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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 517 - (show annotations) (download)
Wed May 5 10:44:20 2010 UTC (4 years, 3 months ago) by ph10
File size: 186132 byte(s)
Add new special properties Xan, Xps, Xsp, Xwd to help with \w etc.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.10 03 May-2010
5 ------------------------
6
7 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
8 THEN.
9
10 2. (*ACCEPT) was not working when inside an atomic group.
11
12 3. Inside a character class, \B is treated as a literal by default, but
13 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14 causes the error). The code is unchanged, but I tidied the documentation.
15
16 4. Inside a character class, PCRE always treated \R and \X as literals,
17 whereas Perl faults them if its -w option is set. I have changed PCRE so
18 that it faults them when PCRE_EXTRA is set.
19
20 5. Added support for \N, which always matches any character other than
21 newline. (It is the same as "." when PCRE_DOTALL is not set.)
22
23 6. When compiling pcregrep with newer versions of gcc which may have
24 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25 declared with attribute warn_unused_result" were given. Just casting the
26 result to (void) does not stop the warnings; a more elaborate fudge is
27 needed. I've used a macro to implement this.
28
29 7. Minor change to pcretest.c to avoid a compiler warning.
30
31 8. Added four artifical Unicode properties to help with an option to make
32 \s etc use properties. The new properties are: Xan (alphanumeric), Xsp
33 (Perl space), Xps (POSIX space), and Xwd (word).
34
35
36 Version 8.02 19-Mar-2010
37 ------------------------
38
39 1. The Unicode data tables have been updated to Unicode 5.2.0.
40
41 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
42 configured.
43
44 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
45 original author of that file, following a query about its status.
46
47 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
48 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
49
50 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
51 quantifier applied to a forward-referencing subroutine call, could compile
52 incorrect code or give the error "internal error: previously-checked
53 referenced subpattern not found".
54
55 6. Both MS Visual Studio and Symbian OS have problems with initializing
56 variables to point to external functions. For these systems, therefore,
57 pcre_malloc etc. are now initialized to local functions that call the
58 relevant global functions.
59
60 7. There were two entries missing in the vectors called coptable and poptable
61 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
62 I've fixed the data, and added a kludgy way of testing at compile time that
63 the lengths are correct (equal to the number of opcodes).
64
65 8. Following on from 7, I added a similar kludge to check the length of the
66 eint vector in pcreposix.c.
67
68 9. Error texts for pcre_compile() are held as one long string to avoid too
69 much relocation at load time. To find a text, the string is searched,
70 counting zeros. There was no check for running off the end of the string,
71 which could happen if a new error number was added without updating the
72 string.
73
74 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
75
76 11. \K was not working if it appeared in an atomic group or in a group that
77 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
78 \K is "not well defined" if used in an assertion. PCRE now accepts it if
79 the assertion is positive, but not if it is negative.
80
81 12. Change 11 fortuitously reduced the size of the stack frame used in the
82 "match()" function of pcre_exec.c by one pointer. Forthcoming
83 implementation of support for (*MARK) will need an extra pointer on the
84 stack; I have reserved it now, so that the stack frame size does not
85 decrease.
86
87 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
88 item in branch that calls a recursion is a subroutine call - as in the
89 second branch in the above example - was incorrectly given the compile-
90 time error "recursive call could loop indefinitely" because pcre_compile()
91 was not correctly checking the subroutine for matching a non-empty string.
92
93 14. The checks for overrunning compiling workspace could trigger after an
94 overrun had occurred. This is a "should never occur" error, but it can be
95 triggered by pathological patterns such as hundreds of nested parentheses.
96 The checks now trigger 100 bytes before the end of the workspace.
97
98 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
99
100
101 Version 8.01 19-Jan-2010
102 ------------------------
103
104 1. If a pattern contained a conditional subpattern with only one branch (in
105 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
106 computed the wrong minimum data length (which is of course zero for such
107 subpatterns). This could cause incorrect "no match" results.
108
109 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
110 the pattern is reset in the first branch, pcre_compile() failed with
111 "internal error: code overflow at offset...". This happened only when
112 the reset was to the original external option setting. (An optimization
113 abstracts leading options settings into an external setting, which was the
114 cause of this.)
115
116 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
117 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
118 assertion pattern did not match (meaning that the assertion was true), it
119 was incorrectly treated as false if the SKIP had been reached during the
120 matching. This also applied to assertions used as conditions.
121
122 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
123 assertion subpattern, including such a pattern used as a condition,
124 unpredictable results occurred, instead of the error return
125 PCRE_ERROR_DFA_UITEM.
126
127 5. The C++ GlobalReplace function was not working like Perl for the special
128 situation when an empty string is matched. It now does the fancy magic
129 stuff that is necessary.
130
131 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
132 removed. (These were left over from very, very early versions of PCRE.)
133
134 7. Some cosmetic changes to the code to make life easier when compiling it
135 as part of something else:
136
137 (a) Change DEBUG to PCRE_DEBUG.
138
139 (b) In pcre_compile(), rename the member of the "branch_chain" structure
140 called "current" as "current_branch", to prevent a collision with the
141 Linux macro when compiled as a kernel module.
142
143 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
144 prevent a collision with the Linux macro when compiled as a kernel
145 module.
146
147 8. In pcre_compile() there are some checks for integer overflows that used to
148 cast potentially large values to (double). This has been changed to that
149 when building, a check for int64_t is made, and if it is found, it is used
150 instead, thus avoiding the use of floating point arithmetic. (There is no
151 other use of FP in PCRE.) If int64_t is not found, the fallback is to
152 double.
153
154 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
155 2005 (difference between two addresses compared to an unsigned value).
156
157 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
158 custom one, because of the following reported problem in Windows:
159
160 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
161 under Win32.
162 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
163 therefore missing the function definition.
164 - The compiler thus generates a "C" signature for the test function.
165 - The linker fails to find the "C" function.
166 - PCRE fails to configure if asked to do so against libbz2.
167
168 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
169 messages were output:
170
171 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
172 rerunning libtoolize, to keep the correct libtool macros in-tree.
173 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
174
175 I have done both of these things.
176
177 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
178 most of the time, it *can* run out if it is given a pattern that contains a
179 runaway infinite recursion. I updated the discussion in the pcrestack man
180 page.
181
182 13. Now that we have gone to the x.xx style of version numbers, the minor
183 version may start with zero. Using 08 or 09 is a bad idea because users
184 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
185 interpreted as invalid octal numbers. I've updated the previous comment in
186 configure.ac, and also added a check that gives an error if 08 or 09 are
187 used.
188
189 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
190 causing partial matching to fail when the end of the subject matched \W
191 in a UTF-8 pattern where \W was quantified with a minimum of 3.
192
193 15. There were some discrepancies between the declarations in pcre_internal.h
194 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
195 their definitions. The declarations used "const uschar *" and the
196 definitions used USPTR. Even though USPTR is normally defined as "const
197 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
198 reported that: "This difference in casting confuses some C++ compilers, for
199 example, SunCC recognizes above declarations as different functions and
200 generates broken code for hbpcre." I have changed the declarations to use
201 USPTR.
202
203 16. GNU libtool is named differently on some systems. The autogen.sh script now
204 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
205 (FreeBSD).
206
207 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
208 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
209 comment: "Figure out how to create a longlong from a string: strtoll and
210 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
211 instance, but it only takes 2 args instead of 3!"
212
213 18. A subtle bug concerned with back references has been fixed by a change of
214 specification, with a corresponding code fix. A pattern such as
215 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
216 refers, was giving matches when it shouldn't. For example, xa=xaaa would
217 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
218 same bug. Such groups have to be quantified to be useful, or contained
219 inside another quantified group. (If there's no repetition, the reference
220 can never match.) The problem arises because, having left the group and
221 moved on to the rest of the pattern, a later failure that backtracks into
222 the group uses the captured value from the final iteration of the group
223 rather than the correct earlier one. I have fixed this in PCRE by forcing
224 any group that contains a reference to itself to be an atomic group; that
225 is, there cannot be any backtracking into it once it has completed. This is
226 similar to recursive and subroutine calls.
227
228
229 Version 8.00 19-Oct-09
230 ----------------------
231
232 1. The table for translating pcre_compile() error codes into POSIX error codes
233 was out-of-date, and there was no check on the pcre_compile() error code
234 being within the table. This could lead to an OK return being given in
235 error.
236
237 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
238 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
239 in a Windows environment.
240
241 3. The pcregrep --count option prints the count for each file even when it is
242 zero, as does GNU grep. However, pcregrep was also printing all files when
243 --files-with-matches was added. Now, when both options are given, it prints
244 counts only for those files that have at least one match. (GNU grep just
245 prints the file name in this circumstance, but including the count seems
246 more useful - otherwise, why use --count?) Also ensured that the
247 combination -clh just lists non-zero counts, with no names.
248
249 4. The long form of the pcregrep -F option was incorrectly implemented as
250 --fixed_strings instead of --fixed-strings. This is an incompatible change,
251 but it seems right to fix it, and I didn't think it was worth preserving
252 the old behaviour.
253
254 5. The command line items --regex=pattern and --regexp=pattern were not
255 recognized by pcregrep, which required --regex pattern or --regexp pattern
256 (with a space rather than an '='). The man page documented the '=' forms,
257 which are compatible with GNU grep; these now work.
258
259 6. No libpcreposix.pc file was created for pkg-config; there was just
260 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
261
262 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
263 when UCP support is not needed, by modifying the Python script that
264 generates it from Unicode data files. This should not matter if the module
265 is correctly used as a library, but I received one complaint about 50K of
266 unwanted data. My guess is that the person linked everything into his
267 program rather than using a library. Anyway, it does no harm.
268
269 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
270 was a minimum greater than 1 for a wide character in a possessive
271 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
272 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
273 character. Chaos in the form of incorrect output or a compiling loop could
274 result.
275
276 9. The restrictions on what a pattern can contain when partial matching is
277 requested for pcre_exec() have been removed. All patterns can now be
278 partially matched by this function. In addition, if there are at least two
279 slots in the offset vector, the offset of the earliest inspected character
280 for the match and the offset of the end of the subject are set in them when
281 PCRE_ERROR_PARTIAL is returned.
282
283 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
284 synonymous with PCRE_PARTIAL, for backwards compatibility, and
285 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
286 and may be more useful for multi-segment matching.
287
288 11. Partial matching with pcre_exec() is now more intuitive. A partial match
289 used to be given if ever the end of the subject was reached; now it is
290 given only if matching could not proceed because another character was
291 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
292 string "Z", which now yields "no match" instead of "partial match". In the
293 case of pcre_dfa_exec(), "no match" is given if every matching path for the
294 final character ended with (*FAIL).
295
296 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
297 if the pattern had a "must contain" character that was already found in the
298 earlier partial match, unless partial matching was again requested. For
299 example, with the pattern /dog.(body)?/, the "must contain" character is
300 "g". If the first part-match was for the string "dog", restarting with
301 "sbody" failed. This bug has been fixed.
302
303 13. The string returned by pcre_dfa_exec() after a partial match has been
304 changed so that it starts at the first inspected character rather than the
305 first character of the match. This makes a difference only if the pattern
306 starts with a lookbehind assertion or \b or \B (\K is not supported by
307 pcre_dfa_exec()). It's an incompatible change, but it makes the two
308 matching functions compatible, and I think it's the right thing to do.
309
310 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
311 so that the demonstration program is easily available in environments where
312 PCRE has not been installed from source.
313
314 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
315 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
316 library.
317
318 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
319 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
320 is not the first non-POSIX option to be added. Clearly some people find
321 these options useful.
322
323 17. If a caller to the POSIX matching function regexec() passes a non-zero
324 value for nmatch with a NULL value for pmatch, the value of
325 nmatch is forced to zero.
326
327 18. RunGrepTest did not have a test for the availability of the -u option of
328 the diff command, as RunTest does. It now checks in the same way as
329 RunTest, and also checks for the -b option.
330
331 19. If an odd number of negated classes containing just a single character
332 interposed, within parentheses, between a forward reference to a named
333 subpattern and the definition of the subpattern, compilation crashed with
334 an internal error, complaining that it could not find the referenced
335 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
336 [The bug was that it was starting one character too far in when skipping
337 over the character class, thus treating the ] as data rather than
338 terminating the class. This meant it could skip too much.]
339
340 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
341 /g option in pcretest when the pattern contains \K, which makes it possible
342 to have an empty string match not at the start, even when the pattern is
343 anchored. Updated pcretest and pcredemo to use this option.
344
345 21. If the maximum number of capturing subpatterns in a recursion was greater
346 than the maximum at the outer level, the higher number was returned, but
347 with unset values at the outer level. The correct (outer level) value is
348 now given.
349
350 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
351 PCRE did not set those parentheses (unlike Perl). I have now found a way to
352 make it do so. The string so far is captured, making this feature
353 compatible with Perl.
354
355 23. The tests have been re-organized, adding tests 11 and 12, to make it
356 possible to check the Perl 5.10 features against Perl 5.10.
357
358 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
359 pattern matches a fixed length string. PCRE did not allow this; now it
360 does. Neither allows recursion.
361
362 25. I finally figured out how to implement a request to provide the minimum
363 length of subject string that was needed in order to match a given pattern.
364 (It was back references and recursion that I had previously got hung up
365 on.) This code has now been added to pcre_study(); it finds a lower bound
366 to the length of subject needed. It is not necessarily the greatest lower
367 bound, but using it to avoid searching strings that are too short does give
368 some useful speed-ups. The value is available to calling programs via
369 pcre_fullinfo().
370
371 26. While implementing 25, I discovered to my embarrassment that pcretest had
372 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
373 study optimizations had never been tested with that matching function.
374 Oops. What is worse, even when it was passed study data, there was a bug in
375 pcre_dfa_exec() that meant it never actually used it. Double oops. There
376 were also very few tests of studied patterns with pcre_dfa_exec().
377
378 27. If (?| is used to create subpatterns with duplicate numbers, they are now
379 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
380 on the other side of the coin, they are no longer allowed to have different
381 names, because these cannot be distinguished in PCRE, and this has caused
382 confusion. (This is a difference from Perl.)
383
384 28. When duplicate subpattern names are present (necessarily with different
385 numbers, as required by 27 above), and a test is made by name in a
386 conditional pattern, either for a subpattern having been matched, or for
387 recursion in such a pattern, all the associated numbered subpatterns are
388 tested, and the overall condition is true if the condition is true for any
389 one of them. This is the way Perl works, and is also more like the way
390 testing by number works.
391
392
393 Version 7.9 11-Apr-09
394 ---------------------
395
396 1. When building with support for bzlib/zlib (pcregrep) and/or readline
397 (pcretest), all targets were linked against these libraries. This included
398 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
399 libraries. This caused unwanted dependencies to be created. This problem
400 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
401 pcretest is linked with readline.
402
403 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
404 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
405 moved outside it again, because FALSE and TRUE are already defined in AIX,
406 but BOOL is not.
407
408 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
409 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
410
411 4. The pcregrep documentation said spaces were inserted as well as colons (or
412 hyphens) following file names and line numbers when outputting matching
413 lines. This is not true; no spaces are inserted. I have also clarified the
414 wording for the --colour (or --color) option.
415
416 5. In pcregrep, when --colour was used with -o, the list of matching strings
417 was not coloured; this is different to GNU grep, so I have changed it to be
418 the same.
419
420 6. When --colo(u)r was used in pcregrep, only the first matching substring in
421 each matching line was coloured. Now it goes on to look for further matches
422 of any of the test patterns, which is the same behaviour as GNU grep.
423
424 7. A pattern that could match an empty string could cause pcregrep to loop; it
425 doesn't make sense to accept an empty string match in pcregrep, so I have
426 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
427 seems to be how GNU grep behaves.
428
429 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
430 start or after a newline", because the conditional assertion was not being
431 correctly handled. The rule now is that both the assertion and what follows
432 in the first alternative must satisfy the test.
433
434 9. If auto-callout was enabled in a pattern with a conditional group whose
435 condition was an assertion, PCRE could crash during matching, both with
436 pcre_exec() and pcre_dfa_exec().
437
438 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
439 used for matching.
440
441 11. Unicode property support in character classes was not working for
442 characters (bytes) greater than 127 when not in UTF-8 mode.
443
444 12. Added the -M command line option to pcretest.
445
446 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
447
448 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
449
450 16. Added comments and documentation about mis-use of no_arg in the C++
451 wrapper.
452
453 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
454 from Martin Jerabek that uses macro names for all relevant character and
455 string constants.
456
457 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
458 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
459 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
460 these, but not everybody uses configure.
461
462 19. A conditional group that had only one branch was not being correctly
463 recognized as an item that could match an empty string. This meant that an
464 enclosing group might also not be so recognized, causing infinite looping
465 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
466 with the subject "ab", where knowledge that the repeated group can match
467 nothing is needed in order to break the loop.
468
469 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
470 exec(), but without supplying a callout function, matching went wrong.
471
472 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
473 leak if the size of the offset vector was greater than 30. When the vector
474 is smaller, the saved offsets during recursion go onto a local stack
475 vector, but for larger vectors malloc() is used. It was failing to free
476 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
477 error, in fact).
478
479 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
480 heapframe that is used only when UTF-8 support is enabled. This caused no
481 problem, but was untidy.
482
483 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
484 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
485 included within another project.
486
487 24. Steven Van Ingelgem's patches to add more options to the CMake support,
488 slightly modified by me:
489
490 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
491 not building pcregrep.
492
493 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
494 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
495
496 25. Forward references, both numeric and by name, in patterns that made use of
497 duplicate group numbers, could behave incorrectly or give incorrect errors,
498 because when scanning forward to find the reference group, PCRE was not
499 taking into account the duplicate group numbers. A pattern such as
500 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
501
502 26. Changed a few more instances of "const unsigned char *" to USPTR, making
503 the feature of a custom pointer more persuasive (as requested by a user).
504
505 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
506 pcretest.c, inside #ifndefs, because it seems they are sometimes already
507 pre-defined.
508
509 28. Added support for (*UTF8) at the start of a pattern.
510
511 29. Arrange for flags added by the "release type" setting in CMake to be shown
512 in the configuration summary.
513
514
515 Version 7.8 05-Sep-08
516 ---------------------
517
518 1. Replaced UCP searching code with optimized version as implemented for Ad
519 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
520 stage table and inline lookup instead of a function, giving speed ups of 2
521 to 5 times on some simple patterns that I tested. Permission was given to
522 distribute the MultiStage2.py script that generates the tables (it's not in
523 the tarball, but is in the Subversion repository).
524
525 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
526 scripts.
527
528 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
529 a group with a zero qualifier. The result of the study could be incorrect,
530 or the function might crash, depending on the pattern.
531
532 4. Caseless matching was not working for non-ASCII characters in back
533 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
534 It now works when Unicode Property Support is available.
535
536 5. In pcretest, an escape such as \x{de} in the data was always generating
537 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
538 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
539 truncation.
540
541 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
542
543 7. Added two (int) casts to pcregrep when printing the difference of two
544 pointers, in case they are 64-bit values.
545
546 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
547 test 2 if it fails.
548
549 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
550 and a #define of that name to empty if it is not externally set. This is to
551 allow users of MSVC to set it if necessary.
552
553 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
554 the convenience functions in the pcre_get.c source file.
555
556 11. An option change at the start of a pattern that had top-level alternatives
557 could cause overwriting and/or a crash. This command provoked a crash in
558 some environments:
559
560 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
561
562 This potential security problem was recorded as CVE-2008-2371.
563
564 12. For a pattern where the match had to start at the beginning or immediately
565 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
566 pcre_dfa_exec() could read past the end of the passed subject if there was
567 no match. To help with detecting such bugs (e.g. with valgrind), I modified
568 pcretest so that it places the subject at the end of its malloc-ed buffer.
569
570 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
571 exec() might read past the end of the data buffer in UTF-8 mode.
572
573 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
574 the data contained the byte 0x85 as part of a UTF-8 character within its
575 first line. This applied both to normal and DFA matching.
576
577 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
578 /^[^d]*?$/8 failed to match "abc".
579
580 16. Added a missing copyright notice to pcrecpp_internal.h.
581
582 17. Make it more clear in the documentation that values returned from
583 pcre_exec() in ovector are byte offsets, not character counts.
584
585 18. Tidied a few places to stop certain compilers from issuing warnings.
586
587 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
588 supplied by Stefan Weber. I made a further small update for 7.8 because
589 there is a change of source arrangements: the pcre_searchfuncs.c module is
590 replaced by pcre_ucd.c.
591
592
593 Version 7.7 07-May-08
594 ---------------------
595
596 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
597 a string to a long long, pretend we don't even have a long long." This is
598 done by checking for the strtoq, strtoll, and _strtoi64 functions.
599
600 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
601 pre-7.6 versions, which defined a global no_arg variable instead of putting
602 it in the RE class. (See also #8 below.)
603
604 3. Remove a line of dead code, identified by coverity and reported by Nuno
605 Lopes.
606
607 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
608
609 (1) The include/exclude patterns were being applied to the whole pathnames
610 of files, instead of just to the final components.
611
612 (2) If there was more than one level of directory, the subdirectories were
613 skipped unless they satisfied the include/exclude conditions. This is
614 inconsistent with GNU grep (and could even be seen as contrary to the
615 pcregrep specification - which I improved to make it absolutely clear).
616 The action now is always to scan all levels of directory, and just
617 apply the include/exclude patterns to regular files.
618
619 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
620 --exclude_dir in the tests to avoid scanning .svn directories.
621
622 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
623 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
624 doesn't support NULs in patterns.
625
626 7. Added some missing "const"s to declarations of static tables in
627 pcre_compile.c and pcre_dfa_exec.c.
628
629 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
630 caused by fix #2 above. (Subsequently also a second patch to fix the
631 first patch. And a third patch - this was a messy problem.)
632
633 9. Applied Craig's patch to remove the use of push_back().
634
635 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
636 matching function regexec().
637
638 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
639 which, however, unlike Perl's \g{...}, are subroutine calls, not back
640 references. PCRE supports relative numbers with this syntax (I don't think
641 Oniguruma does).
642
643 12. Previously, a group with a zero repeat such as (...){0} was completely
644 omitted from the compiled regex. However, this means that if the group
645 was called as a subroutine from elsewhere in the pattern, things went wrong
646 (an internal error was given). Such groups are now left in the compiled
647 pattern, with a new opcode that causes them to be skipped at execution
648 time.
649
650 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
651 to the way PCRE behaves:
652
653 (a) A lone ] character is dis-allowed (Perl treats it as data).
654
655 (b) A back reference to an unmatched subpattern matches an empty string
656 (Perl fails the current match path).
657
658 (c) A data ] in a character class must be notated as \] because if the
659 first data character in a class is ], it defines an empty class. (In
660 Perl it is not possible to have an empty class.) The empty class []
661 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
662 The negative empty class [^] matches any one character, independently
663 of the DOTALL setting.
664
665 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
666 non-existent subpattern following a character class starting with ']' and
667 containing () gave an internal compiling error instead of "reference to
668 non-existent subpattern". Fortunately, when the pattern did exist, the
669 compiled code was correct. (When scanning forwards to check for the
670 existencd of the subpattern, it was treating the data ']' as terminating
671 the class, so got the count wrong. When actually compiling, the reference
672 was subsequently set up correctly.)
673
674 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
675 it was being rejected as not supported by pcre_dfa_exec(), even though
676 other assertions are supported. I have made pcre_dfa_exec() support
677 (*FAIL).
678
679 16. The implementation of 13c above involved the invention of a new opcode,
680 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
681 cannot be changed at match time, I realized I could make a small
682 improvement to matching performance by compiling OP_ALLANY instead of
683 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
684 on the OP_ANY path.
685
686 17. Compiling pcretest on Windows with readline support failed without the
687 following two fixes: (1) Make the unistd.h include conditional on
688 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
689
690 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
691 ncurses library to be included for pcretest when ReadLine support is
692 requested, but also to allow for it to be overridden. This patch came from
693 Daniel Bergström.
694
695 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
696 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
697 any errors with the current Unicode tables. Thanks to Peter Kankowski for
698 spotting this.
699
700
701 Version 7.6 28-Jan-08
702 ---------------------
703
704 1. A character class containing a very large number of characters with
705 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
706 overflow.
707
708 2. Patch to cut out the "long long" test in pcrecpp_unittest when
709 HAVE_LONG_LONG is not defined.
710
711 3. Applied Christian Ehrlicher's patch to update the CMake build files to
712 bring them up to date and include new features. This patch includes:
713
714 - Fixed PH's badly added libz and libbz2 support.
715 - Fixed a problem with static linking.
716 - Added pcredemo. [But later removed - see 7 below.]
717 - Fixed dftables problem and added an option.
718 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
719 HAVE_LONG_LONG.
720 - Added readline support for pcretest.
721 - Added an listing of the option settings after cmake has run.
722
723 4. A user submitted a patch to Makefile that makes it easy to create
724 "pcre.dll" under mingw when using Configure/Make. I added stuff to
725 Makefile.am that cause it to include this special target, without
726 affecting anything else. Note that the same mingw target plus all
727 the other distribution libraries and programs are now supported
728 when configuring with CMake (see 6 below) instead of with
729 Configure/Make.
730
731 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
732 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
733 exported in the Windows port". It has not yet been confirmed that the patch
734 solves the problem, but it does no harm.
735
736 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
737 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
738 with CMake, and also correct the comment about stack recursion.
739
740 7. Remove the automatic building of pcredemo from the ./configure system and
741 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
742 of a program that users should build themselves after PCRE is installed, so
743 building it automatically is not really right. What is more, it gave
744 trouble in some build environments.
745
746 8. Further tidies to CMakeLists.txt from Sheri and Christian.
747
748
749 Version 7.5 10-Jan-08
750 ---------------------
751
752 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
753 values in parens when parsing an RE using the C++ wrapper."
754
755 2. Negative specials like \S did not work in character classes in UTF-8 mode.
756 Characters greater than 255 were excluded from the class instead of being
757 included.
758
759 3. The same bug as (2) above applied to negated POSIX classes such as
760 [:^space:].
761
762 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
763 defined or documented. It seems to have been a typo for PCRE_STATIC, so
764 I have changed it.
765
766 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
767 first named subpattern) and a construct such as (?&a) would reference the
768 first named subpattern whose name started with "a" (in other words, the
769 length check was missing). Both these problems are fixed. "Subpattern name
770 expected" is now given for (?&) (a zero-length name), and this patch also
771 makes it give the same error for \k'' (previously it complained that that
772 was a reference to a non-existent subpattern).
773
774 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
775 this is right because (?- can be followed by option settings as well as by
776 digits. I have, however, made the messages clearer.
777
778 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
779 than the number used in the conditional) now cause a compile-time error.
780 This is actually not compatible with Perl, which accepts such patterns, but
781 treats the conditional as always being FALSE (as PCRE used to), but it
782 seems to me that giving a diagnostic is better.
783
784 8. Change "alphameric" to the more common word "alphanumeric" in comments
785 and messages.
786
787 9. Fix two occurrences of "backslash" in comments that should have been
788 "backspace".
789
790 10. Remove two redundant lines of code that can never be obeyed (their function
791 was moved elsewhere).
792
793 11. The program that makes PCRE's Unicode character property table had a bug
794 which caused it to generate incorrect table entries for sequences of
795 characters that have the same character type, but are in different scripts.
796 It amalgamated them into a single range, with the script of the first of
797 them. In other words, some characters were in the wrong script. There were
798 thirteen such cases, affecting characters in the following ranges:
799
800 U+002b0 - U+002c1
801 U+0060c - U+0060d
802 U+0061e - U+00612
803 U+0064b - U+0065e
804 U+0074d - U+0076d
805 U+01800 - U+01805
806 U+01d00 - U+01d77
807 U+01d9b - U+01dbf
808 U+0200b - U+0200f
809 U+030fc - U+030fe
810 U+03260 - U+0327f
811 U+0fb46 - U+0fbb1
812 U+10450 - U+1049d
813
814 12. The -o option (show only the matching part of a line) for pcregrep was not
815 compatible with GNU grep in that, if there was more than one match in a
816 line, it showed only the first of them. It now behaves in the same way as
817 GNU grep.
818
819 13. If the -o and -v options were combined for pcregrep, it printed a blank
820 line for every non-matching line. GNU grep prints nothing, and pcregrep now
821 does the same. The return code can be used to tell if there were any
822 non-matching lines.
823
824 14. Added --file-offsets and --line-offsets to pcregrep.
825
826 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
827 infinitely looping recursion. The bug was that positive lookaheads were not
828 being skipped when checking for a possible empty match (negative lookaheads
829 and both kinds of lookbehind were skipped).
830
831 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
832 inclusion of <windows.h> to before rather than after the definition of
833 INVALID_FILE_ATTRIBUTES (patch from David Byron).
834
835 17. Specifying a possessive quantifier with a specific limit for a Unicode
836 character property caused pcre_compile() to compile bad code, which led at
837 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
838 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
839 caused the error; without that there was no problem.
840
841 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
842
843 19. Added --enable-pcretest-libreadline.
844
845 20. In pcrecpp.cc, the variable 'count' was incremented twice in
846 RE::GlobalReplace(). As a result, the number of replacements returned was
847 double what it should be. I removed one of the increments, but Craig sent a
848 later patch that removed the other one (the right fix) and added unit tests
849 that check the return values (which was not done before).
850
851 21. Several CMake things:
852
853 (1) Arranged that, when cmake is used on Unix, the libraries end up with
854 the names libpcre and libpcreposix, not just pcre and pcreposix.
855
856 (2) The above change means that pcretest and pcregrep are now correctly
857 linked with the newly-built libraries, not previously installed ones.
858
859 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
860
861 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
862 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
863 UTF-8 newline character). The key issue is that the pattern starts .*;
864 this means that the match must be either at the beginning, or after a
865 newline. The bug was in the code for advancing after a failed match and
866 checking that the new position followed a newline. It was not taking
867 account of UTF-8 characters correctly.
868
869 23. PCRE was behaving differently from Perl in the way it recognized POSIX
870 character classes. PCRE was not treating the sequence [:...:] as a
871 character class unless the ... were all letters. Perl, however, seems to
872 allow any characters between [: and :], though of course it rejects as
873 unknown any "names" that contain non-letters, because all the known class
874 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
875 for example, whereas PCRE did not - it did not recognize a POSIX character
876 class. This seemed a bit dangerous, so the code has been changed to be
877 closer to Perl. The behaviour is not identical to Perl, because PCRE will
878 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
879 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
880 Perl does, and where it didn't before.
881
882 24. Rewrite so as to remove the single use of %n from pcregrep because in some
883 Windows environments %n is disabled by default.
884
885
886 Version 7.4 21-Sep-07
887 ---------------------
888
889 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
890 means that a class such as [\s] counted as "explicit reference to CR or
891 LF". That isn't really right - the whole point of the change was to try to
892 help when there was an actual mention of one of the two characters. So now
893 the change happens only if \r or \n (or a literal CR or LF) character is
894 encountered.
895
896 2. The 32-bit options word was also used for 6 internal flags, but the numbers
897 of both had grown to the point where there were only 3 bits left.
898 Fortunately, there was spare space in the data structure, and so I have
899 moved the internal flags into a new 16-bit field to free up more option
900 bits.
901
902 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
903 but did not set the internal JCHANGED flag - either of these is enough to
904 control the way the "get" function works - but the PCRE_INFO_JCHANGED
905 facility is supposed to tell if (?J) was ever used, so now (?J) at the
906 start sets both bits.
907
908 4. Added options (at build time, compile time, exec time) to change \R from
909 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
910
911 5. doc/pcresyntax.html was missing from the distribution.
912
913 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
914 compatibility, even though it is no longer used.
915
916 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
917 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
918 windows.h file is present (where different names are used). [This was
919 reversed later after testing - see 16 below.]
920
921 8. Changed all #include <config.h> to #include "config.h". There were also
922 some further <pcre.h> cases that I changed to "pcre.h".
923
924 9. When pcregrep was used with the --colour option, it missed the line ending
925 sequence off the lines that it output.
926
927 10. It was pointed out to me that arrays of string pointers cause lots of
928 relocations when a shared library is dynamically loaded. A technique of
929 using a single long string with a table of offsets can drastically reduce
930 these. I have refactored PCRE in four places to do this. The result is
931 dramatic:
932
933 Originally: 290
934 After changing UCP table: 187
935 After changing error message table: 43
936 After changing table of "verbs" 36
937 After changing table of Posix names 22
938
939 Thanks to the folks working on Gregex for glib for this insight.
940
941 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
942 unicode-properties was also set.
943
944 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
945
946 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
947 checked only for CRLF.
948
949 14. Added casts to pcretest.c to avoid compiler warnings.
950
951 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
952
953 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
954 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
955 entirely. This removes changes made in 7 above.
956
957 17. The CMake files have been updated, and there is now more information about
958 building with CMake in the NON-UNIX-USE document.
959
960
961 Version 7.3 28-Aug-07
962 ---------------------
963
964 1. In the rejigging of the build system that eventually resulted in 7.1, the
965 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
966 brackets there is not right, since it causes compilers to look for an
967 installed pcre.h, not the version that is in the source that is being
968 compiled (which of course may be different). I have changed it back to:
969
970 #include "pcre.h"
971
972 I have a vague recollection that the change was concerned with compiling in
973 different directories, but in the new build system, that is taken care of
974 by the VPATH setting the Makefile.
975
976 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
977 when the subject happened to end in the byte 0x85 (e.g. if the last
978 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
979 characters but of course it shouldn't be taken as a newline when it is part
980 of another character. The bug was that, for an unlimited repeat of . in
981 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
982 characters when looking for a newline.
983
984 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
985
986 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
987 in debug output.
988
989 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
990 long printing in the pcrecpp unittest when running under MinGW.
991
992 6. ESC_K was left out of the EBCDIC table.
993
994 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
995 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
996 limit also applies to "virtual nesting" when a pattern is recursive, and in
997 this case 1000 isn't so big. I have been able to remove this limit at the
998 expense of backing off one optimization in certain circumstances. Normally,
999 when pcre_exec() would call its internal match() function recursively and
1000 immediately return the result unconditionally, it uses a "tail recursion"
1001 feature to save stack. However, when a subpattern that can match an empty
1002 string has an unlimited repetition quantifier, it no longer makes this
1003 optimization. That gives it a stack frame in which to save the data for
1004 checking that an empty string has been matched. Previously this was taken
1005 from the 1000-entry workspace that had been reserved. So now there is no
1006 explicit limit, but more stack is used.
1007
1008 8. Applied Daniel's patches to solve problems with the import/export magic
1009 syntax that is required for Windows, and which was going wrong for the
1010 pcreposix and pcrecpp parts of the library. These were overlooked when this
1011 problem was solved for the main library.
1012
1013 9. There were some crude static tests to avoid integer overflow when computing
1014 the size of patterns that contain repeated groups with explicit upper
1015 limits. As the maximum quantifier is 65535, the maximum group length was
1016 set at 30,000 so that the product of these two numbers did not overflow a
1017 32-bit integer. However, it turns out that people want to use groups that
1018 are longer than 30,000 bytes (though not repeat them that many times).
1019 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1020 made it possible to implement the integer overflow checks in a much more
1021 dynamic way, which I have now done. The artificial limitation on group
1022 length has been removed - we now have only the limit on the total length of
1023 the compiled pattern, which depends on the LINK_SIZE setting.
1024
1025 10. Fixed a bug in the documentation for get/copy named substring when
1026 duplicate names are permitted. If none of the named substrings are set, the
1027 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1028 empty string.
1029
1030 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1031 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1032 because the ] is interpreted as the first data character and the
1033 terminating ] is not found. PCRE has been made compatible with Perl in this
1034 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1035 cause memory overwriting.
1036
1037 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1038 string has been matched (to stop an infinite loop). It was not recognizing
1039 a conditional subpattern that could match an empty string if that
1040 subpattern was within another subpattern. For example, it looped when
1041 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1042 condition was not nested. This bug has been fixed.
1043
1044 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1045 past the start of the subject in the presence of bytes with the top bit
1046 set, for example "\x8aBCD".
1047
1048 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1049 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1050
1051 14. Optimized (?!) to (*FAIL).
1052
1053 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1054 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1055 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1056 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1057 does: it's just the validity check that is more restrictive.
1058
1059 16. Inserted checks for integer overflows during escape sequence (backslash)
1060 processing, and also fixed erroneous offset values for syntax errors during
1061 backslash processing.
1062
1063 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1064 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1065
1066 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1067 caused an overrun.
1068
1069 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1070 something other than just ASCII characters) inside a group that had an
1071 unlimited repeat caused a loop at compile time (while checking to see
1072 whether the group could match an empty string).
1073
1074 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1075 [\P{Any}] did so. (Error in the code for printing property names.)
1076
1077 21. An orphan \E inside a character class could cause a crash.
1078
1079 22. A repeated capturing bracket such as (A)? could cause a wild memory
1080 reference during compilation.
1081
1082 23. There are several functions in pcre_compile() that scan along a compiled
1083 expression for various reasons (e.g. to see if it's fixed length for look
1084 behind). There were bugs in these functions when a repeated \p or \P was
1085 present in the pattern. These operators have additional parameters compared
1086 with \d, etc, and these were not being taken into account when moving along
1087 the compiled data. Specifically:
1088
1089 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1090 length.
1091
1092 (b) An item such as \pL+ within a repeated group could cause crashes or
1093 loops.
1094
1095 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1096 "reference to non-existent subpattern" error.
1097
1098 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1099
1100 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1101 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1102
1103 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1104
1105 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1106 character were causing crashes (broken optimization).
1107
1108 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1109 \p or \P) caused a compile-time loop.
1110
1111 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1112 break. For example, the unstudied pattern [\r\n]A does not match the string
1113 "\r\nA" because change 7.0/46 below moves the current point on by two
1114 characters after failing to match at the start. However, the pattern \nA
1115 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1116 the same is true. There doesn't seem any very clean way out of this, but
1117 what I have chosen to do makes the common cases work: PCRE now takes note
1118 of whether there can be an explicit match for \r or \n anywhere in the
1119 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1120 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1121 pattern has explicit CR or LF references.
1122
1123 29. Added (*CR) etc for changing newline setting at start of pattern.
1124
1125
1126 Version 7.2 19-Jun-07
1127 ---------------------
1128
1129 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1130 which is apparently normally available under Windows.
1131
1132 2. Re-jig the pcregrep tests with different newline settings in an attempt
1133 to make them independent of the local environment's newline setting.
1134
1135 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1136
1137 4. Some of the "internals" tests were previously cut out when the link size
1138 was not 2, because the output contained actual offsets. The recent new
1139 "Z" feature of pcretest means that these can be cut out, making the tests
1140 usable with all link sizes.
1141
1142 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1143 stack recursion. This gives a massive performance boost under BSD, but just
1144 a small improvement under Linux. However, it saves one field in the frame
1145 in all cases.
1146
1147 6. Added more features from the forthcoming Perl 5.10:
1148
1149 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1150 recursion call. It refers to the nth most recently opened parentheses.
1151
1152 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1153 to be opened parentheses.
1154
1155 (c) Conditions that refer to capturing parentheses can be specified
1156 relatively, for example, (?(-2)... or (?(+3)...
1157
1158 (d) \K resets the start of the current match so that everything before
1159 is not part of it.
1160
1161 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1162
1163 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1164 reference syntax.
1165
1166 (g) (?| introduces a group in which the numbering of parentheses in each
1167 alternative starts with the same number.
1168
1169 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1170
1171 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1172 PCRE_INFO_JCHANGED.
1173
1174 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1175 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1176 for detecting groups that can match an empty string.
1177
1178 9. A pattern with a very large number of alternatives (more than several
1179 hundred) was running out of internal workspace during the pre-compile
1180 phase, where pcre_compile() figures out how much memory will be needed. A
1181 bit of new cunning has reduced the workspace needed for groups with
1182 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1183 workspace instead of running out of the 4096 that are available.
1184
1185 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1186
1187 11. Applied patch from Google to remove an optimization that didn't quite work.
1188 The report of the bug said:
1189
1190 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1191 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1192 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1193
1194 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1195 it matched the wrong number of bytes.
1196
1197
1198 Version 7.1 24-Apr-07
1199 ---------------------
1200
1201 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1202 that is more "standard", making use of automake and other Autotools. There
1203 is some re-arrangement of the files and adjustment of comments consequent
1204 on this.
1205
1206 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1207 for recursive directory scanning broke on some systems because the files
1208 are not scanned in any specific order and on different systems the order
1209 was different. A call to "sort" has been inserted into RunGrepTest for the
1210 approprate test as a short-term fix. In the longer term there may be an
1211 alternative.
1212
1213 3. I had an email from Eric Raymond about problems translating some of PCRE's
1214 man pages to HTML (despite the fact that I distribute HTML pages, some
1215 people do their own conversions for various reasons). The problems
1216 concerned the use of low-level troff macros .br and .in. I have therefore
1217 removed all such uses from the man pages (some were redundant, some could
1218 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1219 HTML has been updated to handle .nf/.fi and to complain if it encounters
1220 .br or .in.
1221
1222 4. Updated comments in configure.ac that get placed in config.h.in and also
1223 arranged for config.h to be included in the distribution, with the name
1224 config.h.generic, for the benefit of those who have to compile without
1225 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1226
1227 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1228 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1229 makevp.bat for the current PCRE, using the additional files
1230 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1231
1232 6. A Windows user reported a minor discrepancy with test 2, which turned out
1233 to be caused by a trailing space on an input line that had got lost in his
1234 copy. The trailing space was an accident, so I've just removed it.
1235
1236 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1237 that is needed.
1238
1239 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1240 as "const" (a) because they are and (b) because it helps the PHP
1241 maintainers who have recently made a script to detect big data structures
1242 in the php code that should be moved to the .rodata section. I remembered
1243 to update Builducptable as well, so it won't revert if ucptable.h is ever
1244 re-created.
1245
1246 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1247 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1248 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1249 support is not required. This saves 1.5-2K of code, which is important in
1250 some applications.
1251
1252 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1253 so as not to refer to the tables, even though these functions will never be
1254 called when UTF-8 support is disabled. Otherwise there are problems with a
1255 shared library.
1256
1257 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1258
1259 (a) It was defining its arguments as char * instead of void *.
1260
1261 (b) It was assuming that all moves were upwards in memory; this was true
1262 a long time ago when I wrote it, but is no longer the case.
1263
1264 The emulated memove() is provided for those environments that have neither
1265 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1266 is clearly not the case, as these two bugs were recently reported.
1267
1268 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1269 and Detrail to create the HTML documentation, the .txt form of the man
1270 pages, and it removes trailing spaces from listed files. It also creates
1271 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1272 case, it wraps all the #defines with #ifndefs. This script should be run
1273 before "make dist".
1274
1275 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1276 with Unicode property support.
1277
1278 (a) For a maximizing quantifier, if the two different cases of the
1279 character were of different lengths in their UTF-8 codings (there are
1280 some cases like this - I found 11), and the matching function had to
1281 back up over a mixture of the two cases, it incorrectly assumed they
1282 were both the same length.
1283
1284 (b) When PCRE was configured to use the heap rather than the stack for
1285 recursion during matching, it was not correctly preserving the data for
1286 the other case of a UTF-8 character when checking ahead for a match
1287 while processing a minimizing repeat. If the check also involved
1288 matching a wide character, but failed, corruption could cause an
1289 erroneous result when trying to check for a repeat of the original
1290 character.
1291
1292 13. Some tidying changes to the testing mechanism:
1293
1294 (a) The RunTest script now detects the internal link size and whether there
1295 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1296 values substituted by "configure". (The RunGrepTest script already did
1297 this for UTF-8.) The configure.ac script no longer substitutes the
1298 relevant variables.
1299
1300 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1301 with length and offset values. This means that the output is different
1302 for different internal link sizes. Test 2 is skipped for link sizes
1303 other than 2 because of this, bypassing the problem. Unfortunately,
1304 there was also a test in test 3 (the locale tests) that used /B and
1305 failed for link sizes other than 2. Rather than cut the whole test out,
1306 I have added a new /Z option to pcretest that replaces the length and
1307 offset values with spaces. This is now used to make test 3 independent
1308 of link size. (Test 2 will be tidied up later.)
1309
1310 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1311 segmentation fault instead of returning the appropriate error message.
1312
1313 15. In multiline mode when the newline sequence was set to "any", the pattern
1314 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1315 This doesn't seem right; it now treats the CRLF combination as the line
1316 ending, and so does not match in that case. It's only a pattern such as ^$
1317 that would hit this one: something like ^ABC$ would have failed after \r
1318 and then tried again after \r\n.
1319
1320 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1321 in an attempt to make files that differ only in their line terminators
1322 compare equal. This works on Linux.
1323
1324 17. Under certain error circumstances pcregrep might try to free random memory
1325 as it exited. This is now fixed, thanks to valgrind.
1326
1327 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1328 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1329 was because its rules for how to advance for /g after matching an empty
1330 string at the end of a line did not allow for this case. They now check for
1331 it specially.
1332
1333 20. pcretest is supposed to handle patterns and data of any length, by
1334 extending its buffers when necessary. It was getting this wrong when the
1335 buffer for a data line had to be extended.
1336
1337 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1338 CRLF as a newline sequence.
1339
1340 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1341 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1342 I have nevertheless tidied it up.
1343
1344 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1345
1346 24. Added a man page for pcre-config.
1347
1348
1349 Version 7.0 19-Dec-06
1350 ---------------------
1351
1352 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1353 moving to gcc 4.1.1.
1354
1355 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1356 sys/time.h, which is documented as needed for this function. It doesn't
1357 seem to matter on Linux, but it showed up on some releases of OS X.
1358
1359 3. It seems that there are systems where bytes whose values are greater than
1360 127 match isprint() in the "C" locale. The "C" locale should be the
1361 default when a C program starts up. In most systems, only ASCII printing
1362 characters match isprint(). This difference caused the output from pcretest
1363 to vary, making some of the tests fail. I have changed pcretest so that:
1364
1365 (a) When it is outputting text in the compiled version of a pattern, bytes
1366 other than 32-126 are always shown as hex escapes.
1367
1368 (b) When it is outputting text that is a matched part of a subject string,
1369 it does the same, unless a different locale has been set for the match
1370 (using the /L modifier). In this case, it uses isprint() to decide.
1371
1372 4. Fixed a major bug that caused incorrect computation of the amount of memory
1373 required for a compiled pattern when options that changed within the
1374 pattern affected the logic of the preliminary scan that determines the
1375 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1376 that the computed length was too small. The symptoms of this bug were
1377 either the PCRE error "internal error: code overflow" from pcre_compile(),
1378 or a glibc crash with a message such as "pcretest: free(): invalid next
1379 size (fast)". Examples of patterns that provoked this bug (shown in
1380 pcretest format) are:
1381
1382 /(?-x: )/x
1383 /(?x)(?-x: \s*#\s*)/
1384 /((?i)[\x{c0}])/8
1385 /(?i:[\x{c0}])/8
1386
1387 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1388 is now done differently.
1389
1390 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1391 wrapper classes; (b) implement a new function in the C++ scanner that is
1392 more efficient than the old way of doing things because it avoids levels of
1393 recursion in the regex matching; (c) add a paragraph to the documentation
1394 for the FullMatch() function.
1395
1396 6. The escape sequence \n was being treated as whatever was defined as
1397 "newline". Not only was this contrary to the documentation, which states
1398 that \n is character 10 (hex 0A), but it also went horribly wrong when
1399 "newline" was defined as CRLF. This has been fixed.
1400
1401 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1402 was being set to -1 for the "end of line" case (supposedly a value that no
1403 character can have). Though this value is never used (the check for end of
1404 line is "zero bytes in current character"), it caused compiler complaints.
1405 I've changed it to 0xffffffff.
1406
1407 8. In pcre_version.c, the version string was being built by a sequence of
1408 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1409 string (as it is for production releases) called a macro with an empty
1410 argument. The C standard says the result of this is undefined. The gcc
1411 compiler treats it as an empty string (which was what was wanted) but it is
1412 reported that Visual C gives an error. The source has been hacked around to
1413 avoid this problem.
1414
1415 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1416 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1417 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1418 of them did).
1419
1420 10. Originally, pcretest opened its input and output without "b"; then I was
1421 told that "b" was needed in some environments, so it was added for release
1422 5.0 to both the input and output. (It makes no difference on Unix-like
1423 systems.) Later I was told that it is wrong for the input on Windows. I've
1424 now abstracted the modes into two macros, to make it easier to fiddle with
1425 them, and removed "b" from the input mode under Windows.
1426
1427 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1428
1429 12. Added -help and --help to pcretest as an official way of being reminded
1430 of the options.
1431
1432 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1433 and pcrecpp.cc because they annoy compilers at high warning levels.
1434
1435 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1436
1437 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1438 scripts are not C programs :-) and which was not noticed because it works
1439 on Linux.
1440
1441 16. pcretest is supposed to handle any length of pattern and data line (as one
1442 line or as a continued sequence of lines) by extending its input buffer if
1443 necessary. This feature was broken for very long pattern lines, leading to
1444 a string of junk being passed to pcre_compile() if the pattern was longer
1445 than about 50K.
1446
1447 17. I have done a major re-factoring of the way pcre_compile() computes the
1448 amount of memory needed for a compiled pattern. Previously, there was code
1449 that made a preliminary scan of the pattern in order to do this. That was
1450 OK when PCRE was new, but as the facilities have expanded, it has become
1451 harder and harder to keep it in step with the real compile phase, and there
1452 have been a number of bugs (see for example, 4 above). I have now found a
1453 cunning way of running the real compile function in a "fake" mode that
1454 enables it to compute how much memory it would need, while actually only
1455 ever using a few hundred bytes of working memory and without too many
1456 tests of the mode. This should make future maintenance and development
1457 easier. A side effect of this work is that the limit of 200 on the nesting
1458 depth of parentheses has been removed (though this was never a serious
1459 limitation, I suspect). However, there is a downside: pcre_compile() now
1460 runs more slowly than before (30% or more, depending on the pattern). I
1461 hope this isn't a big issue. There is no effect on runtime performance.
1462
1463 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1464 newline (only possible for the last line of a file) and it was a
1465 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1466
1467 19. Added additional timing features to pcretest. (1) The -tm option now times
1468 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1469 separate command line item, by a number that specifies the number of
1470 repeats to use when timing. The default is 50000; this gives better
1471 precision, but takes uncomfortably long for very large patterns.
1472
1473 20. Extended pcre_study() to be more clever in cases where a branch of a
1474 subpattern has no definite first character. For example, (a*|b*)[cd] would
1475 previously give no result from pcre_study(). Now it recognizes that the
1476 first character must be a, b, c, or d.
1477
1478 21. There was an incorrect error "recursive call could loop indefinitely" if
1479 a subpattern (or the entire pattern) that was being tested for matching an
1480 empty string contained only one non-empty item after a nested subpattern.
1481 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1482 incorrectly, because the \d was being skipped in the check.
1483
1484 22. The pcretest program now has a new pattern option /B and a command line
1485 option -b, which is equivalent to adding /B to every pattern. This causes
1486 it to show the compiled bytecode, without the additional information that
1487 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1488 is the same as /B/I).
1489
1490 23. A new optimization is now able automatically to treat some sequences such
1491 as a*b as a*+b. More specifically, if something simple (such as a character
1492 or a simple class like \d) has an unlimited quantifier, and is followed by
1493 something that cannot possibly match the quantified thing, the quantifier
1494 is automatically "possessified".
1495
1496 24. A recursive reference to a subpattern whose number was greater than 39
1497 went wrong under certain circumstances in UTF-8 mode. This bug could also
1498 have affected the operation of pcre_study().
1499
1500 25. Realized that a little bit of performance could be had by replacing
1501 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1502
1503 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1504
1505 27. Possessive quantifiers such as a++ were previously implemented by turning
1506 them into atomic groups such as ($>a+). Now they have their own opcodes,
1507 which improves performance. This includes the automatically created ones
1508 from 23 above.
1509
1510 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1511 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1512 the first matched character to be a colon. This applied both to named and
1513 numbered groups.
1514
1515 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1516
1517 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1518 building PCRE on Windows easier, so I have included it in the distribution.
1519
1520 31. There is now a check in pcretest against a ridiculously large number being
1521 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1522 loop, the loop is abandoned.
1523
1524 32. Forward references to subpatterns in conditions such as (?(2)...) where
1525 subpattern 2 is defined later cause pcre_compile() to search forwards in
1526 the pattern for the relevant set of parentheses. This search went wrong
1527 when there were unescaped parentheses in a character class, parentheses
1528 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1529
1530 33. "Subroutine" calls and backreferences were previously restricted to
1531 referencing subpatterns earlier in the regex. This restriction has now
1532 been removed.
1533
1534 34. Added a number of extra features that are going to be in Perl 5.10. On the
1535 whole, these are just syntactic alternatives for features that PCRE had
1536 previously implemented using the Python syntax or my own invention. The
1537 other formats are all retained for compatibility.
1538
1539 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1540 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1541 also .NET compatible.
1542
1543 (b) A recursion or subroutine call to a named group can now be defined as
1544 (?&name) as well as (?P>name).
1545
1546 (c) A backreference to a named group can now be defined as \k<name> or
1547 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1548 5.10, are also .NET compatible.
1549
1550 (d) A conditional reference to a named group can now use the syntax
1551 (?(<name>) or (?('name') as well as (?(name).
1552
1553 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1554 groups (named and numbered) that are never evaluated inline, but can be
1555 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1556 is always false. There may be only one alternative in such a group.
1557
1558 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1559 as the simple (?(R). The condition is true only if the most recent
1560 recursion is that of the given number or name. It does not search out
1561 through the entire recursion stack.
1562
1563 (g) The escape \gN or \g{N} has been added, where N is a positive or
1564 negative number, specifying an absolute or relative reference.
1565
1566 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1567 some "unreachable code" warnings.
1568
1569 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1570 things, this adds five new scripts.
1571
1572 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1573 There were also incompatibilities regarding the handling of \Q..\E inside
1574 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1575 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1576
1577 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1578 matches an empty string, and forcibly breaks the loop. There were bugs in
1579 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1580 against aaaa the result was just "a" rather than "aaaa", for example. Two
1581 separate and independent bugs (that affected different cases) have been
1582 fixed.
1583
1584 39. Refactored the code to abolish the use of different opcodes for small
1585 capturing bracket numbers. This is a tidy that I avoided doing when I
1586 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1587 The new approach is not only tidier, it makes it possible to reduce the
1588 memory needed to fix the previous bug (38).
1589
1590 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1591 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1592 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1593 mode.
1594
1595 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1596 report.
1597
1598 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1599 copy construction and assignment in the C++ wrapper.
1600
1601 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1602 couple of bugs that could have given wrong results in the "--newline=crlf"
1603 case.
1604
1605 44. Added a number of casts and did some reorganization of signed/unsigned int
1606 variables following suggestions from Dair Grant. Also renamed the variable
1607 "this" as "item" because it is a C++ keyword.
1608
1609 45. Arranged for dftables to add
1610
1611 #include "pcre_internal.h"
1612
1613 to pcre_chartables.c because without it, gcc 4.x may remove the array
1614 definition from the final binary if PCRE is built into a static library and
1615 dead code stripping is activated.
1616
1617 46. For an unanchored pattern, if a match attempt fails at the start of a
1618 newline sequence, and the newline setting is CRLF or ANY, and the next two
1619 characters are CRLF, advance by two characters instead of one.
1620
1621
1622 Version 6.7 04-Jul-06
1623 ---------------------
1624
1625 1. In order to handle tests when input lines are enormously long, pcretest has
1626 been re-factored so that it automatically extends its buffers when
1627 necessary. The code is crude, but this _is_ just a test program. The
1628 default size has been increased from 32K to 50K.
1629
1630 2. The code in pcre_study() was using the value of the re argument before
1631 testing it for NULL. (Of course, in any sensible call of the function, it
1632 won't be NULL.)
1633
1634 3. The memmove() emulation function in pcre_internal.h, which is used on
1635 systems that lack both memmove() and bcopy() - that is, hardly ever -
1636 was missing a "static" storage class specifier.
1637
1638 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1639 containing an extended class (one that cannot be represented by a bitmap
1640 because it contains high-valued characters or Unicode property items, e.g.
1641 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1642 pattern, but PCRE should not loop if you do not (it no longer does).
1643 [Detail: two cases were found: (a) a repeated subpattern containing an
1644 extended class; (b) a recursive reference to a subpattern that followed a
1645 previous extended class. It wasn't skipping over the extended class
1646 correctly when UTF-8 mode was not set.]
1647
1648 5. A negated single-character class was not being recognized as fixed-length
1649 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1650 compile error "lookbehind assertion is not fixed length".
1651
1652 6. The RunPerlTest auxiliary script was showing an unexpected difference
1653 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1654 write a Perl script that can interpret lines of an input file either as
1655 byte characters or as UTF-8, which is what "perltest" was being required to
1656 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1657 can't do is switch easily at run time between having the "use utf8;" pragma
1658 or not. In the end, I fudged it by using the RunPerlTest script to insert
1659 "use utf8;" explicitly for the UTF-8 tests.
1660
1661 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1662 the end of the subject string, contrary to the documentation and to what
1663 Perl does. This was true of both matching functions. Now it matches only at
1664 the start of the subject and immediately after *internal* newlines.
1665
1666 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1667 a pointer to an int instead of a pointer to an unsigned long int. This
1668 caused problems on 64-bit systems.
1669
1670 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1671 instance of the 'standard' template library not being so standard".
1672
1673 10. There was no check on the number of named subpatterns nor the maximum
1674 length of a subpattern name. The product of these values is used to compute
1675 the size of the memory block for a compiled pattern. By supplying a very
1676 long subpattern name and a large number of named subpatterns, the size
1677 computation could be caused to overflow. This is now prevented by limiting
1678 the length of names to 32 characters, and the number of named subpatterns
1679 to 10,000.
1680
1681 11. Subpatterns that are repeated with specific counts have to be replicated in
1682 the compiled pattern. The size of memory for this was computed from the
1683 length of the subpattern and the repeat count. The latter is limited to
1684 65535, but there was no limit on the former, meaning that integer overflow
1685 could in principle occur. The compiled length of a repeated subpattern is
1686 now limited to 30,000 bytes in order to prevent this.
1687
1688 12. Added the optional facility to have named substrings with the same name.
1689
1690 13. Added the ability to use a named substring as a condition, using the
1691 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1692 are numbers (not recommended). Forward references are permitted.
1693
1694 14. Added forward references in named backreferences (if you see what I mean).
1695
1696 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1697 pattern could run off the end of the subject. For example, the pattern
1698 "(?s)(.{1,5})"8 did this with the subject "ab".
1699
1700 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1701 PCRE_CASELESS was set when matching characters that were quantified with ?
1702 or *.
1703
1704 17. A character class other than a single negated character that had a minimum
1705 but no maximum quantifier - for example [ab]{6,} - was not handled
1706 correctly by pce_dfa_exec(). It would match only one character.
1707
1708 18. A valid (though odd) pattern that looked like a POSIX character
1709 class but used an invalid character after [ (for example [[,abc,]]) caused
1710 pcre_compile() to give the error "Failed: internal error: code overflow" or
1711 in some cases to crash with a glibc free() error. This could even happen if
1712 the pattern terminated after [[ but there just happened to be a sequence of
1713 letters, a binary zero, and a closing ] in the memory that followed.
1714
1715 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1716 over the years. Originally (before any Unicode support), just the bottom 8
1717 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1718 output from "man perlunicode" includes this:
1719
1720 The regular expression compiler produces polymorphic opcodes. That
1721 is, the pattern adapts to the data and automatically switches to
1722 the Unicode character scheme when presented with Unicode data--or
1723 instead uses a traditional byte scheme when presented with byte
1724 data.
1725
1726 Sadly, a wide octal escape does not cause a switch, and in a string with
1727 no other multibyte characters, these octal escapes are treated as before.
1728 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1729 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1730 Unicode string.
1731
1732 I have not perpetrated such confusion in PCRE. Up till now, it took just
1733 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1734 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1735 translate to the appropriate multibyte character.
1736
1737 29. Applied some refactoring to reduce the number of warnings from Microsoft
1738 and Borland compilers. This has included removing the fudge introduced
1739 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1740 a warning about an unused variable.
1741
1742 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1743 characters since release 4.0, because Perl (from release 5.004) does not.
1744 [Or at least, is documented not to: some releases seem to be in conflict
1745 with the documentation.] However, when a pattern was studied with
1746 pcre_study() and all its branches started with \s, PCRE still included VT
1747 as a possible starting character. Of course, this did no harm; it just
1748 caused an unnecessary match attempt.
1749
1750 22. Removed a now-redundant internal flag bit that recorded the fact that case
1751 dependency changed within the pattern. This was once needed for "required
1752 byte" processing, but is no longer used. This recovers a now-scarce options
1753 bit. Also moved the least significant internal flag bit to the most-
1754 significant bit of the word, which was not previously used (hangover from
1755 the days when it was an int rather than a uint) to free up another bit for
1756 the future.
1757
1758 23. Added support for CRLF line endings as well as CR and LF. As well as the
1759 default being selectable at build time, it can now be changed at runtime
1760 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1761 specify that it is scanning data with non-default line endings.
1762
1763 24. Changed the definition of CXXLINK to make it agree with the definition of
1764 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1765
1766 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1767 recursions. This makes a big different to stack usage for some patterns.
1768
1769 26. If a subpattern containing a named recursion or subroutine reference such
1770 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1771 the space required for the compiled pattern went wrong and gave too small a
1772 value. Depending on the environment, this could lead to "Failed: internal
1773 error: code overflow at offset 49" or "glibc detected double free or
1774 corruption" errors.
1775
1776 27. Applied patches from Google (a) to support the new newline modes and (b) to
1777 advance over multibyte UTF-8 characters in GlobalReplace.
1778
1779 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1780 difference for some implementation of PCRE in some Windows version.
1781
1782 29. Added some extra testing facilities to pcretest:
1783
1784 \q<number> in a data line sets the "match limit" value
1785 \Q<number> in a data line sets the "match recursion limt" value
1786 -S <number> sets the stack size, where <number> is in megabytes
1787
1788 The -S option isn't available for Windows.
1789
1790
1791 Version 6.6 06-Feb-06
1792 ---------------------
1793
1794 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1795 in pcreposix.h. I have copied the definition from pcre.h.
1796
1797 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1798 because pcre.h is no longer a built file.
1799
1800 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1801 not normally included in the compiled code.
1802
1803
1804 Version 6.5 01-Feb-06
1805 ---------------------
1806
1807 1. When using the partial match feature with pcre_dfa_exec(), it was not
1808 anchoring the second and subsequent partial matches at the new starting
1809 point. This could lead to incorrect results. For example, with the pattern
1810 /1234/, partially matching against "123" and then "a4" gave a match.
1811
1812 2. Changes to pcregrep:
1813
1814 (a) All non-match returns from pcre_exec() were being treated as failures
1815 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1816 error message is output. Some extra information is given for the
1817 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1818 probably the only errors that are likely to be caused by users (by
1819 specifying a regex that has nested indefinite repeats, for instance).
1820 If there are more than 20 of these errors, pcregrep is abandoned.
1821
1822 (b) A binary zero was treated as data while matching, but terminated the
1823 output line if it was written out. This has been fixed: binary zeroes
1824 are now no different to any other data bytes.
1825
1826 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1827 used to set a locale for matching. The --locale=xxxx long option has
1828 been added (no short equivalent) to specify a locale explicitly on the
1829 pcregrep command, overriding the environment variables.
1830
1831 (d) When -B was used with -n, some line numbers in the output were one less
1832 than they should have been.
1833
1834 (e) Added the -o (--only-matching) option.
1835
1836 (f) If -A or -C was used with -c (count only), some lines of context were
1837 accidentally printed for the final match.
1838
1839 (g) Added the -H (--with-filename) option.
1840
1841 (h) The combination of options -rh failed to suppress file names for files
1842 that were found from directory arguments.
1843
1844 (i) Added the -D (--devices) and -d (--directories) options.
1845
1846 (j) Added the -F (--fixed-strings) option.
1847
1848 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1849
1850 (l) Added the --colo(u)r option.
1851
1852 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1853 is not present by default.
1854
1855 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1856 items such as (?R) or (?1), when the recursion could match a number of
1857 alternatives. If it matched one of the alternatives, but subsequently,
1858 outside the recursion, there was a failure, the code tried to back up into
1859 the recursion. However, because of the way PCRE is implemented, this is not
1860 possible, and the result was an incorrect result from the match.
1861
1862 In order to prevent this happening, the specification of recursion has
1863 been changed so that all such subpatterns are automatically treated as
1864 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1865
1866 4. I had overlooked the fact that, in some locales, there are characters for
1867 which isalpha() is true but neither isupper() nor islower() are true. In
1868 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1869 and ordfeminine) are like this. This affected the treatment of \w and \W
1870 when they appeared in character classes, but not when they appeared outside
1871 a character class. The bit map for "word" characters is now created
1872 separately from the results of isalnum() instead of just taking it from the
1873 upper, lower, and digit maps. (Plus the underscore character, of course.)
1874
1875 5. The above bug also affected the handling of POSIX character classes such as
1876 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1877 permanent tables. Instead, the bit maps for such a class were previously
1878 created as the appropriate unions of the upper, lower, and digit bitmaps.
1879 Now they are created by subtraction from the [[:word:]] class, which has
1880 its own bitmap.
1881
1882 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1883 It is created by subtracting the vertical space characters (\x09, \x0a,
1884 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1885 subtraction was done in the overall bitmap for a character class, meaning
1886 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1887 be recognized. This bug has been fixed.
1888
1889 7. Patches from the folks at Google:
1890
1891 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1892 real life, but is still worth protecting against".
1893
1894 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1895 regular expressions".
1896
1897 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1898 have it.
1899
1900 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1901 "configure" and the latter not, in order to fix a problem somebody had
1902 with compiling the Arg class on HP-UX.
1903
1904 (e) Improve the error-handling of the C++ wrapper a little bit.
1905
1906 (f) New tests for checking recursion limiting.
1907
1908 8. The pcre_memmove() function, which is used only if the environment does not
1909 have a standard memmove() function (and is therefore rarely compiled),
1910 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1911 returning a result (though PCRE never actually uses the result).
1912
1913 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1914 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1915 returned instead of calling malloc() with an overflowing number that would
1916 most likely cause subsequent chaos.
1917
1918 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1919
1920 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1921 with this option is matched, the nmatch and pmatch options of regexec() are
1922 ignored.
1923
1924 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1925 provided in case anyone wants to the the POSIX interface with UTF-8
1926 strings.
1927
1928 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1929 C++ linking (needed for some HP-UX environments).
1930
1931 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1932 (unused parameter) and in the pcre_printint() function (omitted "default"
1933 switch label when the default is to do nothing).
1934
1935 15. Added some code to make it possible, when PCRE is compiled as a C++
1936 library, to replace subject pointers for pcre_exec() with a smart pointer
1937 class, thus making it possible to process discontinuous strings.
1938
1939 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1940 much the same function. They were added by different people who were trying
1941 to make PCRE easy to compile on non-Unix systems. It has been suggested
1942 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1943 for compiling on Windows systems. I have therefore replaced it with
1944 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1945 defaults to "extern" for C or "extern C" for C++, which works fine on
1946 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1947 SCOPE with something explicit in config.h. In addition:
1948
1949 (a) pcreposix.h still had just "extern" instead of either of these macros;
1950 I have replaced it with PCRE_DATA_SCOPE.
1951
1952 (b) Functions such as _pcre_xclass(), which are internal to the library,
1953 but external in the C sense, all had PCRE_EXPORT in their definitions.
1954 This is apparently wrong for the Windows case, so I have removed it.
1955 (It makes no difference on Unix-like systems.)
1956
1957 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1958 of recursive calls to match(). This is different to MATCH_LIMIT because
1959 that limits the total number of calls to match(), not all of which increase
1960 the depth of recursion. Limiting the recursion depth limits the amount of
1961 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1962 when PCRE is compiled, and changed at run time. A patch from Google adds
1963 this functionality to the C++ interface.
1964
1965 18. Changes to the handling of Unicode character properties:
1966
1967 (a) Updated the table to Unicode 4.1.0.
1968
1969 (b) Recognize characters that are not in the table as "Cn" (undefined).
1970
1971 (c) I revised the way the table is implemented to a much improved format
1972 which includes recognition of ranges. It now supports the ranges that
1973 are defined in UnicodeData.txt, and it also amalgamates other
1974 characters into ranges. This has reduced the number of entries in the
1975 table from around 16,000 to around 3,000, thus reducing its size
1976 considerably. I realized I did not need to use a tree structure after
1977 all - a binary chop search is just as efficient. Having reduced the
1978 number of entries, I extended their size from 6 bytes to 8 bytes to
1979 allow for more data.
1980
1981 (d) Added support for Unicode script names via properties such as \p{Han}.
1982
1983 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1984 matching that character.
1985
1986 20. When matching a repeated Unicode property with a minimum greater than zero,
1987 (for example \pL{2,}), PCRE could look past the end of the subject if it
1988 reached it while seeking the minimum number of characters. This could
1989 happen only if some of the characters were more than one byte long, because
1990 there is a check for at least the minimum number of bytes.
1991
1992 21. Refactored the implementation of \p and \P so as to be more general, to
1993 allow for more different types of property in future. This has changed the
1994 compiled form incompatibly. Anybody with saved compiled patterns that use
1995 \p or \P will have to recompile them.
1996
1997 22. Added "Any" and "L&" to the supported property types.
1998
1999 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2000 but give a compile time error if the value is greater than 0xff.
2001
2002 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2003 accidentally not being installed or uninstalled.
2004
2005 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2006 made were to insert the current release number. This seemed silly, because
2007 it made things harder for people building PCRE on systems that don't run
2008 "configure". I have turned pcre.h into a distributed file, no longer built
2009 by "configure", with the version identification directly included. There is
2010 no longer a pcre.h.in file.
2011
2012 However, this change necessitated a change to the pcre-config script as
2013 well. It is built from pcre-config.in, and one of the substitutions was the
2014 release number. I have updated configure.ac so that ./configure now finds
2015 the release number by grepping pcre.h.
2016
2017 26. Added the ability to run the tests under valgrind.
2018
2019
2020 Version 6.4 05-Sep-05
2021 ---------------------
2022
2023 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2024 "--" to be printed when multiple files were scanned, even when none of the
2025 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2026 consider it to be a bug, and have restored the previous behaviour.
2027
2028 2. A couple of code tidies to get rid of compiler warnings.
2029
2030 3. The pcretest program used to cheat by referring to symbols in the library
2031 whose names begin with _pcre_. These are internal symbols that are not
2032 really supposed to be visible externally, and in some environments it is
2033 possible to suppress them. The cheating is now confined to including
2034 certain files from the library's source, which is a bit cleaner.
2035
2036 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2037 file's purpose clearer.
2038
2039 5. Reorganized pcre_ucp_findchar().
2040
2041
2042 Version 6.3 15-Aug-05
2043 ---------------------
2044
2045 1. The file libpcre.pc.in did not have general read permission in the tarball.
2046
2047 2. There were some problems when building without C++ support:
2048
2049 (a) If C++ support was not built, "make install" and "make test" still
2050 tried to test it.
2051
2052 (b) There were problems when the value of CXX was explicitly set. Some
2053 changes have been made to try to fix these, and ...
2054
2055 (c) --disable-cpp can now be used to explicitly disable C++ support.
2056
2057 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2058 backslash in a target when C++ was disabled. This confuses some
2059 versions of "make", apparently. Using an intermediate variable solves
2060 this. (Same for CPP_LOBJ.)
2061
2062 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2063 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2064 necessary on certain architectures.
2065
2066 4. Added a setting of -export-symbols-regex to the link command to remove
2067 those symbols that are exported in the C sense, but actually are local
2068 within the library, and not documented. Their names all begin with
2069 "_pcre_". This is not a perfect job, because (a) we have to except some
2070 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2071 available (and never for static libraries). I have made a note to try to
2072 find a way round (a) in the future.
2073
2074
2075 Version 6.2 01-Aug-05
2076 ---------------------
2077
2078 1. There was no test for integer overflow of quantifier values. A construction
2079 such as {1111111111111111} would give undefined results. What is worse, if
2080 a minimum quantifier for a parenthesized subpattern overflowed and became
2081 negative, the calculation of the memory size went wrong. This could have
2082 led to memory overwriting.
2083
2084 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2085
2086 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2087 operating environments where this matters.
2088
2089 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2090 PCRE options from within the C++ wrapper.
2091
2092 5. Named capturing subpatterns were not being correctly counted when a pattern
2093 was compiled. This caused two problems: (a) If there were more than 100
2094 such subpatterns, the calculation of the memory needed for the whole
2095 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2096 back references of the form \12, where the number was greater than 9, were
2097 not recognized as back references, even though there were sufficient
2098 previous subpatterns.
2099
2100 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2101 versions of gcc, e.g. 2.95.4.
2102
2103
2104 Version 6.1 21-Jun-05
2105 ---------------------
2106
2107 1. There was one reference to the variable "posix" in pcretest.c that was not
2108 surrounded by "#if !defined NOPOSIX".
2109
2110 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2111 the cross-check on the old pcre_info() function, for the benefit of the
2112 cut-down version of PCRE that is currently imported into Exim.
2113
2114 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2115 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2116 patterns that start (?i) but I don't think that matters. The use of (?i) is
2117 just an example; this all applies to the other options as well.
2118
2119 4. Since libtool seems to echo the compile commands it is issuing, the output
2120 from "make" can be reduced a bit by putting "@" in front of each libtool
2121 compile command.
2122
2123 5. Patch from the folks at Google for configure.in to be a bit more thorough
2124 in checking for a suitable C++ installation before trying to compile the
2125 C++ stuff. This should fix a reported problem when a compiler was present,
2126 but no suitable headers.
2127
2128 6. The man pages all had just "PCRE" as their title. I have changed them to
2129 be the relevant file name. I have also arranged that these names are
2130 retained in the file doc/pcre.txt, which is a concatenation in text format
2131 of all the man pages except the little individual ones for each function.
2132
2133 7. The NON-UNIX-USE file had not been updated for the different set of source
2134 files that come with release 6. I also added a few comments about the C++
2135 wrapper.
2136
2137
2138 Version 6.0 07-Jun-05
2139 ---------------------
2140
2141 1. Some minor internal re-organization to help with my DFA experiments.
2142
2143 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2144 didn't matter for the library itself when fully configured, but did matter
2145 when compiling without UCP support, or within Exim, where the ucp files are
2146 not imported.
2147
2148 3. Refactoring of the library code to split up the various functions into
2149 different source modules. The addition of the new DFA matching code (see
2150 below) to a single monolithic source would have made it really too
2151 unwieldy, quite apart from causing all the code to be include in a
2152 statically linked application, when only some functions are used. This is
2153 relevant even without the DFA addition now that patterns can be compiled in
2154 one application and matched in another.
2155
2156 The downside of splitting up is that there have to be some external
2157 functions and data tables that are used internally in different modules of
2158 the library but which are not part of the API. These have all had their
2159 names changed to start with "_pcre_" so that they are unlikely to clash
2160 with other external names.
2161
2162 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2163 a different (DFA) algorithm. Although it is slower than the original
2164 function, it does have some advantages for certain types of matching
2165 problem.
2166
2167 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2168 including restarting after a partial match.
2169
2170 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2171 defined when compiling for Windows was sent to me. I have put it into the
2172 code, though I have no means of testing or verifying it.
2173
2174 7. Added the pcre_refcount() auxiliary function.
2175
2176 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2177 match before or at the first newline in the subject string. In pcretest,
2178 the /f option on a pattern can be used to set this.
2179
2180 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2181 would behave wrongly. This has been present in PCRE since release 4.0.
2182
2183 10. A number of changes to the pcregrep command:
2184
2185 (a) Refactored how -x works; insert ^(...)$ instead of setting
2186 PCRE_ANCHORED and checking the length, in preparation for adding
2187 something similar for -w.
2188
2189 (b) Added the -w (match as a word) option.
2190
2191 (c) Refactored the way lines are read and buffered so as to have more
2192 than one at a time available.
2193
2194 (d) Implemented a pcregrep test script.
2195
2196 (e) Added the -M (multiline match) option. This allows patterns to match
2197 over several lines of the subject. The buffering ensures that at least
2198 8K, or the rest of the document (whichever is the shorter) is available
2199 for matching (and similarly the previous 8K for lookbehind assertions).
2200
2201 (f) Changed the --help output so that it now says
2202
2203 -w, --word-regex(p)
2204
2205 instead of two lines, one with "regex" and the other with "regexp"
2206 because that confused at least one person since the short forms are the
2207 same. (This required a bit of code, as the output is generated
2208 automatically from a table. It wasn't just a text change.)
2209
2210 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2211 option but starts with a hyphen. Could be a pattern or a path name
2212 starting with a hyphen, for instance.
2213
2214 (h) "-" can be given as a file name to represent stdin.
2215
2216 (i) When file names are being printed, "(standard input)" is used for
2217 the standard input, for compatibility with GNU grep. Previously
2218 "<stdin>" was used.
2219
2220 (j) The option --label=xxx can be used to supply a name to be used for
2221 stdin when file names are being printed. There is no short form.
2222
2223 (k) Re-factored the options decoding logic because we are going to add
2224 two more options that take data. Such options can now be given in four
2225 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2226
2227 (l) Added the -A, -B, and -C options for requesting that lines of context
2228 around matches be printed.
2229
2230 (m) Added the -L option to print the names of files that do not contain
2231 any matching lines, that is, the complement of -l.
2232
2233 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2234 continue to scan other files.
2235
2236 (o) The -s option was incorrectly implemented. For compatibility with other
2237 greps, it now suppresses the error message for a non-existent or non-
2238 accessible file (but not the return code). There is a new option called
2239 -q that suppresses the output of matching lines, which was what -s was
2240 previously doing.
2241
2242 (p) Added --include and --exclude options to specify files for inclusion
2243 and exclusion when recursing.
2244
2245 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2246 Hopefully, it now does.
2247
2248 12. Missing cast in pcre_study().
2249
2250 13. Added an "uninstall" target to the makefile.
2251
2252 14. Replaced "extern" in the function prototypes in Makefile.in with
2253 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2254 world, but is set differently for Windows.
2255
2256 15. Added a second compiling function called pcre_compile2(). The only
2257 difference is that it has an extra argument, which is a pointer to an
2258 integer error code. When there is a compile-time failure, this is set
2259 non-zero, in addition to the error test pointer being set to point to an
2260 error message. The new argument may be NULL if no error number is required
2261 (but then you may as well call pcre_compile(), which is now just a
2262 wrapper). This facility is provided because some applications need a
2263 numeric error indication, but it has also enabled me to tidy up the way
2264 compile-time errors are handled in the POSIX wrapper.
2265
2266 16. Added VPATH=.libs to the makefile; this should help when building with one
2267 prefix path and installing with another. (Or so I'm told by someone who
2268 knows more about this stuff than I do.)
2269
2270 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2271 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2272 match everything, including newlines. This is not POSIX-compatible, but
2273 somebody wanted the feature. From pcretest it can be activated by using
2274 both the P and the s flags.
2275
2276 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2277
2278 19. libpcre.pc was being incorrectly installed as executable.
2279
2280 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2281 it now also looks for '\r' so that it will work unmodified on Windows.
2282
2283 21. Added Google's contributed C++ wrapper to the distribution.
2284
2285 22. Added some untidy missing memory free() calls in pcretest, to keep
2286 Electric Fence happy when testing.
2287
2288
2289
2290 Version 5.0 13-Sep-04
2291 ---------------------
2292
2293 1. Internal change: literal characters are no longer packed up into items
2294 containing multiple characters in a single byte-string. Each character
2295 is now matched using a separate opcode. However, there may be more than one
2296 byte in the character in UTF-8 mode.
2297
2298 2. The pcre_callout_block structure has two new fields: pattern_position and
2299 next_item_length. These contain the offset in the pattern to the next match
2300 item, and its length, respectively.
2301
2302 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2303 insertion of callouts before each pattern item. Added the /C option to
2304 pcretest to make use of this.
2305
2306 4. On the advice of a Windows user, the lines
2307
2308 #if defined(_WIN32) || defined(WIN32)
2309 _setmode( _fileno( stdout ), 0x8000 );
2310 #endif /* defined(_WIN32) || defined(WIN32) */
2311
2312 have been added to the source of pcretest. This apparently does useful
2313 magic in relation to line terminators.
2314
2315 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2316 for the benefit of those environments where the "b" makes a difference.
2317
2318 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2319 to know about it. I have put a hack into configure.in that adds in code
2320 to set GCC=yes if CC=icc. This seems to end up at a point in the
2321 generated configure script that is early enough to affect the setting of
2322 compiler options, which is what is needed, but I have no means of testing
2323 whether it really works. (The user who reported this had patched the
2324 generated configure script, which of course I cannot do.)
2325
2326 LATER: After change 22 below (new libtool files), the configure script
2327 seems to know about icc (and also ecc). Therefore, I have commented out
2328 this hack in configure.in.
2329
2330 7. Added support for pkg-config (2 patches were sent in).
2331
2332 8. Negated POSIX character classes that used a combination of internal tables
2333 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2334 [[:^ascii]]. Typically, they would match almost any characters. The other
2335 POSIX classes were not broken in this way.
2336
2337 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2338 to find the match, as PCRE was deluded into thinking that the match had to
2339 start at the start point or following a newline. The same bug applied to
2340 patterns with negative forward assertions or any backward assertions
2341 preceding ".*" at the start, unless the pattern required a fixed first
2342 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2343
2344 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2345 starting at the last subject character, bytes beyond the end of the subject
2346 string were read.
2347
2348 11. Renamed the variable "class" as "classbits" to make life easier for C++
2349 users. (Previously there was a macro definition, but it apparently wasn't
2350 enough.)
2351
2352 12. Added the new field "tables" to the extra data so that tables can be passed
2353 in at exec time, or the internal tables can be re-selected. This allows
2354 a compiled regex to be saved and re-used at a later time by a different
2355 program that might have everything at different addresses.
2356
2357 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2358 -R library as well as a -L library.
2359
2360 14. The debugging options of pcretest (-d on the command line or D on a
2361 pattern) showed incorrect output for anything following an extended class
2362 that contained multibyte characters and which was followed by a quantifier.
2363
2364 15. Added optional support for general category Unicode character properties
2365 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2366 support. It adds about 90K to the size of the library. The meanings of the
2367 inbuilt class escapes such as \d and \s have NOT been changed.
2368
2369 16. Updated pcredemo.c to include calls to free() to release the memory for the
2370 compiled pattern.
2371
2372 17. The generated file chartables.c was being created in the source directory
2373 instead of in the building directory. This caused the build to fail if the
2374 source directory was different from the building directory, and was
2375 read-only.
2376
2377 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2378 file. No doubt somebody will tell me if they don't make sense... Also added
2379 Dan Mooney's comments about building on OpenVMS.
2380
2381 19. Added support for partial matching via the PCRE_PARTIAL option for
2382 pcre_exec() and the \P data escape in pcretest.
2383
2384 20. Extended pcretest with 3 new pattern features:
2385
2386 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2387 write the compiled pattern to the file whose name is "rest-of-line".
2388 This is a straight binary dump of the data, with the saved pointer to
2389 the character tables forced to be NULL. The study data, if any, is
2390 written too. After writing, pcretest reads a new pattern.
2391
2392 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2393 compiled pattern from the given file. There must not be any
2394 occurrences of "<" in the file name (pretty unlikely); if there are,
2395 pcretest will instead treat the initial "<" as a pattern delimiter.
2396 After reading in the pattern, pcretest goes on to read data lines as
2397 usual.
2398
2399 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2400 and 16-bit fields in a compiled pattern, to simulate a pattern that
2401 was compiled on a host of opposite endianness.
2402
2403 21. The pcre-exec() function can now cope with patterns that were compiled on
2404 hosts of opposite endianness, with this restriction:
2405
2406 As for any compiled expression that is saved and used later, the tables
2407 pointer field cannot be preserved; the extra_data field in the arguments
2408 to pcre_exec() should be used to pass in a tables address if a value
2409 other than the default internal tables were used at compile time.
2410
2411 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2412 now diagnosed as an error. Previously, most of the time, a negative number
2413 would have been treated as zero, but if in addition "ovector" was passed as
2414 NULL, a crash could occur.
2415
2416 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2417 new versions from the libtool 1.5 distribution (the last one is a copy of
2418 a file called libtool.m4). This seems to have fixed the need to patch
2419 "configure" to support Darwin 1.3 (which I used to do). However, I still
2420 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2421 workstation).
2422
2423 24. Changed the PCRE licence to be the more standard "BSD" licence.
2424
2425
2426 Version 4.5 01-Dec-03
2427 ---------------------
2428
2429 1. There has been some re-arrangement of the code for the match() function so
2430 that it can be compiled in a version that does not call itself recursively.
2431 Instead, it keeps those local variables that need separate instances for
2432 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2433 needs to "recurse". Keeping track of where control must go is done by means
2434 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2435 hide most of the details from the main code, and operates only if
2436 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2437 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2438 operating.
2439
2440 To make it easier for callers to provide specially tailored get/free
2441 functions for this usage, two new functions, pcre_stack_malloc, and
2442 pcre_stack_free, are used. They are always called in strict stacking order,
2443 and the size of block requested is always the same.
2444
2445 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2446 PCRE has been compiled to use the stack or the heap for recursion. The
2447 -C option of pcretest uses this to show which version is compiled.
2448
2449 A new data escape \S, is added to pcretest; it causes the amounts of store
2450 obtained and freed by both kinds of malloc/free at match time to be added
2451 to the output.
2452
2453 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2454 what's available on my current Linux desktop machine.
2455
2456 3. When matching a UTF-8 string, the test for a valid string at the start has
2457 been extended. If start_offset is not zero, PCRE now checks that it points
2458 to a byte that is the start of a UTF-8 character. If not, it returns
2459 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2460 this is necessary because there may be backward assertions in the pattern.
2461 When matching the same subject several times, it may save resources to use
2462 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2463
2464 4. The code for checking the validity of UTF-8 strings has been tightened so
2465 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2466 containing "overlong sequences".
2467
2468 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2469 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2470 should have been "&", but it just so happened that all the cases this let
2471 through by mistake were picked up later in the function.
2472
2473 6. I had used a variable called "isblank" - this is a C99 function, causing
2474 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2475
2476 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2477 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2478 so the version is shown; (c) stop "make test" echoing "./RunTest".
2479
2480 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2481
2482 9. The prototype for memmove() for systems that don't have it was using
2483 size_t, but the inclusion of the header that defines size_t was later. I've
2484 moved the #includes for the C headers earlier to avoid this.
2485
2486 10. Added some adjustments to the code to make it easier to compiler on certain
2487 special systems:
2488
2489 (a) Some "const" qualifiers were missing.
2490 (b) Added the macro EXPORT before all exported functions; by default this
2491 is defined to be empty.
2492 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2493 that it reads its output file name as an argument instead of writing
2494 to the standard output and assuming this can be redirected.
2495
2496 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2497 class containing characters with values greater than 255, PCRE compilation
2498 went into a loop.
2499
2500 12. A recursive reference to a subpattern that was within another subpattern
2501 that had a minimum quantifier of zero caused PCRE to crash. For example,
2502 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2503 recursion. If the recursively-called subpattern itself had a zero repeat,
2504 that was OK.
2505
2506 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2507 buffer into which it was copied (for escape processing) was still set at
2508 1024, so long lines caused crashes.
2509
2510 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2511 "internal error: code overflow...". This applied to any character class
2512 that was followed by a possessive quantifier.
2513
2514 15. Modified the Makefile to add libpcre.la as a prerequisite for
2515 libpcreposix.la because I was told this is needed for a parallel build to
2516 work.
2517
2518 16. If a pattern that contained .* following optional items at the start was
2519 studied, the wrong optimizing data was generated, leading to matching
2520 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2521 matching string must start with a or b or c. The correct conclusion for
2522 this pattern is that a match can start with any character.
2523
2524
2525 Version 4.4 13-Aug-03
2526 ---------------------
2527
2528 1. In UTF-8 mode, a character class containing characters with values between
2529 127 and 255 was not handled correctly if the compiled pattern was studied.
2530 In fixing this, I have also improved the studying algorithm for such
2531 classes (slightly).
2532
2533 2. Three internal functions had redundant arguments passed to them. Removal
2534 might give a very teeny performance improvement.
2535
2536 3. Documentation bug: the value of the capture_top field in a callout is *one
2537 more than* the number of the hightest numbered captured substring.
2538
2539 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2540 in incorrectly linking with a previously installed version. They now link
2541 explicitly with libpcre.la.
2542
2543 5. configure.in no longer needs to recognize Cygwin specially.
2544
2545 6. A problem in pcre.in for Windows platforms is fixed.
2546
2547 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2548 pcretest, it used to include the size of the study block as part of its
2549 output. Unfortunately, the structure contains a field that has a different
2550 size on different hardware architectures. This meant that the tests that
2551 showed this size failed. As the block is currently always of a fixed size,
2552 this information isn't actually particularly useful in pcretest output, so
2553 I have just removed it.
2554
2555 8. Three pre-processor statements accidentally did not start in column 1.
2556 Sadly, there are *still* compilers around that complain, even though
2557 standard C has not required this for well over a decade. Sigh.
2558
2559 9. In pcretest, the code for checking callouts passed small integers in the
2560 callout_data field, which is a void * field. However, some picky compilers
2561 complained about the casts involved for this on 64-bit systems. Now
2562 pcretest passes the address of the small integer instead, which should get
2563 rid of the warnings.
2564
2565 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2566 both compile and run time, and gives an error if an invalid UTF-8 sequence
2567 is found. There is a option for disabling this check in cases where the
2568 string is known to be correct and/or the maximum performance is wanted.
2569
2570 11. In response to a bug report, I changed one line in Makefile.in from
2571
2572 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2573 to
2574 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2575
2576 to look similar to other lines, but I have no way of telling whether this
2577 is the right thing to do, as I do not use Windows. No doubt I'll get told
2578 if it's wrong...
2579
2580
2581 Version 4.3 21-May-03
2582 ---------------------
2583
2584 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2585 Makefile.
2586
2587 2. Some refactoring to improve the quality of the code:
2588
2589 (i) The utf8_table... variables are now declared "const".
2590
2591 (ii) The code for \cx, which used the "case flipping" table to upper case
2592 lower case letters, now just substracts 32. This is ASCII-specific,
2593 but the whole concept of \cx is ASCII-specific, so it seems
2594 reasonable.
2595
2596 (iii) PCRE was using its character types table to recognize decimal and
2597 hexadecimal digits in the pattern. This is silly, because it handles
2598 only 0-9, a-f, and A-F, but the character types table is locale-
2599 specific, which means strange things might happen. A private
2600 table is now used for this - though it costs 256 bytes, a table is
2601 much faster than multiple explicit tests. Of course, the standard
2602 character types table is still used for matching digits in subject
2603 strings against \d.
2604
2605 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2606 ending in _t are). So I've renamed it as ESC_tee.
2607
2608 3. The first argument for regexec() in the POSIX wrapper should have been
2609 defined as "const".
2610
2611 4. Changed pcretest to use malloc() for its buffers so that they can be
2612 Electric Fenced for debugging.
2613
2614 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2615 to read one or more bytes before the start of the subject string. Often this
2616 had no effect on PCRE's behaviour, but in some circumstances it could
2617 provoke a segmentation fault.
2618
2619 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2620 to try to read one or more bytes before the start of the subject string.
2621
2622 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2623 UTF-8 support could misbehave in various ways if the subject string
2624 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2625 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2626 back over UTF-8 characters.)
2627
2628
2629 Version 4.2 14-Apr-03
2630 ---------------------
2631
2632 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2633
2634 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2635 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2636 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2637 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2638 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2639 and BUILD_EXEEXT
2640 Note: automatic setting of the BUILD variables is not yet working
2641 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2642 compile-time but not at link-time
2643 [LINK]: use for linking executables only
2644 make different versions for Windows and non-Windows
2645 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2646 libraries
2647 [LINK_FOR_BUILD]: new variable
2648 [OBJEXT]: use throughout
2649 [EXEEXT]: use throughout
2650 <winshared>: new target
2651 <wininstall>: new target
2652 <dftables.o>: use native compiler
2653 <dftables>: use native linker
2654 <install>: handle Windows platform correctly
2655 <clean>: ditto
2656 <check>: ditto
2657 copy DLL to top builddir before testing
2658
2659 As part of these changes, -no-undefined was removed again. This was reported
2660 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2661 in any case.
2662
2663 3. Some tidies to get rid of compiler warnings:
2664
2665 . In the match_data structure, match_limit was an unsigned long int, whereas
2666 match_call_count was an int. I've made them both unsigned long ints.
2667
2668 . In pcretest the fact that a const uschar * doesn't automatically cast to
2669 a void * provoked a warning.
2670
2671 . Turning on some more compiler warnings threw up some "shadow" variables
2672 and a few more missing casts.
2673
2674 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2675 option, a class that contained a single character with a value between 128
2676 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2677
2678 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2679 option, a class that contained several characters, but with at least one
2680 whose value was between 128 and 255 caused PCRE to crash.
2681
2682
2683 Version 4.1 12-Mar-03
2684 ---------------------
2685
2686 1. Compiling with gcc -pedantic found a couple of places where casts were
2687 needed, and a string in dftables.c that was longer than standard compilers are
2688 required to support.
2689
2690 2. Compiling with Sun's compiler found a few more places where the code could
2691 be tidied up in order to avoid warnings.
2692
2693 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2694 first of these names is deprecated in the latest Autoconf in favour of the name
2695 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2696 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2697 analogy I have changed it to CFLAGS_FOR_BUILD.
2698
2699 4. Added -no-undefined to the linking command in the Makefile, because this is
2700 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2701 linking step for the pcreposix library.
2702
2703 5. PCRE was failing to diagnose the case of two named groups with the same
2704 name.
2705
2706 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2707 literal character that is needed in the subject for a match, and scans along to
2708 ensure that it is present before embarking on the full matching process. This
2709 saves time in cases of nested unlimited repeats that are never going to match.
2710 Problem: the scan can take a lot of time if the subject is very long (e.g.
2711 megabytes), thus penalizing straightforward matches. It is now done only if the
2712 amount of subject to be scanned is less than 1000 bytes.
2713
2714 7. A lesser problem with the same optimization is that it was recording the
2715 first character of an anchored pattern as "needed", thus provoking a search
2716 right along the subject, even when the first match of the pattern was going to
2717 fail. The "needed" character is now not set for anchored patterns, unless it
2718 follows something in the pattern that is of non-fixed length. Thus, it still
2719 fulfils its original purpose of finding quick non-matches in cases of nested
2720 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2721
2722
2723 Version 4.0 17-Feb-03
2724 ---------------------
2725
2726 1. If a comment in an extended regex that started immediately after a meta-item
2727 extended to the end of string, PCRE compiled incorrect data. This could lead to
2728 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2729
2730 2. Moved to autoconf 2.53 and libtool 1.4.2.
2731
2732 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2733 the special perltest8 script is no longer needed - all the tests can be run
2734 from a single perltest script.
2735
2736 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2737 by \s. It has now been removed in PCRE. This means it isn't recognized as
2738 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2739 class [:space:] *does* include VT, thereby creating a mess.
2740
2741 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2742 space and tab.
2743
2744 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2745 its new features into the main test script, reducing the number of scripts.
2746
2747 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2748 were backward compatible, and made the (?i) apply to the whole pattern, as if
2749 /i were given. Now it behaves more logically, and applies the option setting
2750 only to what follows. PCRE has been changed to follow suit. However, if it
2751 finds options settings right at the start of the pattern, it extracts them into
2752 the global options, as before. Thus, they show up in the info data.
2753
2754 8. Added support for the \Q...\E escape sequence. Characters in between are
2755 treated as literals. This is slightly different from Perl in that $ and @ are
2756 also handled as literals inside the quotes. In Perl, they will cause variable
2757 interpolation. Note the following examples:
2758
2759 Pattern PCRE matches Perl matches
2760
2761 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2762 \Qabc\$xyz\E abc\$xyz abc\$xyz
2763 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2764
2765 For compatibility with Perl, \Q...\E sequences are recognized inside character
2766 classes as well as outside them.
2767
2768 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2769 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2770 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2771 signed/unsigned warnings.
2772
2773 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2774 option for pcretest, so I've replaced it by a simple function that does just
2775 that job.
2776
2777 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2778 "pcregrep -".
2779
2780 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2781 Java package. This provides some syntactic sugar for simple cases of what my
2782 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2783 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2784 item, you can use this simplified notation. Note that only makes sense with
2785 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2786 greediness, whatever the setting of the PCRE_UNGREEDY option.
2787
2788 13. A change of greediness default within a pattern was not taking effect at
2789 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2790 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2791 was abstracted outside.
2792
2793 14. PCRE now supports the \G assertion. It is true when the current matching
2794 position is at the start point of the match. This differs from \A when the
2795 starting offset is non-zero. Used with the /g option of pcretest (or similar
2796 code), it works in the same way as it does for Perl's /g option. If all
2797 alternatives of a regex begin with \G, the expression is anchored to the start
2798 match position, and the "anchored" flag is set in the compiled expression.
2799
2800 15. Some bugs concerning the handling of certain option changes within patterns
2801 have been fixed. These applied to options other than (?ims). For example,
2802 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2803 been the other way round. Some of this was related to change 7 above.
2804
2805 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2806 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2807 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2808 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2809
2810 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2811 mode. Unlike ".", it always matches newline, whatever the setting of
2812 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2813 assertions. Perl allows it, but it doesn't (in general) work because it can't
2814 calculate the length of the lookbehind. At least, that's the case for Perl
2815 5.8.0 - I've been told they are going to document that it doesn't work in
2816 future.
2817
2818 18. Added an error diagnosis for escapes that PCRE does not support: these are
2819 \L, \l, \N, \P, \p, \U, \u, and \X.
2820
2821 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2822 reading past the end of the pattern in cases such as /[abcd/.
2823
2824 20. PCRE was getting more memory than necessary for patterns with classes that
2825 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2826
2827 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2828 compiling PCRE for use with Virtual Pascal.
2829
2830 22. Small fix to the Makefile to make it work properly if the build is done
2831 outside the source tree.
2832
2833 23. Added a new extension: a condition to go with recursion. If a conditional
2834 subpattern starts with (?(R) the "true" branch is used if recursion has
2835 happened, whereas the "false" branch is used only at the top level.
2836
2837 24. When there was a very long string of literal characters (over 255 bytes
2838 without UTF support, over 250 bytes with UTF support), the computation of how
2839 much memory was required could be incorrect, leading to segfaults or other
2840 strange effects.
2841
2842 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2843 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2844 there was a subsequent back reference to those brackets. This meant that, for
2845 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2846 possible to check for precisely this case. All we can do is abandon the
2847 optimization if .* occurs inside capturing brackets when there are any back
2848 references whatsoever. (See below for a better fix that came later.)
2849
2850 26. The handling of the optimization for finding the first character of a
2851 non-anchored pattern, and for finding a character that is required later in the
2852 match were failing in some cases. This didn't break the matching; it just
2853 failed to optimize when it could. The way this is done has been re-implemented.
2854
2855 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2856
2857 28. Added a new feature that provides some of the functionality that Perl
2858 provides with (?{...}). The facility is termed a "callout". The way it is done
2859 in PCRE is for the caller to provide an optional function, by setting
2860 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2861 global variable. By default it is unset, which disables all calling out. To get
2862 the function called, the regex must include (?C) at appropriate points. This
2863 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2864 This provides a means of identifying different callout points. When PCRE
2865 reaches such a point in the regex, if pcre_callout has been set, the external
2866 function is called. It is provided with data in a structure called
2867 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2868 matching continues; if it returns a non-zero value, the match at the current
2869 point fails. However, backtracking will occur if possible. [This was changed
2870 later and other features added - see item 49 below.]
2871
2872 29. pcretest is upgraded to test the callout functionality. It provides a
2873 callout function that displays information. By default, it shows the start of
2874 the match and the current position in the text. There are some new data escapes
2875 to vary what happens:
2876
2877 \C+ in addition, show current contents of captured substrings
2878 \C- do not supply a callout function
2879 \C!n return 1 when callout number n is reached
2880 \C!n!m return 1 when callout number n is reached for the mth time
2881
2882 30. If pcregrep was called with the -l option and just a single file name, it
2883 output "<stdin>" if a match was found, instead of the file name.
2884
2885 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2886 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2887 pcre_exec(). This saves a malloc/free per call. The default value of
2888 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2889 when configuring.
2890
2891 32. The default maximum size of a compiled pattern is 64K. There have been a
2892 few cases of people hitting this limit. The code now uses macros to handle the
2893 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2894 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2895 configuring. Tests 2 and 5 work only with 2-byte links because they output
2896 debugging information about compiled patterns.
2897
2898 33. Internal code re-arrangements:
2899
2900 (a) Moved the debugging function for printing out a compiled regex into
2901 its own source file (printint.c) and used #include to pull it into
2902 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2903 separate copies.
2904
2905 (b) Defined the list of op-code names for debugging as a macro in
2906 internal.h so that it is next to the definition of the opcodes.
2907
2908 (c) Defined a table of op-code lengths for simpler skipping along compiled
2909 code. This is again a macro in internal.h so that it is next to the
2910 definition of the opcodes.
2911
2912 34. Added support for recursive calls to individual subpatterns, along the
2913 lines of Robin Houston's patch (but implemented somewhat differently).
2914
2915 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2916 allow it to read and process whole directories in Win32. This code was
2917 contributed by Lionel Fourquaux; it has not been tested by me.
2918
2919 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2920 used to name a group. Names consist of alphanumerics and underscores, and must
2921 be unique. Back references use the syntax (?P=name) and recursive calls use
2922 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2923 numbers. The function pcre_fullinfo() can be used after compilation to extract
2924 a name/number map. There are three relevant calls:
2925
2926 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2927 PCRE_INFO_NAMECOUNT yields the number of entries
2928 PCRE_INFO_NAMETABLE yields a pointer to the map.
2929
2930 The map is a vector of fixed-size entries. The size of each entry depends on
2931 the length of the longest name used. The first two bytes of each entry are the
2932 group number, most significant byte first. There follows the corresponding
2933 name, zero terminated. The names are in alphabetical order.
2934
2935 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2936 case instead of 255. Making it the same both with and without UTF-8 support
2937 means that the same test output works with both.
2938
2939 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2940 calling malloc() with a zero argument.
2941
2942 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2943 optimization. I've improved things by keeping a bitmap of backreferences with
2944 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2945 fact referenced, the optimization can be applied. It is unlikely that a
2946 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2947 the match to follow \n) will appear inside brackets with a number greater than
2948 31, but if it does, any back reference > 31 suppresses the optimization.
2949
2950 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2951 of disabling numbered capturing parentheses. Any opening parenthesis that is
2952 not followed by ? behaves as if it were followed by ?: but named parentheses
2953 can still be used for capturing (and they will acquire numbers in the usual
2954 way).
2955
2956 41. Redesigned the return codes from the match() function into yes/no/error so
2957 that errors can be passed back from deep inside the nested calls. A malloc
2958 failure while inside a recursive subpattern call now causes the
2959 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2960
2961 42. It is now possible to set a limit on the number of times the match()
2962 function is called in a call to pcre_exec(). This facility makes it possible to
2963 limit the amount of recursion and backtracking, though not in a directly
2964 obvious way, because the match() function is used in a number of different
2965 circumstances. The count starts from zero for each position in the subject
2966 string (for non-anchored patterns). The default limit is, for compatibility, a
2967 large number, namely 10 000 000. You can change this in two ways:
2968
2969 (a) When configuring PCRE before making, you can use --with-match-limit=n
2970 to set a default value for the compiled library.
2971
2972 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2973 a different value is set. See 45 below.
2974
2975 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2976
2977 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2978 of things that can be changed at compile time. The first argument specifies
2979 what is wanted and the second points to where the information is to be placed.
2980 The current list of available information is:
2981
2982 PCRE_CONFIG_UTF8
2983
2984 The output is an integer that is set to one if UTF-8 support is available;
2985 otherwise it is set to zero.
2986
2987 PCRE_CONFIG_NEWLINE
2988
2989 The output is an integer that it set to the value of the code that is used for
2990 newline. It is either LF (10) or CR (13).
2991
2992 PCRE_CONFIG_LINK_SIZE
2993
2994 The output is an integer that contains the number of bytes used for internal
2995 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2996
2997 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2998
2999 The output is an integer that contains the threshold above which the POSIX
3000 interface uses malloc() for output vectors. See item 31 above.
3001
3002 PCRE_CONFIG_MATCH_LIMIT
3003
3004 The output is an unsigned integer that contains the default limit of the number
3005 of match() calls in a pcre_exec() execution. See 42 above.
3006
3007 44. pcretest has been upgraded by the addition of the -C option. This causes it
3008 to extract all the available output from the new pcre_config() function, and to
3009 output it. The program then exits immediately.
3010
3011 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3012 order to support additional features. One way would have been to define
3013 pcre_exec2() (for example) with extra arguments, but this would not have been
3014 extensible, and would also have required all calls to the original function to
3015 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3016 is used for passing in "extra" data from pcre_study().
3017
3018 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3019 contains the following fields:
3020
3021 flags a bitmap indicating which of the following fields are set
3022 study_data opaque data from pcre_study()
3023 match_limit a way of specifying a limit on match() calls for a specific
3024 call to pcre_exec()
3025 callout_data data for callouts (see 49 below)
3026
3027 The flag bits are also defined in pcre.h, and are
3028
3029 PCRE_EXTRA_STUDY_DATA
3030 PCRE_EXTRA_MATCH_LIMIT
3031 PCRE_EXTRA_CALLOUT_DATA
3032
3033 The pcre_study() function now returns one of these new pcre_extra blocks, with
3034 the actual study data pointed to by the study_data field, and the
3035 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3036 before. That is, this change is entirely upwards-compatible and requires no
3037 change to existing code.
3038
3039 If you want to pass in additional data to pcre_exec(), you can either place it
3040 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3041 block.
3042
3043 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3044 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3045 times with different match limits, until it finds the minimum value needed for
3046 pcre_exec() to complete. The value is then output. This can be instructive; for
3047 most simple matches the number is quite small, but for pathological cases it
3048 gets very large very quickly.
3049
3050 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3051 returns the size of the data block pointed to by the study_data field in a
3052 pcre_extra block, that is, the value that was passed as the argument to
3053 pcre_malloc() when PCRE was getting memory in which to place the information
3054 created by pcre_study(). The fourth argument should point to a size_t variable.
3055 pcretest has been extended so that this information is shown after a successful
3056 pcre_study() call when information about the compiled regex is being displayed.
3057
3058 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3059 because what follows is always an absolute path. (Later: it turns out that this
3060 is more than cosmetic for MinGW, because it doesn't like empty path
3061 components.)
3062
3063 49. Some changes have been made to the callout feature (see 28 above):
3064
3065 (i) A callout function now has three choices for what it returns:
3066
3067 0 => success, carry on matching
3068 > 0 => failure at this point, but backtrack if possible
3069 < 0 => serious error, return this value from pcre_exec()
3070
3071 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3072 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3073 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3074 use by callout functions. It will never be used by PCRE itself.
3075
3076 (ii) The pcre_extra structure (see 45 above) has a void * field called
3077 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3078 pcre_callout_block structure has a field of the same name. The contents of
3079 the field passed in the pcre_extra structure are passed to the callout
3080 function in the corresponding field in the callout block. This makes it
3081 easier to use the same callout-containing regex from multiple threads. For
3082 testing, the pcretest program has a new data escape
3083
3084 \C*n pass the number n (may be negative) as callout_data
3085
3086 If the callout function in pcretest receives a non-zero value as
3087 callout_data, it returns that value.
3088
3089 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3090 there were some redundant $(CFLAGS) in commands that are now specified as
3091 $(LINK), which already includes $(CFLAGS).
3092
3093 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3094 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3095 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3096 one-byte characters throughout. Note that case-insensitive matching applies
3097 only to characters whose values are less than 256. PCRE doesn't support the
3098 notion of cases for higher-valued characters.
3099
3100 (i) A character class whose characters are all within 0-255 is handled as
3101 a bit map, and the map is inverted for negative classes. Previously, a
3102 character > 255 always failed to match such a class; however it should
3103 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3104
3105 (ii) A negated character class with a single character < 255 is coded as
3106 "not this character" (OP_NOT). This wasn't working properly when the test
3107 character was multibyte, either singly or repeated.
3108
3109 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3110 mode, for example: \x{100}{2,3}.
3111
3112 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3113 singly or repeated) now correctly test multibyte characters. However,
3114 PCRE doesn't recognize any characters with values greater than 255 as
3115 digits, spaces, or word characters. Such characters always match \D, \S,
3116 and \W, and never match \d, \s, or \w.
3117
3118 (v) Classes may now contain characters and character ranges with values
3119 greater than 255. For example: [ab\x{100}-\x{400}].
3120
3121 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3122 PCRE in UTF-8 mode.
3123
3124 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3125 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3126 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3127 value.)
3128
3129 53. The single man page has become too large. I have therefore split it up into
3130 a number of separate man pages. These also give rise to individual HTML pages;
3131 these are now put in a separate directory, and there is an index.html page that
3132 lists them all. Some hyperlinking between the pages has been installed.
3133
3134 54. Added convenience functions for handling named capturing parentheses.
3135
3136 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3137 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3138 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3139 are faulted.
3140
3141 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3142 calling configure. These values are used when compiling the dftables.c program
3143 which is run to generate the source of the default character tables. They
3144 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3145 you will need to set these values.
3146
3147 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3148
3149
3150 Version 3.9 02-Jan-02
3151 ---------------------
3152
3153 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3154
3155 2. If --disable-static was given, the building process failed when trying to
3156 build pcretest and pcregrep. (For some reason it was using libtool to compile
3157 them, which is not right, as they aren't part of the library.)
3158
3159
3160 Version 3.8 18-Dec-01
3161 ---------------------
3162
3163 1. The experimental UTF-8 code was completely screwed up. It was packing the
3164 bytes in the wrong order. How dumb can you get?
3165
3166
3167 Version 3.7 29-Oct-01
3168 ---------------------
3169
3170 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3171 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3172 this didn't happen under Solaris 8, where I normally test things.
3173
3174 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3175 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3176 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3177 this fix an hour or so after the initial 3.7 release.)
3178
3179
3180 Version 3.6 23-Oct-01
3181 ---------------------
3182
3183 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3184 offsets passed as NULL with zero offset count.
3185
3186 2. The config.guess and config.sub files had not been updated when I moved to
3187 the latest autoconf.
3188
3189
3190 Version 3.5 15-Aug-01
3191 ---------------------
3192
3193 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3194 had been forgotten.
3195
3196 2. By using declared but undefined structures, we can avoid using "void"
3197 definitions in pcre.h while keeping the internal definitions of the structures
3198 private.
3199
3200 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3201 user point of view, this means that both static and shared libraries are built
3202 by default, but this can be individually controlled. More of the work of
3203 handling this static/shared cases is now inside libtool instead of PCRE's make
3204 file.
3205
3206 4. The pcretest utility is now installed along with pcregrep because it is
3207 useful for users (to test regexs) and by doing this, it automatically gets
3208 relinked by libtool. The documentation has been turned into a man page, so
3209 there are now .1, .txt, and .html versions in /doc.
3210
3211 5. Upgrades to pcregrep:
3212 (i) Added long-form option names like gnu grep.
3213 (ii) Added --help to list all options with an explanatory phrase.
3214 (iii) Added -r, --recursive to recurse into sub-directories.
3215 (iv) Added -f, --file to read patterns from a file.
3216
3217 6. pcre_exec() was referring to its "code" argument before testing that
3218 argument for NULL (and giving an error if it was NULL).
3219
3220 7. Upgraded Makefile.in to allow for compiling in a different directory from
3221 the source directory.
3222
3223 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3224 options bits, the pointer it was passed was to an int instead of to an unsigned
3225 long int. This mattered only on 64-bit systems.
3226
3227 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3228 generated) instead of pcre.in, which it its source. Also made the same change
3229 in several of the .c files.
3230
3231 10. A new release of gcc defines printf() as a macro, which broke pcretest
3232 because it had an ifdef in the middle of a string argument for printf(). Fixed
3233 by using separate calls to printf().
3234
3235 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3236 script, to force use of CR or LF instead of \n in the source. On non-Unix
3237 systems, the value can be set in config.h.
3238
3239 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3240 absolute limit. Changed the text of the error message to make this clear, and
3241 likewise updated the man page.
3242
3243 13. The limit of 99 on the number of capturing subpatterns has been removed.
3244 The new limit is 65535, which I hope will not be a "real" limit.
3245
3246
3247 Version 3.4 22-Aug-00
3248 ---------------------
3249
3250 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3251
3252 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3253
3254
3255 Version 3.3 01-Aug-00
3256 ---------------------
3257
3258 1. If an octal character was given, but the value was greater than \377, it
3259 was not getting masked to the least significant bits, as documented. This could
3260 lead to crashes in some systems.
3261
3262 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3263 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3264
3265 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3266 These just pass their arguments on to (pcre_free)(), but they are provided
3267 because some uses of PCRE bind it to non-C systems that can call its functions,
3268 but cannot call free() or pcre_free() directly.
3269
3270 4. Add "make test" as a synonym for "make check". Corrected some comments in
3271 the Makefile.
3272
3273 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3274 Makefile.
3275
3276 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3277 command called pgrep for grepping around the active processes.
3278
3279 7. Added the beginnings of support for UTF-8 character strings.
3280
3281 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3282 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3283 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3284 out for the ar command.)
3285
3286
3287 Version 3.2 12-May-00
3288 ---------------------
3289
3290 This is purely a bug fixing release.
3291
3292 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3293 of ZA. This was just one example of several cases that could provoke this bug,
3294 which was introduced by change 9 of version 2.00. The code for breaking
3295 infinite loops after an iteration that matches an empty string was't working
3296 correctly.
3297
3298 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3299 when matched against abbab (for example). After matching an empty string, it
3300 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3301 caused it to match further down the string than it should.
3302
3303 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3304 was there because it doesn't seem to be needed, and it causes trouble on some
3305 systems, as it is not a Standard C header. It has been removed.
3306
3307 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3308 were reported on the Macintosh. The changes were from
3309
3310 while ((c = *(++ptr)) != 0 && c != '\n');
3311 to
3312 while ((c = *(++ptr)) != 0 && c != '\n') ;
3313
3314 Totally extraordinary, but if that's what it takes...
3315
3316 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3317 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3318 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3319 assumes the way PCRE uses memmove() (always moving upwards).
3320
3321 6. PCRE is being used in one environment where strchr() is not available. There
3322 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3323 faster code anyway.
3324
3325
3326 Version 3.1 09-Feb-00
3327 ---------------------
3328
3329 The only change in this release is the fixing of some bugs in Makefile.in for
3330 the "install" target:
3331
3332 (1) It was failing to install pcreposix.h.
3333
3334 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3335
3336
3337 Version 3.0 01-Feb-00
3338 ---------------------
3339
3340 1. Add support for the /+ modifier to perltest (to output $` like it does in
3341 pcretest).
3342
3343 2. Add support for the /g modifier to perltest.
3344
3345 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3346 matches null strings.
3347
3348 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3349 pattern. Perl treats empty patterns specially - it reuses the most recent
3350 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3351 effect.
3352
3353 5. The POSIX interface was broken in that it was just handing over the POSIX
3354 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3355 required a bigger vector, with some working space on the end. This means that
3356 the POSIX wrapper now has to get and free some memory, and copy the results.
3357
3358 6. Added some simple autoconf support, placing the test data and the
3359 documentation in separate directories, re-organizing some of the
3360 information files, and making it build pcre-config (a GNU standard). Also added
3361 libtool support for building PCRE as a shared library, which is now the
3362 default.
3363
3364 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3365 09 are not valid octal constants. Single digits will be used for minor values
3366 less than 10.
3367
3368 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3369 existing programs that set these in the POSIX interface can use PCRE without
3370 modification.
3371
3372 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3373 return all that pcre_info() returns, plus additional data. The pcre_info()
3374 function is retained for compatibility, but is considered to be obsolete.
3375
3376 10. Added experimental recursion feature (?R) to handle one common case that
3377 Perl 5.6 will be able to do with (?p{...}).
3378
3379 11. Added support for POSIX character classes like [:alpha:], which Perl is
3380 adopting.
3381
3382
3383 Version 2.08 31-Aug-99
3384 ----------------------
3385
3386 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3387 trying to match at the startoffset position, but instead was moving forward to
3388 the next newline as if a previous match had failed.
3389
3390 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3391 and could get into a loop if a null string was matched other than at the start
3392 of the subject.
3393
3394 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3395 be distinguished at compile time, and for completeness also added PCRE_DATE.
3396
3397 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3398 in GnuWin32 environments.
3399
3400
3401 Version 2.07 29-Jul-99
3402 ----------------------
3403
3404 1. The documentation is now supplied in plain text form and HTML as well as in
3405 the form of man page sources.
3406
3407 2. C++ compilers don't like assigning (void *) values to other pointer types.
3408 In particular this affects malloc(). Although there is no problem in Standard
3409 C, I've put in casts to keep C++ compilers happy.
3410
3411 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3412 should be (const char *).
3413
3414 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3415 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3416 However, I haven't made this a standard facility. The documentation doesn't
3417 mention it, and the Makefile doesn't support it.
3418
3419 5. The Makefile now contains an "install" target, with editable destinations at
3420 the top of the file. The pcretest program is not installed.
3421
3422 6. pgrep -V now gives the PCRE version number and date.
3423
3424 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3425 causing the entire string to be ignored, instead of just the last character.
3426
3427 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3428 non-matching string, it can take a very, very long time, even for strings of
3429 quite modest length, because of the nested recursion. PCRE now does better in
3430 some of these cases. It does this by remembering the last required literal
3431 character in the pattern, and pre-searching the subject to ensure it is present
3432 before running the real match. In other words, it applies a heuristic to detect
3433 some types of certain failure quickly, and in the above example, if presented
3434 with a string that has no trailing " it gives "no match" very quickly.
3435
3436 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3437 other alternatives are tried instead.
3438
3439
3440 Version 2.06 09-Jun-99
3441 ----------------------
3442
3443 1. Change pcretest's output for amount of store used to show just the code
3444 space, because the remainder (the data block) varies in size between 32-bit and
3445 64-bit systems.
3446
3447 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3448 start matching at. This allows lookbehinds to work when searching for multiple
3449 occurrences in a string.
3450
3451 3. Added additional options to pcretest for testing multiple occurrences:
3452
3453 /+ outputs the rest of the string that follows a match
3454 /g loops for multiple occurrences, using the new startoffset argument
3455 /G loops for multiple occurrences by passing an incremented pointer
3456
3457 4. PCRE wasn't doing the "first character" optimization for patterns starting
3458 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3459 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3460 the letter 'x'. On long subject strings, this gives a significant speed-up.
3461
3462
3463 Version 2.05 21-Apr-99
3464 ----------------------
3465
3466 1. Changed the type of magic_number from int to long int so that it works
3467 properly on 16-bit systems.
3468
3469 2. Fixed a bug which caused patterns starting with .* not to work correctly
3470 when the subject string contained newline characters. PCRE was assuming
3471 anchoring for such patterns in all cases, which is not correct because .* will
3472 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3473 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3474 must be retried after every newline in the subject.
3475
3476
3477 Version 2.04 18-Feb-99
3478 ----------------------
3479
3480 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3481 computation of the store needed to hold the pattern was incorrect (too large).
3482 If such patterns were nested a few deep, this could multiply and become a real
3483 problem.
3484
3485 2. Added /M option to pcretest to show the memory requirement of a specific
3486 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3487
3488 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3489 compiled in such a way that the backtracking after subsequent failure was
3490 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3491 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3492
3493
3494 Version 2.03 02-Feb-99
3495 ----------------------
3496
3497 1. Fixed typo and small mistake in man page.
3498
3499 2. Added 4th condition (GPL supersedes if conflict) and created separate
3500 LICENCE file containing the conditions.
3501
3502 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3503 Perl, that is the internal \ allows the delimiter to be included in the
3504 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3505 the final delimiter, add \ to the end of the pattern (to test the error).
3506
3507 4. Added the convenience functions for extracting substrings after a successful
3508 match. Updated pcretest to make it able to test these functions.
3509
3510
3511 Version 2.02 14-Jan-99
3512 ----------------------
3513
3514 1. Initialized the working variables associated with each extraction so that
3515 their saving and restoring doesn't refer to uninitialized store.
3516
3517 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3518 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3519 fix the problem.
3520
3521 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3522 calls, and wasn't printing the correct value for compiling calls. Increased the
3523 default value of LOOPREPEAT, and the number of significant figures in the
3524 times.
3525
3526 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3527
3528 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3529 a building problem on Windows NT with a FAT file system.
3530
3531
3532 Version 2.01 21-Oct-98
3533 ----------------------
3534
3535 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3536 to character tables built by pcre_maketables() in the current locale. If NULL
3537 is passed, the default tables are used.
3538
3539
3540 Version 2.00 24-Sep-98
3541 ----------------------
3542
3543 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3544 it any more.
3545
3546 2. Allow quantification of (?>) groups, and make it work correctly.
3547
3548 3. The first character computation wasn't working for (?>) groups.
3549
3550 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3551 end of the subject) and add 5.005's \z, which really does match only at the
3552 very end of the subject.
3553
3554 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3555
3556 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3557 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3558 localized options. All options to pcre_study() were also removed.
3559
3560 7. Add other new features from 5.005:
3561
3562 $(?<= positive lookbehind
3563 $(?<! negative lookbehind
3564 (?imsx-imsx) added the unsetting capability
3565 such a setting is global if at outer level; local otherwise
3566 (?imsx-imsx:) non-capturing groups with option setting
3567 (?(cond)re|re) conditional pattern matching
3568
3569 A backreference to itself in a repeated group matches the previous
3570 captured string.
3571
3572 8. General tidying up of studying (both automatic and via "study")
3573 consequential on the addition of new assertions.
3574
3575 9. As in 5.005, unlimited repeated groups that could match an empty substring
3576 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3577 runtime if any iteration does actually match an empty substring.
3578
3579 10. Include the RunTest script in the distribution.
3580
3581 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3582 discrepancies, some of which were old and were also with respect to 5.004. They
3583 have now been fixed.
3584
3585
3586 Version 1.09 28-Apr-98
3587 ----------------------
3588
3589 1. A negated single character class followed by a quantifier with a minimum
3590 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3591 program crashes, or just wrong answers. This did not apply to negated classes
3592 containing more than one character, or to minima other than one.
3593
3594
3595 Version 1.08 27-Mar-98
3596 ----------------------
3597
3598 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3599
3600 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3601 latter must appear before anything that relies on it in the pattern.
3602
3603
3604 Version 1.07 16-Feb-98
3605 ----------------------
3606
3607 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3608 repeat of a potentially empty string).
3609
3610
3611 Version 1.06 23-Jan-98
3612 ----------------------
3613
3614 1. Added Markus Oberhumer's little patches for C++.
3615
3616 2. Literal strings longer than 255 characters were broken.
3617
3618
3619 Version 1.05 23-Dec-97
3620 ----------------------
3621
3622 1. Negated character classes containing more than one character were failing if
3623 PCRE_CASELESS was set at run time.
3624
3625
3626 Version 1.04 19-Dec-97
3627 ----------------------
3628
3629 1. Corrected the man page, where some "const" qualifiers had been omitted.
3630
3631 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3632 input syntax.
3633
3634 3. Fixed memory leak which occurred when a regex with back references was
3635 matched with an offsets vector that wasn't big enough. The temporary memory
3636 that is used in this case wasn't being freed if the match failed.
3637
3638 4. Tidied pcretest to ensure it frees memory that it gets.
3639
3640 5. Temporary memory was being obtained in the case where the passed offsets
3641 vector was exactly big enough.
3642
3643 6. Corrected definition of offsetof() from change 5 below.
3644
3645 7. I had screwed up change 6 below and broken the rules for the use of
3646 setjmp(). Now fixed.
3647
3648
3649 Version 1.03 18-Dec-97
3650 ----------------------
3651
3652 1. A erroneous regex with a missing opening parenthesis was correctly
3653 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3654 on some systems.
3655
3656 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3657 it was reported that one broken compiler failed on the former because "code" is
3658 also an independent variable.
3659
3660 3. The erroneous regex a[]b caused an array overrun reference.
3661
3662 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3663 fail on data ending with that character. (It was going on too far, and checking
3664 the next character, typically a binary zero.) This was specific to the
3665 optimized code for single-character negative classes.
3666
3667 5. Added a contributed patch from the TIN world which does the following:
3668
3669 + Add an undef for memmove, in case the the system defines a macro for it.
3670
3671 + Add a definition of offsetof(), in case there isn't one. (I don't know
3672 the reason behind this - offsetof() is part of the ANSI standard - but
3673 it does no harm).
3674
3675 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3676 most of the places where whitespace preceded '#'. I have given up and
3677 allowed the remaining 2 cases to be at the margin.
3678
3679 + Rename some variables in pcre to eliminate shadowing. This seems very
3680 pedantic, but does no harm, of course.
3681
3682 6. Moved the call to setjmp() into its own function, to get rid of warnings
3683 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3684
3685 7. Constructs such as \d{8,} were compiling into the equivalent of
3686 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3687 outcome, but in this particular case used more store than had been allocated,
3688 which caused the bug to be discovered because it threw up an internal error.
3689
3690 8. The debugging code in both pcre and pcretest for outputting the compiled
3691 form of a regex was going wrong in the case of back references followed by
3692 curly-bracketed repeats.
3693
3694
3695 Version 1.02 12-Dec-97
3696 ----------------------
3697
3698 1. Typos in pcre.3 and comments in the source fixed.
3699
3700 2. Applied a contributed patch to get rid of places where it used to remove
3701 'const' from variables, and fixed some signed/unsigned and uninitialized
3702 variable warnings.
3703
3704 3. Added the "runtest" target to Makefile.
3705
3706 4. Set default compiler flag to -O2 rather than just -O.
3707
3708
3709 Version 1.01 19-Nov-97
3710 ----------------------
3711
3712 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3713 like /([ab]*)*/, that is, for classes with more than one character in them.
3714
3715 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3716 as /((?>a*))*/ (a PCRE_EXTRA facility).
3717
3718
3719 Version 1.00 18-Nov-97
3720 ----------------------
3721
3722 1. Added compile-time macros to support systems such as SunOS4 which don't have
3723 memmove() or strerror() but have other things that can be used instead.
3724
3725 2. Arranged that "make clean" removes the executables.
3726
3727
3728 Version 0.99 27-Oct-97
3729 ----------------------
3730
3731 1. Fixed bug in code for optimizing classes with only one character. It was
3732 initializing a 32-byte map regardless, which could cause it to run off the end
3733 of the memory it had got.
3734
3735 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3736
3737
3738 Version 0.98 22-Oct-97
3739 ----------------------
3740
3741 1. Fixed bug in code for handling temporary memory usage when there are more
3742 back references than supplied space in the ovector. This could cause segfaults.
3743
3744
3745 Version 0.97 21-Oct-97
3746 ----------------------
3747
3748 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3749
3750 2. Optimized negated single characters not to use a bit map.
3751
3752 3. Brought error texts together as macro definitions; clarified some of them;
3753 fixed one that was wrong - it said "range out of order" when it meant "invalid
3754 escape sequence".
3755
3756 4. Changed some char * arguments to const char *.
3757
3758 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3759
3760 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3761 pcretest.
3762
3763
3764 Version 0.96 16-Oct-97
3765 ----------------------
3766
3767 1. Added a simple "pgrep" utility to the distribution.
3768
3769 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3770 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3771 where "ddd" means "one or more decimal digits".
3772
3773 3. Fixed serious bug. If a pattern had a back reference, but the call to
3774 pcre_exec() didn't supply a large enough ovector to record the related
3775 identifying subpattern, the match always failed. PCRE now remembers the number
3776 of the largest back reference, and gets some temporary memory in which to save
3777 the offsets during matching if necessary, in order to ensure that
3778 backreferences always work.
3779
3780 4. Increased the compatibility with Perl in a number of ways:
3781
3782 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3783 to request this handling. The option can be set at compile or exec time.
3784
3785 (b) $ matches before a terminating newline by default; an option
3786 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3787 mode). The option can be set at compile or exec time.
3788
3789 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3790 the same as Perl's. If the decimal number it represents is less than 10
3791 or there aren't that many previous left capturing parentheses, an octal
3792 escape is read. Inside a character class, it's always an octal escape,
3793 even if it is a single digit.
3794
3795 (d) An escaped but undefined alphabetic character is taken as a literal,
3796 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3797 escapes.
3798
3799 (e) {0} is now permitted. (The previous item is removed from the compiled
3800 pattern).
3801
3802 5. Changed all the names of code files so that the basic parts are no longer
3803 than 10 characters, and abolished the teeny "globals.c" file.
3804
3805 6. Changed the handling of character classes; they are now done with a 32-byte
3806 bit map always.
3807
3808 7. Added the -d and /D options to pcretest to make it possible to look at the
3809 internals of compilation without having to recompile pcre.
3810
3811
3812 Version 0.95 23-Sep-97
3813 ----------------------
3814
3815 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3816 \x20 at the start of a run of normal characters. These were being treated as
3817 real characters, instead of the source characters being re-checked.
3818
3819
3820 Version 0.94 18-Sep-97
3821 ----------------------
3822
3823 1. The functions are now thread-safe, with the caveat that the global variables
3824 containing pointers to malloc() and free() or alternative functions are the
3825 same for all threads.
3826
3827 2. Get pcre_study() to generate a bitmap of initial characters for non-
3828 anchored patterns when this is possible, and use it if passed to pcre_exec().
3829
3830
3831 Version 0.93 15-Sep-97
3832 ----------------------
3833
3834 1. /(b)|(:+)/ was computing an incorrect first character.
3835
3836 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3837 but not actually doing anything yet.
3838
3839 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3840 as Perl does (e.g. [-az] or [az-]).
3841
3842 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3843 all possible positions.
3844
3845 5. Split up into different modules to avoid including unneeded functions in a
3846 compiled binary. However, compile and exec are still in one module. The "study"
3847 function is split off.
3848
3849 6. The character tables are now in a separate module whose source is generated
3850 by an auxiliary program - but can then be edited by hand if required. There are
3851 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3852 toupper() in the code.
3853
3854 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3855 make them global. Abolish the function for setting them, as the caller can now
3856 set them directly.
3857
3858
3859 Version 0.92 11-Sep-97
3860 ----------------------
3861
3862 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3863 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3864
3865 2. Caseless matching was not working in character classes if the characters in
3866 the pattern were in upper case.
3867
3868 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3869
3870 4. Make PCRE_ANCHORED public and accept as a compile option.
3871
3872 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3873 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3874 pass them.
3875
3876 6. Give an error if bad option bits passed at compile or run time.
3877
3878 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3879 pcretest to cause it to pass that flag.
3880
3881 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3882 options, and the first character, if set.
3883
3884 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3885
3886
3887 Version 0.91 10-Sep-97
3888 ----------------------
3889
3890 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3891 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3892
3893 2. PCRE was looping on encountering an indefinitely repeated back reference to
3894 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3895 Perl does - treats the match as successful.
3896
3897 ****

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