
3667
3668 (b) $ matches before a terminating newline by default; an option
3669 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3670 mode). The option can be set at compile or exec time.
3671
3672 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3673 the same as Perl's. If the decimal number it represents is less than 10
3674 or there aren't that many previous left capturing parentheses, an octal
3675 escape is read. Inside a character class, it's always an octal escape,
3676 even if it is a single digit.
3677
3678 (d) An escaped but undefined alphabetic character is taken as a literal,
3679 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3680 escapes.
3681
3682 (e) {0} is now permitted. (The previous item is removed from the compiled
3683 pattern).
3684
3685 5. Changed all the names of code files so that the basic parts are no longer
3686 than 10 characters, and abolished the teeny "globals.c" file.
3687
3688 6. Changed the handling of character classes; they are now done with a 32-byte
3689 bit map always.
3690
3691 7. Added the -d and /D options to pcretest to make it possible to look at the
3692 internals of compilation without having to recompile pcre.
3693
3694
3695 Version 0.95 23-Sep-97
3696 ----------------------
3697
3698 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3699 \x20 at the start of a run of normal characters. These were being treated as
3700 real characters, instead of the source characters being re-checked.
3701
3702
3703 Version 0.94 18-Sep-97
3704 ----------------------
3705
3706 1. The functions are now thread-safe, with the caveat that the global variables
3707 containing pointers to malloc() and free() or alternative functions are the
3708 same for all threads.
3709
3710 2. Get pcre_study() to generate a bitmap of initial characters for non-
3711 anchored patterns when this is possible, and use it if passed to pcre_exec().
3712
3713
3714 Version 0.93 15-Sep-97
3715 ----------------------
3716
3717 1. /(b)|(:+)/ was computing an incorrect first character.
3718
3719 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3720 but not actually doing anything yet.
3721
3722 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3723 as Perl does (e.g. [-az] or [az-]).
3724
3725 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3726 all possible positions.
3727
3728 5. Split up into different modules to avoid including unneeded functions in a
3729 compiled binary. However, compile and exec are still in one module. The "study"
3730 function is split off.
3731
3732 6. The character tables are now in a separate module whose source is generated
3733 by an auxiliary program - but can then be edited by hand if required. There are
3734 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3735 toupper() in the code.
3736
3737 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3738 make them global. Abolish the function for setting them, as the caller can now
3739 set them directly.
3740
3741
3742 Version 0.92 11-Sep-97
3743 ----------------------
3744
3745 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3746 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3747
3748 2. Caseless matching was not working in character classes if the characters in
3749 the pattern were in upper case.
3750
3751 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3752
3753 4. Make PCRE_ANCHORED public and accept as a compile option.
3754
3755 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3756 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3757 pass them.
3758
3759 6. Give an error if bad option bits passed at compile or run time.
3760
3761 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3762 pcretest to cause it to pass that flag.
3763
3764 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3765 options, and the first character, if set.
3766
3767 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3768
3769
3770 Version 0.91 10-Sep-97
3771 ----------------------
3772
3773 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3774 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3775
3776 2. PCRE was looping on encountering an indefinitely repeated back reference to
3777 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3778 Perl does - treats the match as successful.
3779
3780 ****

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