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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 539 - (show annotations) (download)
Sun Jun 13 21:35:04 2010 UTC (4 years, 1 month ago) by ph10
File size: 188975 byte(s)
A more correct fix for the chartables bug with UTF-8 and non-std tables.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.10 03-Jun-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 (see next item). The new properties are: Xan
33 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
34
35 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
36 use Unicode properties. (*UCP) at the start of a pattern can be used to set
37 this option. Modified pcretest to add /W to test this facility. Added
38 REG_UCP to make it available via the POSIX interface.
39
40 10. Added --line-buffered to pcregrep.
41
42 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
43 studied, and the match started with a letter with a code point greater than
44 127 whose first byte was different to the first byte of the other case of
45 the letter, the other case of this starting letter was not recognized
46 (#976).
47
48 12. If a pattern that was studied started with a repeated Unicode property
49 test, for example, \p{Nd}+, there was the theoretical possibility of
50 setting up an incorrect bitmap of starting bytes, but fortunately it could
51 not have actually happened in practice until change 8 above was made (it
52 added property types that matched character-matching opcodes).
53
54 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
55 possible starting bytes for non-anchored patterns.
56
57 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
58 \R, and also a number of cases that involve Unicode properties, both
59 explicit and implicit when PCRE_UCP is set.
60
61 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
62 input, it could crash or give wrong results if characters with values
63 greater than 0xc0 were present in the subject string. (Detail: it assumed
64 UTF-8 input when processing these items.)
65
66 16. Added a lot of (int) casts to avoid compiler warnings in systems where
67 size_t is 64-bit (#991).
68
69 17. Added a check for running out of memory when PCRE is compiled with
70 --disable-stack-for-recursion (#990).
71
72 18. If the last data line in a file for pcretest does not have a newline on
73 the end, a newline was missing in the output.
74
75 19. The default pcre_chartables.c file recognizes only ASCII characters (values
76 less than 128) in its various bitmaps. However, there is a facility for
77 generating tables according to the current locale when PCRE is compiled. It
78 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
79 characters, are recognized by isspace() and therefore were getting set in
80 these tables. This caused a problem in UTF-8 mode when pcre_study() was
81 used to create a list of bytes that can start a match. For \s, it was
82 including 0x85 and 0xa0, which of course cannot start UTF-8 characters. I
83 have changed the code so that only real ASCII characters (less than 128)
84 and the correct starting bytes for UTF-8 encodings are set in this case.
85 (When PCRE_UCP is set - see 9 above - the code is different altogether.)
86
87
88 Version 8.02 19-Mar-2010
89 ------------------------
90
91 1. The Unicode data tables have been updated to Unicode 5.2.0.
92
93 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
94 configured.
95
96 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
97 original author of that file, following a query about its status.
98
99 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
100 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
101
102 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
103 quantifier applied to a forward-referencing subroutine call, could compile
104 incorrect code or give the error "internal error: previously-checked
105 referenced subpattern not found".
106
107 6. Both MS Visual Studio and Symbian OS have problems with initializing
108 variables to point to external functions. For these systems, therefore,
109 pcre_malloc etc. are now initialized to local functions that call the
110 relevant global functions.
111
112 7. There were two entries missing in the vectors called coptable and poptable
113 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
114 I've fixed the data, and added a kludgy way of testing at compile time that
115 the lengths are correct (equal to the number of opcodes).
116
117 8. Following on from 7, I added a similar kludge to check the length of the
118 eint vector in pcreposix.c.
119
120 9. Error texts for pcre_compile() are held as one long string to avoid too
121 much relocation at load time. To find a text, the string is searched,
122 counting zeros. There was no check for running off the end of the string,
123 which could happen if a new error number was added without updating the
124 string.
125
126 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
127
128 11. \K was not working if it appeared in an atomic group or in a group that
129 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
130 \K is "not well defined" if used in an assertion. PCRE now accepts it if
131 the assertion is positive, but not if it is negative.
132
133 12. Change 11 fortuitously reduced the size of the stack frame used in the
134 "match()" function of pcre_exec.c by one pointer. Forthcoming
135 implementation of support for (*MARK) will need an extra pointer on the
136 stack; I have reserved it now, so that the stack frame size does not
137 decrease.
138
139 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
140 item in branch that calls a recursion is a subroutine call - as in the
141 second branch in the above example - was incorrectly given the compile-
142 time error "recursive call could loop indefinitely" because pcre_compile()
143 was not correctly checking the subroutine for matching a non-empty string.
144
145 14. The checks for overrunning compiling workspace could trigger after an
146 overrun had occurred. This is a "should never occur" error, but it can be
147 triggered by pathological patterns such as hundreds of nested parentheses.
148 The checks now trigger 100 bytes before the end of the workspace.
149
150 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
151
152
153 Version 8.01 19-Jan-2010
154 ------------------------
155
156 1. If a pattern contained a conditional subpattern with only one branch (in
157 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
158 computed the wrong minimum data length (which is of course zero for such
159 subpatterns). This could cause incorrect "no match" results.
160
161 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
162 the pattern is reset in the first branch, pcre_compile() failed with
163 "internal error: code overflow at offset...". This happened only when
164 the reset was to the original external option setting. (An optimization
165 abstracts leading options settings into an external setting, which was the
166 cause of this.)
167
168 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
169 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
170 assertion pattern did not match (meaning that the assertion was true), it
171 was incorrectly treated as false if the SKIP had been reached during the
172 matching. This also applied to assertions used as conditions.
173
174 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
175 assertion subpattern, including such a pattern used as a condition,
176 unpredictable results occurred, instead of the error return
177 PCRE_ERROR_DFA_UITEM.
178
179 5. The C++ GlobalReplace function was not working like Perl for the special
180 situation when an empty string is matched. It now does the fancy magic
181 stuff that is necessary.
182
183 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
184 removed. (These were left over from very, very early versions of PCRE.)
185
186 7. Some cosmetic changes to the code to make life easier when compiling it
187 as part of something else:
188
189 (a) Change DEBUG to PCRE_DEBUG.
190
191 (b) In pcre_compile(), rename the member of the "branch_chain" structure
192 called "current" as "current_branch", to prevent a collision with the
193 Linux macro when compiled as a kernel module.
194
195 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
196 prevent a collision with the Linux macro when compiled as a kernel
197 module.
198
199 8. In pcre_compile() there are some checks for integer overflows that used to
200 cast potentially large values to (double). This has been changed to that
201 when building, a check for int64_t is made, and if it is found, it is used
202 instead, thus avoiding the use of floating point arithmetic. (There is no
203 other use of FP in PCRE.) If int64_t is not found, the fallback is to
204 double.
205
206 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
207 2005 (difference between two addresses compared to an unsigned value).
208
209 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
210 custom one, because of the following reported problem in Windows:
211
212 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
213 under Win32.
214 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
215 therefore missing the function definition.
216 - The compiler thus generates a "C" signature for the test function.
217 - The linker fails to find the "C" function.
218 - PCRE fails to configure if asked to do so against libbz2.
219
220 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
221 messages were output:
222
223 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
224 rerunning libtoolize, to keep the correct libtool macros in-tree.
225 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
226
227 I have done both of these things.
228
229 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
230 most of the time, it *can* run out if it is given a pattern that contains a
231 runaway infinite recursion. I updated the discussion in the pcrestack man
232 page.
233
234 13. Now that we have gone to the x.xx style of version numbers, the minor
235 version may start with zero. Using 08 or 09 is a bad idea because users
236 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
237 interpreted as invalid octal numbers. I've updated the previous comment in
238 configure.ac, and also added a check that gives an error if 08 or 09 are
239 used.
240
241 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
242 causing partial matching to fail when the end of the subject matched \W
243 in a UTF-8 pattern where \W was quantified with a minimum of 3.
244
245 15. There were some discrepancies between the declarations in pcre_internal.h
246 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
247 their definitions. The declarations used "const uschar *" and the
248 definitions used USPTR. Even though USPTR is normally defined as "const
249 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
250 reported that: "This difference in casting confuses some C++ compilers, for
251 example, SunCC recognizes above declarations as different functions and
252 generates broken code for hbpcre." I have changed the declarations to use
253 USPTR.
254
255 16. GNU libtool is named differently on some systems. The autogen.sh script now
256 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
257 (FreeBSD).
258
259 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
260 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
261 comment: "Figure out how to create a longlong from a string: strtoll and
262 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
263 instance, but it only takes 2 args instead of 3!"
264
265 18. A subtle bug concerned with back references has been fixed by a change of
266 specification, with a corresponding code fix. A pattern such as
267 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
268 refers, was giving matches when it shouldn't. For example, xa=xaaa would
269 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
270 same bug. Such groups have to be quantified to be useful, or contained
271 inside another quantified group. (If there's no repetition, the reference
272 can never match.) The problem arises because, having left the group and
273 moved on to the rest of the pattern, a later failure that backtracks into
274 the group uses the captured value from the final iteration of the group
275 rather than the correct earlier one. I have fixed this in PCRE by forcing
276 any group that contains a reference to itself to be an atomic group; that
277 is, there cannot be any backtracking into it once it has completed. This is
278 similar to recursive and subroutine calls.
279
280
281 Version 8.00 19-Oct-09
282 ----------------------
283
284 1. The table for translating pcre_compile() error codes into POSIX error codes
285 was out-of-date, and there was no check on the pcre_compile() error code
286 being within the table. This could lead to an OK return being given in
287 error.
288
289 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
290 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
291 in a Windows environment.
292
293 3. The pcregrep --count option prints the count for each file even when it is
294 zero, as does GNU grep. However, pcregrep was also printing all files when
295 --files-with-matches was added. Now, when both options are given, it prints
296 counts only for those files that have at least one match. (GNU grep just
297 prints the file name in this circumstance, but including the count seems
298 more useful - otherwise, why use --count?) Also ensured that the
299 combination -clh just lists non-zero counts, with no names.
300
301 4. The long form of the pcregrep -F option was incorrectly implemented as
302 --fixed_strings instead of --fixed-strings. This is an incompatible change,
303 but it seems right to fix it, and I didn't think it was worth preserving
304 the old behaviour.
305
306 5. The command line items --regex=pattern and --regexp=pattern were not
307 recognized by pcregrep, which required --regex pattern or --regexp pattern
308 (with a space rather than an '='). The man page documented the '=' forms,
309 which are compatible with GNU grep; these now work.
310
311 6. No libpcreposix.pc file was created for pkg-config; there was just
312 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
313
314 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
315 when UCP support is not needed, by modifying the Python script that
316 generates it from Unicode data files. This should not matter if the module
317 is correctly used as a library, but I received one complaint about 50K of
318 unwanted data. My guess is that the person linked everything into his
319 program rather than using a library. Anyway, it does no harm.
320
321 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
322 was a minimum greater than 1 for a wide character in a possessive
323 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
324 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
325 character. Chaos in the form of incorrect output or a compiling loop could
326 result.
327
328 9. The restrictions on what a pattern can contain when partial matching is
329 requested for pcre_exec() have been removed. All patterns can now be
330 partially matched by this function. In addition, if there are at least two
331 slots in the offset vector, the offset of the earliest inspected character
332 for the match and the offset of the end of the subject are set in them when
333 PCRE_ERROR_PARTIAL is returned.
334
335 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
336 synonymous with PCRE_PARTIAL, for backwards compatibility, and
337 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
338 and may be more useful for multi-segment matching.
339
340 11. Partial matching with pcre_exec() is now more intuitive. A partial match
341 used to be given if ever the end of the subject was reached; now it is
342 given only if matching could not proceed because another character was
343 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
344 string "Z", which now yields "no match" instead of "partial match". In the
345 case of pcre_dfa_exec(), "no match" is given if every matching path for the
346 final character ended with (*FAIL).
347
348 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
349 if the pattern had a "must contain" character that was already found in the
350 earlier partial match, unless partial matching was again requested. For
351 example, with the pattern /dog.(body)?/, the "must contain" character is
352 "g". If the first part-match was for the string "dog", restarting with
353 "sbody" failed. This bug has been fixed.
354
355 13. The string returned by pcre_dfa_exec() after a partial match has been
356 changed so that it starts at the first inspected character rather than the
357 first character of the match. This makes a difference only if the pattern
358 starts with a lookbehind assertion or \b or \B (\K is not supported by
359 pcre_dfa_exec()). It's an incompatible change, but it makes the two
360 matching functions compatible, and I think it's the right thing to do.
361
362 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
363 so that the demonstration program is easily available in environments where
364 PCRE has not been installed from source.
365
366 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
367 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
368 library.
369
370 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
371 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
372 is not the first non-POSIX option to be added. Clearly some people find
373 these options useful.
374
375 17. If a caller to the POSIX matching function regexec() passes a non-zero
376 value for nmatch with a NULL value for pmatch, the value of
377 nmatch is forced to zero.
378
379 18. RunGrepTest did not have a test for the availability of the -u option of
380 the diff command, as RunTest does. It now checks in the same way as
381 RunTest, and also checks for the -b option.
382
383 19. If an odd number of negated classes containing just a single character
384 interposed, within parentheses, between a forward reference to a named
385 subpattern and the definition of the subpattern, compilation crashed with
386 an internal error, complaining that it could not find the referenced
387 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
388 [The bug was that it was starting one character too far in when skipping
389 over the character class, thus treating the ] as data rather than
390 terminating the class. This meant it could skip too much.]
391
392 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
393 /g option in pcretest when the pattern contains \K, which makes it possible
394 to have an empty string match not at the start, even when the pattern is
395 anchored. Updated pcretest and pcredemo to use this option.
396
397 21. If the maximum number of capturing subpatterns in a recursion was greater
398 than the maximum at the outer level, the higher number was returned, but
399 with unset values at the outer level. The correct (outer level) value is
400 now given.
401
402 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
403 PCRE did not set those parentheses (unlike Perl). I have now found a way to
404 make it do so. The string so far is captured, making this feature
405 compatible with Perl.
406
407 23. The tests have been re-organized, adding tests 11 and 12, to make it
408 possible to check the Perl 5.10 features against Perl 5.10.
409
410 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
411 pattern matches a fixed length string. PCRE did not allow this; now it
412 does. Neither allows recursion.
413
414 25. I finally figured out how to implement a request to provide the minimum
415 length of subject string that was needed in order to match a given pattern.
416 (It was back references and recursion that I had previously got hung up
417 on.) This code has now been added to pcre_study(); it finds a lower bound
418 to the length of subject needed. It is not necessarily the greatest lower
419 bound, but using it to avoid searching strings that are too short does give
420 some useful speed-ups. The value is available to calling programs via
421 pcre_fullinfo().
422
423 26. While implementing 25, I discovered to my embarrassment that pcretest had
424 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
425 study optimizations had never been tested with that matching function.
426 Oops. What is worse, even when it was passed study data, there was a bug in
427 pcre_dfa_exec() that meant it never actually used it. Double oops. There
428 were also very few tests of studied patterns with pcre_dfa_exec().
429
430 27. If (?| is used to create subpatterns with duplicate numbers, they are now
431 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
432 on the other side of the coin, they are no longer allowed to have different
433 names, because these cannot be distinguished in PCRE, and this has caused
434 confusion. (This is a difference from Perl.)
435
436 28. When duplicate subpattern names are present (necessarily with different
437 numbers, as required by 27 above), and a test is made by name in a
438 conditional pattern, either for a subpattern having been matched, or for
439 recursion in such a pattern, all the associated numbered subpatterns are
440 tested, and the overall condition is true if the condition is true for any
441 one of them. This is the way Perl works, and is also more like the way
442 testing by number works.
443
444
445 Version 7.9 11-Apr-09
446 ---------------------
447
448 1. When building with support for bzlib/zlib (pcregrep) and/or readline
449 (pcretest), all targets were linked against these libraries. This included
450 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
451 libraries. This caused unwanted dependencies to be created. This problem
452 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
453 pcretest is linked with readline.
454
455 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
456 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
457 moved outside it again, because FALSE and TRUE are already defined in AIX,
458 but BOOL is not.
459
460 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
461 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
462
463 4. The pcregrep documentation said spaces were inserted as well as colons (or
464 hyphens) following file names and line numbers when outputting matching
465 lines. This is not true; no spaces are inserted. I have also clarified the
466 wording for the --colour (or --color) option.
467
468 5. In pcregrep, when --colour was used with -o, the list of matching strings
469 was not coloured; this is different to GNU grep, so I have changed it to be
470 the same.
471
472 6. When --colo(u)r was used in pcregrep, only the first matching substring in
473 each matching line was coloured. Now it goes on to look for further matches
474 of any of the test patterns, which is the same behaviour as GNU grep.
475
476 7. A pattern that could match an empty string could cause pcregrep to loop; it
477 doesn't make sense to accept an empty string match in pcregrep, so I have
478 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
479 seems to be how GNU grep behaves.
480
481 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
482 start or after a newline", because the conditional assertion was not being
483 correctly handled. The rule now is that both the assertion and what follows
484 in the first alternative must satisfy the test.
485
486 9. If auto-callout was enabled in a pattern with a conditional group whose
487 condition was an assertion, PCRE could crash during matching, both with
488 pcre_exec() and pcre_dfa_exec().
489
490 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
491 used for matching.
492
493 11. Unicode property support in character classes was not working for
494 characters (bytes) greater than 127 when not in UTF-8 mode.
495
496 12. Added the -M command line option to pcretest.
497
498 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
499
500 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
501
502 16. Added comments and documentation about mis-use of no_arg in the C++
503 wrapper.
504
505 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
506 from Martin Jerabek that uses macro names for all relevant character and
507 string constants.
508
509 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
510 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
511 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
512 these, but not everybody uses configure.
513
514 19. A conditional group that had only one branch was not being correctly
515 recognized as an item that could match an empty string. This meant that an
516 enclosing group might also not be so recognized, causing infinite looping
517 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
518 with the subject "ab", where knowledge that the repeated group can match
519 nothing is needed in order to break the loop.
520
521 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
522 exec(), but without supplying a callout function, matching went wrong.
523
524 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
525 leak if the size of the offset vector was greater than 30. When the vector
526 is smaller, the saved offsets during recursion go onto a local stack
527 vector, but for larger vectors malloc() is used. It was failing to free
528 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
529 error, in fact).
530
531 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
532 heapframe that is used only when UTF-8 support is enabled. This caused no
533 problem, but was untidy.
534
535 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
536 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
537 included within another project.
538
539 24. Steven Van Ingelgem's patches to add more options to the CMake support,
540 slightly modified by me:
541
542 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
543 not building pcregrep.
544
545 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
546 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
547
548 25. Forward references, both numeric and by name, in patterns that made use of
549 duplicate group numbers, could behave incorrectly or give incorrect errors,
550 because when scanning forward to find the reference group, PCRE was not
551 taking into account the duplicate group numbers. A pattern such as
552 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
553
554 26. Changed a few more instances of "const unsigned char *" to USPTR, making
555 the feature of a custom pointer more persuasive (as requested by a user).
556
557 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
558 pcretest.c, inside #ifndefs, because it seems they are sometimes already
559 pre-defined.
560
561 28. Added support for (*UTF8) at the start of a pattern.
562
563 29. Arrange for flags added by the "release type" setting in CMake to be shown
564 in the configuration summary.
565
566
567 Version 7.8 05-Sep-08
568 ---------------------
569
570 1. Replaced UCP searching code with optimized version as implemented for Ad
571 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
572 stage table and inline lookup instead of a function, giving speed ups of 2
573 to 5 times on some simple patterns that I tested. Permission was given to
574 distribute the MultiStage2.py script that generates the tables (it's not in
575 the tarball, but is in the Subversion repository).
576
577 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
578 scripts.
579
580 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
581 a group with a zero qualifier. The result of the study could be incorrect,
582 or the function might crash, depending on the pattern.
583
584 4. Caseless matching was not working for non-ASCII characters in back
585 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
586 It now works when Unicode Property Support is available.
587
588 5. In pcretest, an escape such as \x{de} in the data was always generating
589 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
590 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
591 truncation.
592
593 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
594
595 7. Added two (int) casts to pcregrep when printing the difference of two
596 pointers, in case they are 64-bit values.
597
598 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
599 test 2 if it fails.
600
601 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
602 and a #define of that name to empty if it is not externally set. This is to
603 allow users of MSVC to set it if necessary.
604
605 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
606 the convenience functions in the pcre_get.c source file.
607
608 11. An option change at the start of a pattern that had top-level alternatives
609 could cause overwriting and/or a crash. This command provoked a crash in
610 some environments:
611
612 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
613
614 This potential security problem was recorded as CVE-2008-2371.
615
616 12. For a pattern where the match had to start at the beginning or immediately
617 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
618 pcre_dfa_exec() could read past the end of the passed subject if there was
619 no match. To help with detecting such bugs (e.g. with valgrind), I modified
620 pcretest so that it places the subject at the end of its malloc-ed buffer.
621
622 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
623 exec() might read past the end of the data buffer in UTF-8 mode.
624
625 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
626 the data contained the byte 0x85 as part of a UTF-8 character within its
627 first line. This applied both to normal and DFA matching.
628
629 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
630 /^[^d]*?$/8 failed to match "abc".
631
632 16. Added a missing copyright notice to pcrecpp_internal.h.
633
634 17. Make it more clear in the documentation that values returned from
635 pcre_exec() in ovector are byte offsets, not character counts.
636
637 18. Tidied a few places to stop certain compilers from issuing warnings.
638
639 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
640 supplied by Stefan Weber. I made a further small update for 7.8 because
641 there is a change of source arrangements: the pcre_searchfuncs.c module is
642 replaced by pcre_ucd.c.
643
644
645 Version 7.7 07-May-08
646 ---------------------
647
648 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
649 a string to a long long, pretend we don't even have a long long." This is
650 done by checking for the strtoq, strtoll, and _strtoi64 functions.
651
652 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
653 pre-7.6 versions, which defined a global no_arg variable instead of putting
654 it in the RE class. (See also #8 below.)
655
656 3. Remove a line of dead code, identified by coverity and reported by Nuno
657 Lopes.
658
659 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
660
661 (1) The include/exclude patterns were being applied to the whole pathnames
662 of files, instead of just to the final components.
663
664 (2) If there was more than one level of directory, the subdirectories were
665 skipped unless they satisfied the include/exclude conditions. This is
666 inconsistent with GNU grep (and could even be seen as contrary to the
667 pcregrep specification - which I improved to make it absolutely clear).
668 The action now is always to scan all levels of directory, and just
669 apply the include/exclude patterns to regular files.
670
671 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
672 --exclude_dir in the tests to avoid scanning .svn directories.
673
674 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
675 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
676 doesn't support NULs in patterns.
677
678 7. Added some missing "const"s to declarations of static tables in
679 pcre_compile.c and pcre_dfa_exec.c.
680
681 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
682 caused by fix #2 above. (Subsequently also a second patch to fix the
683 first patch. And a third patch - this was a messy problem.)
684
685 9. Applied Craig's patch to remove the use of push_back().
686
687 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
688 matching function regexec().
689
690 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
691 which, however, unlike Perl's \g{...}, are subroutine calls, not back
692 references. PCRE supports relative numbers with this syntax (I don't think
693 Oniguruma does).
694
695 12. Previously, a group with a zero repeat such as (...){0} was completely
696 omitted from the compiled regex. However, this means that if the group
697 was called as a subroutine from elsewhere in the pattern, things went wrong
698 (an internal error was given). Such groups are now left in the compiled
699 pattern, with a new opcode that causes them to be skipped at execution
700 time.
701
702 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
703 to the way PCRE behaves:
704
705 (a) A lone ] character is dis-allowed (Perl treats it as data).
706
707 (b) A back reference to an unmatched subpattern matches an empty string
708 (Perl fails the current match path).
709
710 (c) A data ] in a character class must be notated as \] because if the
711 first data character in a class is ], it defines an empty class. (In
712 Perl it is not possible to have an empty class.) The empty class []
713 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
714 The negative empty class [^] matches any one character, independently
715 of the DOTALL setting.
716
717 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
718 non-existent subpattern following a character class starting with ']' and
719 containing () gave an internal compiling error instead of "reference to
720 non-existent subpattern". Fortunately, when the pattern did exist, the
721 compiled code was correct. (When scanning forwards to check for the
722 existencd of the subpattern, it was treating the data ']' as terminating
723 the class, so got the count wrong. When actually compiling, the reference
724 was subsequently set up correctly.)
725
726 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
727 it was being rejected as not supported by pcre_dfa_exec(), even though
728 other assertions are supported. I have made pcre_dfa_exec() support
729 (*FAIL).
730
731 16. The implementation of 13c above involved the invention of a new opcode,
732 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
733 cannot be changed at match time, I realized I could make a small
734 improvement to matching performance by compiling OP_ALLANY instead of
735 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
736 on the OP_ANY path.
737
738 17. Compiling pcretest on Windows with readline support failed without the
739 following two fixes: (1) Make the unistd.h include conditional on
740 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
741
742 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
743 ncurses library to be included for pcretest when ReadLine support is
744 requested, but also to allow for it to be overridden. This patch came from
745 Daniel Bergström.
746
747 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
748 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
749 any errors with the current Unicode tables. Thanks to Peter Kankowski for
750 spotting this.
751
752
753 Version 7.6 28-Jan-08
754 ---------------------
755
756 1. A character class containing a very large number of characters with
757 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
758 overflow.
759
760 2. Patch to cut out the "long long" test in pcrecpp_unittest when
761 HAVE_LONG_LONG is not defined.
762
763 3. Applied Christian Ehrlicher's patch to update the CMake build files to
764 bring them up to date and include new features. This patch includes:
765
766 - Fixed PH's badly added libz and libbz2 support.
767 - Fixed a problem with static linking.
768 - Added pcredemo. [But later removed - see 7 below.]
769 - Fixed dftables problem and added an option.
770 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
771 HAVE_LONG_LONG.
772 - Added readline support for pcretest.
773 - Added an listing of the option settings after cmake has run.
774
775 4. A user submitted a patch to Makefile that makes it easy to create
776 "pcre.dll" under mingw when using Configure/Make. I added stuff to
777 Makefile.am that cause it to include this special target, without
778 affecting anything else. Note that the same mingw target plus all
779 the other distribution libraries and programs are now supported
780 when configuring with CMake (see 6 below) instead of with
781 Configure/Make.
782
783 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
784 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
785 exported in the Windows port". It has not yet been confirmed that the patch
786 solves the problem, but it does no harm.
787
788 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
789 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
790 with CMake, and also correct the comment about stack recursion.
791
792 7. Remove the automatic building of pcredemo from the ./configure system and
793 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
794 of a program that users should build themselves after PCRE is installed, so
795 building it automatically is not really right. What is more, it gave
796 trouble in some build environments.
797
798 8. Further tidies to CMakeLists.txt from Sheri and Christian.
799
800
801 Version 7.5 10-Jan-08
802 ---------------------
803
804 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
805 values in parens when parsing an RE using the C++ wrapper."
806
807 2. Negative specials like \S did not work in character classes in UTF-8 mode.
808 Characters greater than 255 were excluded from the class instead of being
809 included.
810
811 3. The same bug as (2) above applied to negated POSIX classes such as
812 [:^space:].
813
814 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
815 defined or documented. It seems to have been a typo for PCRE_STATIC, so
816 I have changed it.
817
818 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
819 first named subpattern) and a construct such as (?&a) would reference the
820 first named subpattern whose name started with "a" (in other words, the
821 length check was missing). Both these problems are fixed. "Subpattern name
822 expected" is now given for (?&) (a zero-length name), and this patch also
823 makes it give the same error for \k'' (previously it complained that that
824 was a reference to a non-existent subpattern).
825
826 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
827 this is right because (?- can be followed by option settings as well as by
828 digits. I have, however, made the messages clearer.
829
830 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
831 than the number used in the conditional) now cause a compile-time error.
832 This is actually not compatible with Perl, which accepts such patterns, but
833 treats the conditional as always being FALSE (as PCRE used to), but it
834 seems to me that giving a diagnostic is better.
835
836 8. Change "alphameric" to the more common word "alphanumeric" in comments
837 and messages.
838
839 9. Fix two occurrences of "backslash" in comments that should have been
840 "backspace".
841
842 10. Remove two redundant lines of code that can never be obeyed (their function
843 was moved elsewhere).
844
845 11. The program that makes PCRE's Unicode character property table had a bug
846 which caused it to generate incorrect table entries for sequences of
847 characters that have the same character type, but are in different scripts.
848 It amalgamated them into a single range, with the script of the first of
849 them. In other words, some characters were in the wrong script. There were
850 thirteen such cases, affecting characters in the following ranges:
851
852 U+002b0 - U+002c1
853 U+0060c - U+0060d
854 U+0061e - U+00612
855 U+0064b - U+0065e
856 U+0074d - U+0076d
857 U+01800 - U+01805
858 U+01d00 - U+01d77
859 U+01d9b - U+01dbf
860 U+0200b - U+0200f
861 U+030fc - U+030fe
862 U+03260 - U+0327f
863 U+0fb46 - U+0fbb1
864 U+10450 - U+1049d
865
866 12. The -o option (show only the matching part of a line) for pcregrep was not
867 compatible with GNU grep in that, if there was more than one match in a
868 line, it showed only the first of them. It now behaves in the same way as
869 GNU grep.
870
871 13. If the -o and -v options were combined for pcregrep, it printed a blank
872 line for every non-matching line. GNU grep prints nothing, and pcregrep now
873 does the same. The return code can be used to tell if there were any
874 non-matching lines.
875
876 14. Added --file-offsets and --line-offsets to pcregrep.
877
878 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
879 infinitely looping recursion. The bug was that positive lookaheads were not
880 being skipped when checking for a possible empty match (negative lookaheads
881 and both kinds of lookbehind were skipped).
882
883 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
884 inclusion of <windows.h> to before rather than after the definition of
885 INVALID_FILE_ATTRIBUTES (patch from David Byron).
886
887 17. Specifying a possessive quantifier with a specific limit for a Unicode
888 character property caused pcre_compile() to compile bad code, which led at
889 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
890 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
891 caused the error; without that there was no problem.
892
893 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
894
895 19. Added --enable-pcretest-libreadline.
896
897 20. In pcrecpp.cc, the variable 'count' was incremented twice in
898 RE::GlobalReplace(). As a result, the number of replacements returned was
899 double what it should be. I removed one of the increments, but Craig sent a
900 later patch that removed the other one (the right fix) and added unit tests
901 that check the return values (which was not done before).
902
903 21. Several CMake things:
904
905 (1) Arranged that, when cmake is used on Unix, the libraries end up with
906 the names libpcre and libpcreposix, not just pcre and pcreposix.
907
908 (2) The above change means that pcretest and pcregrep are now correctly
909 linked with the newly-built libraries, not previously installed ones.
910
911 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
912
913 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
914 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
915 UTF-8 newline character). The key issue is that the pattern starts .*;
916 this means that the match must be either at the beginning, or after a
917 newline. The bug was in the code for advancing after a failed match and
918 checking that the new position followed a newline. It was not taking
919 account of UTF-8 characters correctly.
920
921 23. PCRE was behaving differently from Perl in the way it recognized POSIX
922 character classes. PCRE was not treating the sequence [:...:] as a
923 character class unless the ... were all letters. Perl, however, seems to
924 allow any characters between [: and :], though of course it rejects as
925 unknown any "names" that contain non-letters, because all the known class
926 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
927 for example, whereas PCRE did not - it did not recognize a POSIX character
928 class. This seemed a bit dangerous, so the code has been changed to be
929 closer to Perl. The behaviour is not identical to Perl, because PCRE will
930 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
931 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
932 Perl does, and where it didn't before.
933
934 24. Rewrite so as to remove the single use of %n from pcregrep because in some
935 Windows environments %n is disabled by default.
936
937
938 Version 7.4 21-Sep-07
939 ---------------------
940
941 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
942 means that a class such as [\s] counted as "explicit reference to CR or
943 LF". That isn't really right - the whole point of the change was to try to
944 help when there was an actual mention of one of the two characters. So now
945 the change happens only if \r or \n (or a literal CR or LF) character is
946 encountered.
947
948 2. The 32-bit options word was also used for 6 internal flags, but the numbers
949 of both had grown to the point where there were only 3 bits left.
950 Fortunately, there was spare space in the data structure, and so I have
951 moved the internal flags into a new 16-bit field to free up more option
952 bits.
953
954 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
955 but did not set the internal JCHANGED flag - either of these is enough to
956 control the way the "get" function works - but the PCRE_INFO_JCHANGED
957 facility is supposed to tell if (?J) was ever used, so now (?J) at the
958 start sets both bits.
959
960 4. Added options (at build time, compile time, exec time) to change \R from
961 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
962
963 5. doc/pcresyntax.html was missing from the distribution.
964
965 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
966 compatibility, even though it is no longer used.
967
968 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
969 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
970 windows.h file is present (where different names are used). [This was
971 reversed later after testing - see 16 below.]
972
973 8. Changed all #include <config.h> to #include "config.h". There were also
974 some further <pcre.h> cases that I changed to "pcre.h".
975
976 9. When pcregrep was used with the --colour option, it missed the line ending
977 sequence off the lines that it output.
978
979 10. It was pointed out to me that arrays of string pointers cause lots of
980 relocations when a shared library is dynamically loaded. A technique of
981 using a single long string with a table of offsets can drastically reduce
982 these. I have refactored PCRE in four places to do this. The result is
983 dramatic:
984
985 Originally: 290
986 After changing UCP table: 187
987 After changing error message table: 43
988 After changing table of "verbs" 36
989 After changing table of Posix names 22
990
991 Thanks to the folks working on Gregex for glib for this insight.
992
993 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
994 unicode-properties was also set.
995
996 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
997
998 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
999 checked only for CRLF.
1000
1001 14. Added casts to pcretest.c to avoid compiler warnings.
1002
1003 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1004
1005 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1006 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1007 entirely. This removes changes made in 7 above.
1008
1009 17. The CMake files have been updated, and there is now more information about
1010 building with CMake in the NON-UNIX-USE document.
1011
1012
1013 Version 7.3 28-Aug-07
1014 ---------------------
1015
1016 1. In the rejigging of the build system that eventually resulted in 7.1, the
1017 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1018 brackets there is not right, since it causes compilers to look for an
1019 installed pcre.h, not the version that is in the source that is being
1020 compiled (which of course may be different). I have changed it back to:
1021
1022 #include "pcre.h"
1023
1024 I have a vague recollection that the change was concerned with compiling in
1025 different directories, but in the new build system, that is taken care of
1026 by the VPATH setting the Makefile.
1027
1028 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1029 when the subject happened to end in the byte 0x85 (e.g. if the last
1030 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1031 characters but of course it shouldn't be taken as a newline when it is part
1032 of another character. The bug was that, for an unlimited repeat of . in
1033 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1034 characters when looking for a newline.
1035
1036 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1037
1038 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1039 in debug output.
1040
1041 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1042 long printing in the pcrecpp unittest when running under MinGW.
1043
1044 6. ESC_K was left out of the EBCDIC table.
1045
1046 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1047 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1048 limit also applies to "virtual nesting" when a pattern is recursive, and in
1049 this case 1000 isn't so big. I have been able to remove this limit at the
1050 expense of backing off one optimization in certain circumstances. Normally,
1051 when pcre_exec() would call its internal match() function recursively and
1052 immediately return the result unconditionally, it uses a "tail recursion"
1053 feature to save stack. However, when a subpattern that can match an empty
1054 string has an unlimited repetition quantifier, it no longer makes this
1055 optimization. That gives it a stack frame in which to save the data for
1056 checking that an empty string has been matched. Previously this was taken
1057 from the 1000-entry workspace that had been reserved. So now there is no
1058 explicit limit, but more stack is used.
1059
1060 8. Applied Daniel's patches to solve problems with the import/export magic
1061 syntax that is required for Windows, and which was going wrong for the
1062 pcreposix and pcrecpp parts of the library. These were overlooked when this
1063 problem was solved for the main library.
1064
1065 9. There were some crude static tests to avoid integer overflow when computing
1066 the size of patterns that contain repeated groups with explicit upper
1067 limits. As the maximum quantifier is 65535, the maximum group length was
1068 set at 30,000 so that the product of these two numbers did not overflow a
1069 32-bit integer. However, it turns out that people want to use groups that
1070 are longer than 30,000 bytes (though not repeat them that many times).
1071 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1072 made it possible to implement the integer overflow checks in a much more
1073 dynamic way, which I have now done. The artificial limitation on group
1074 length has been removed - we now have only the limit on the total length of
1075 the compiled pattern, which depends on the LINK_SIZE setting.
1076
1077 10. Fixed a bug in the documentation for get/copy named substring when
1078 duplicate names are permitted. If none of the named substrings are set, the
1079 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1080 empty string.
1081
1082 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1083 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1084 because the ] is interpreted as the first data character and the
1085 terminating ] is not found. PCRE has been made compatible with Perl in this
1086 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1087 cause memory overwriting.
1088
1089 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1090 string has been matched (to stop an infinite loop). It was not recognizing
1091 a conditional subpattern that could match an empty string if that
1092 subpattern was within another subpattern. For example, it looped when
1093 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1094 condition was not nested. This bug has been fixed.
1095
1096 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1097 past the start of the subject in the presence of bytes with the top bit
1098 set, for example "\x8aBCD".
1099
1100 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1101 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1102
1103 14. Optimized (?!) to (*FAIL).
1104
1105 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1106 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1107 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1108 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1109 does: it's just the validity check that is more restrictive.
1110
1111 16. Inserted checks for integer overflows during escape sequence (backslash)
1112 processing, and also fixed erroneous offset values for syntax errors during
1113 backslash processing.
1114
1115 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1116 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1117
1118 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1119 caused an overrun.
1120
1121 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1122 something other than just ASCII characters) inside a group that had an
1123 unlimited repeat caused a loop at compile time (while checking to see
1124 whether the group could match an empty string).
1125
1126 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1127 [\P{Any}] did so. (Error in the code for printing property names.)
1128
1129 21. An orphan \E inside a character class could cause a crash.
1130
1131 22. A repeated capturing bracket such as (A)? could cause a wild memory
1132 reference during compilation.
1133
1134 23. There are several functions in pcre_compile() that scan along a compiled
1135 expression for various reasons (e.g. to see if it's fixed length for look
1136 behind). There were bugs in these functions when a repeated \p or \P was
1137 present in the pattern. These operators have additional parameters compared
1138 with \d, etc, and these were not being taken into account when moving along
1139 the compiled data. Specifically:
1140
1141 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1142 length.
1143
1144 (b) An item such as \pL+ within a repeated group could cause crashes or
1145 loops.
1146
1147 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1148 "reference to non-existent subpattern" error.
1149
1150 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1151
1152 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1153 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1154
1155 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1156
1157 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1158 character were causing crashes (broken optimization).
1159
1160 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1161 \p or \P) caused a compile-time loop.
1162
1163 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1164 break. For example, the unstudied pattern [\r\n]A does not match the string
1165 "\r\nA" because change 7.0/46 below moves the current point on by two
1166 characters after failing to match at the start. However, the pattern \nA
1167 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1168 the same is true. There doesn't seem any very clean way out of this, but
1169 what I have chosen to do makes the common cases work: PCRE now takes note
1170 of whether there can be an explicit match for \r or \n anywhere in the
1171 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1172 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1173 pattern has explicit CR or LF references.
1174
1175 29. Added (*CR) etc for changing newline setting at start of pattern.
1176
1177
1178 Version 7.2 19-Jun-07
1179 ---------------------
1180
1181 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1182 which is apparently normally available under Windows.
1183
1184 2. Re-jig the pcregrep tests with different newline settings in an attempt
1185 to make them independent of the local environment's newline setting.
1186
1187 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1188
1189 4. Some of the "internals" tests were previously cut out when the link size
1190 was not 2, because the output contained actual offsets. The recent new
1191 "Z" feature of pcretest means that these can be cut out, making the tests
1192 usable with all link sizes.
1193
1194 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1195 stack recursion. This gives a massive performance boost under BSD, but just
1196 a small improvement under Linux. However, it saves one field in the frame
1197 in all cases.
1198
1199 6. Added more features from the forthcoming Perl 5.10:
1200
1201 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1202 recursion call. It refers to the nth most recently opened parentheses.
1203
1204 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1205 to be opened parentheses.
1206
1207 (c) Conditions that refer to capturing parentheses can be specified
1208 relatively, for example, (?(-2)... or (?(+3)...
1209
1210 (d) \K resets the start of the current match so that everything before
1211 is not part of it.
1212
1213 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1214
1215 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1216 reference syntax.
1217
1218 (g) (?| introduces a group in which the numbering of parentheses in each
1219 alternative starts with the same number.
1220
1221 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1222
1223 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1224 PCRE_INFO_JCHANGED.
1225
1226 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1227 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1228 for detecting groups that can match an empty string.
1229
1230 9. A pattern with a very large number of alternatives (more than several
1231 hundred) was running out of internal workspace during the pre-compile
1232 phase, where pcre_compile() figures out how much memory will be needed. A
1233 bit of new cunning has reduced the workspace needed for groups with
1234 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1235 workspace instead of running out of the 4096 that are available.
1236
1237 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1238
1239 11. Applied patch from Google to remove an optimization that didn't quite work.
1240 The report of the bug said:
1241
1242 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1243 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1244 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1245
1246 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1247 it matched the wrong number of bytes.
1248
1249
1250 Version 7.1 24-Apr-07
1251 ---------------------
1252
1253 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1254 that is more "standard", making use of automake and other Autotools. There
1255 is some re-arrangement of the files and adjustment of comments consequent
1256 on this.
1257
1258 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1259 for recursive directory scanning broke on some systems because the files
1260 are not scanned in any specific order and on different systems the order
1261 was different. A call to "sort" has been inserted into RunGrepTest for the
1262 approprate test as a short-term fix. In the longer term there may be an
1263 alternative.
1264
1265 3. I had an email from Eric Raymond about problems translating some of PCRE's
1266 man pages to HTML (despite the fact that I distribute HTML pages, some
1267 people do their own conversions for various reasons). The problems
1268 concerned the use of low-level troff macros .br and .in. I have therefore
1269 removed all such uses from the man pages (some were redundant, some could
1270 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1271 HTML has been updated to handle .nf/.fi and to complain if it encounters
1272 .br or .in.
1273
1274 4. Updated comments in configure.ac that get placed in config.h.in and also
1275 arranged for config.h to be included in the distribution, with the name
1276 config.h.generic, for the benefit of those who have to compile without
1277 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1278
1279 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1280 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1281 makevp.bat for the current PCRE, using the additional files
1282 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1283
1284 6. A Windows user reported a minor discrepancy with test 2, which turned out
1285 to be caused by a trailing space on an input line that had got lost in his
1286 copy. The trailing space was an accident, so I've just removed it.
1287
1288 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1289 that is needed.
1290
1291 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1292 as "const" (a) because they are and (b) because it helps the PHP
1293 maintainers who have recently made a script to detect big data structures
1294 in the php code that should be moved to the .rodata section. I remembered
1295 to update Builducptable as well, so it won't revert if ucptable.h is ever
1296 re-created.
1297
1298 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1299 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1300 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1301 support is not required. This saves 1.5-2K of code, which is important in
1302 some applications.
1303
1304 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1305 so as not to refer to the tables, even though these functions will never be
1306 called when UTF-8 support is disabled. Otherwise there are problems with a
1307 shared library.
1308
1309 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1310
1311 (a) It was defining its arguments as char * instead of void *.
1312
1313 (b) It was assuming that all moves were upwards in memory; this was true
1314 a long time ago when I wrote it, but is no longer the case.
1315
1316 The emulated memove() is provided for those environments that have neither
1317 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1318 is clearly not the case, as these two bugs were recently reported.
1319
1320 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1321 and Detrail to create the HTML documentation, the .txt form of the man
1322 pages, and it removes trailing spaces from listed files. It also creates
1323 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1324 case, it wraps all the #defines with #ifndefs. This script should be run
1325 before "make dist".
1326
1327 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1328 with Unicode property support.
1329
1330 (a) For a maximizing quantifier, if the two different cases of the
1331 character were of different lengths in their UTF-8 codings (there are
1332 some cases like this - I found 11), and the matching function had to
1333 back up over a mixture of the two cases, it incorrectly assumed they
1334 were both the same length.
1335
1336 (b) When PCRE was configured to use the heap rather than the stack for
1337 recursion during matching, it was not correctly preserving the data for
1338 the other case of a UTF-8 character when checking ahead for a match
1339 while processing a minimizing repeat. If the check also involved
1340 matching a wide character, but failed, corruption could cause an
1341 erroneous result when trying to check for a repeat of the original
1342 character.
1343
1344 13. Some tidying changes to the testing mechanism:
1345
1346 (a) The RunTest script now detects the internal link size and whether there
1347 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1348 values substituted by "configure". (The RunGrepTest script already did
1349 this for UTF-8.) The configure.ac script no longer substitutes the
1350 relevant variables.
1351
1352 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1353 with length and offset values. This means that the output is different
1354 for different internal link sizes. Test 2 is skipped for link sizes
1355 other than 2 because of this, bypassing the problem. Unfortunately,
1356 there was also a test in test 3 (the locale tests) that used /B and
1357 failed for link sizes other than 2. Rather than cut the whole test out,
1358 I have added a new /Z option to pcretest that replaces the length and
1359 offset values with spaces. This is now used to make test 3 independent
1360 of link size. (Test 2 will be tidied up later.)
1361
1362 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1363 segmentation fault instead of returning the appropriate error message.
1364
1365 15. In multiline mode when the newline sequence was set to "any", the pattern
1366 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1367 This doesn't seem right; it now treats the CRLF combination as the line
1368 ending, and so does not match in that case. It's only a pattern such as ^$
1369 that would hit this one: something like ^ABC$ would have failed after \r
1370 and then tried again after \r\n.
1371
1372 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1373 in an attempt to make files that differ only in their line terminators
1374 compare equal. This works on Linux.
1375
1376 17. Under certain error circumstances pcregrep might try to free random memory
1377 as it exited. This is now fixed, thanks to valgrind.
1378
1379 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1380 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1381 was because its rules for how to advance for /g after matching an empty
1382 string at the end of a line did not allow for this case. They now check for
1383 it specially.
1384
1385 20. pcretest is supposed to handle patterns and data of any length, by
1386 extending its buffers when necessary. It was getting this wrong when the
1387 buffer for a data line had to be extended.
1388
1389 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1390 CRLF as a newline sequence.
1391
1392 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1393 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1394 I have nevertheless tidied it up.
1395
1396 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1397
1398 24. Added a man page for pcre-config.
1399
1400
1401 Version 7.0 19-Dec-06
1402 ---------------------
1403
1404 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1405 moving to gcc 4.1.1.
1406
1407 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1408 sys/time.h, which is documented as needed for this function. It doesn't
1409 seem to matter on Linux, but it showed up on some releases of OS X.
1410
1411 3. It seems that there are systems where bytes whose values are greater than
1412 127 match isprint() in the "C" locale. The "C" locale should be the
1413 default when a C program starts up. In most systems, only ASCII printing
1414 characters match isprint(). This difference caused the output from pcretest
1415 to vary, making some of the tests fail. I have changed pcretest so that:
1416
1417 (a) When it is outputting text in the compiled version of a pattern, bytes
1418 other than 32-126 are always shown as hex escapes.
1419
1420 (b) When it is outputting text that is a matched part of a subject string,
1421 it does the same, unless a different locale has been set for the match
1422 (using the /L modifier). In this case, it uses isprint() to decide.
1423
1424 4. Fixed a major bug that caused incorrect computation of the amount of memory
1425 required for a compiled pattern when options that changed within the
1426 pattern affected the logic of the preliminary scan that determines the
1427 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1428 that the computed length was too small. The symptoms of this bug were
1429 either the PCRE error "internal error: code overflow" from pcre_compile(),
1430 or a glibc crash with a message such as "pcretest: free(): invalid next
1431 size (fast)". Examples of patterns that provoked this bug (shown in
1432 pcretest format) are:
1433
1434 /(?-x: )/x
1435 /(?x)(?-x: \s*#\s*)/
1436 /((?i)[\x{c0}])/8
1437 /(?i:[\x{c0}])/8
1438
1439 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1440 is now done differently.
1441
1442 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1443 wrapper classes; (b) implement a new function in the C++ scanner that is
1444 more efficient than the old way of doing things because it avoids levels of
1445 recursion in the regex matching; (c) add a paragraph to the documentation
1446 for the FullMatch() function.
1447
1448 6. The escape sequence \n was being treated as whatever was defined as
1449 "newline". Not only was this contrary to the documentation, which states
1450 that \n is character 10 (hex 0A), but it also went horribly wrong when
1451 "newline" was defined as CRLF. This has been fixed.
1452
1453 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1454 was being set to -1 for the "end of line" case (supposedly a value that no
1455 character can have). Though this value is never used (the check for end of
1456 line is "zero bytes in current character"), it caused compiler complaints.
1457 I've changed it to 0xffffffff.
1458
1459 8. In pcre_version.c, the version string was being built by a sequence of
1460 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1461 string (as it is for production releases) called a macro with an empty
1462 argument. The C standard says the result of this is undefined. The gcc
1463 compiler treats it as an empty string (which was what was wanted) but it is
1464 reported that Visual C gives an error. The source has been hacked around to
1465 avoid this problem.
1466
1467 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1468 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1469 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1470 of them did).
1471
1472 10. Originally, pcretest opened its input and output without "b"; then I was
1473 told that "b" was needed in some environments, so it was added for release
1474 5.0 to both the input and output. (It makes no difference on Unix-like
1475 systems.) Later I was told that it is wrong for the input on Windows. I've
1476 now abstracted the modes into two macros, to make it easier to fiddle with
1477 them, and removed "b" from the input mode under Windows.
1478
1479 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1480
1481 12. Added -help and --help to pcretest as an official way of being reminded
1482 of the options.
1483
1484 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1485 and pcrecpp.cc because they annoy compilers at high warning levels.
1486
1487 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1488
1489 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1490 scripts are not C programs :-) and which was not noticed because it works
1491 on Linux.
1492
1493 16. pcretest is supposed to handle any length of pattern and data line (as one
1494 line or as a continued sequence of lines) by extending its input buffer if
1495 necessary. This feature was broken for very long pattern lines, leading to
1496 a string of junk being passed to pcre_compile() if the pattern was longer
1497 than about 50K.
1498
1499 17. I have done a major re-factoring of the way pcre_compile() computes the
1500 amount of memory needed for a compiled pattern. Previously, there was code
1501 that made a preliminary scan of the pattern in order to do this. That was
1502 OK when PCRE was new, but as the facilities have expanded, it has become
1503 harder and harder to keep it in step with the real compile phase, and there
1504 have been a number of bugs (see for example, 4 above). I have now found a
1505 cunning way of running the real compile function in a "fake" mode that
1506 enables it to compute how much memory it would need, while actually only
1507 ever using a few hundred bytes of working memory and without too many
1508 tests of the mode. This should make future maintenance and development
1509 easier. A side effect of this work is that the limit of 200 on the nesting
1510 depth of parentheses has been removed (though this was never a serious
1511 limitation, I suspect). However, there is a downside: pcre_compile() now
1512 runs more slowly than before (30% or more, depending on the pattern). I
1513 hope this isn't a big issue. There is no effect on runtime performance.
1514
1515 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1516 newline (only possible for the last line of a file) and it was a
1517 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1518
1519 19. Added additional timing features to pcretest. (1) The -tm option now times
1520 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1521 separate command line item, by a number that specifies the number of
1522 repeats to use when timing. The default is 50000; this gives better
1523 precision, but takes uncomfortably long for very large patterns.
1524
1525 20. Extended pcre_study() to be more clever in cases where a branch of a
1526 subpattern has no definite first character. For example, (a*|b*)[cd] would
1527 previously give no result from pcre_study(). Now it recognizes that the
1528 first character must be a, b, c, or d.
1529
1530 21. There was an incorrect error "recursive call could loop indefinitely" if
1531 a subpattern (or the entire pattern) that was being tested for matching an
1532 empty string contained only one non-empty item after a nested subpattern.
1533 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1534 incorrectly, because the \d was being skipped in the check.
1535
1536 22. The pcretest program now has a new pattern option /B and a command line
1537 option -b, which is equivalent to adding /B to every pattern. This causes
1538 it to show the compiled bytecode, without the additional information that
1539 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1540 is the same as /B/I).
1541
1542 23. A new optimization is now able automatically to treat some sequences such
1543 as a*b as a*+b. More specifically, if something simple (such as a character
1544 or a simple class like \d) has an unlimited quantifier, and is followed by
1545 something that cannot possibly match the quantified thing, the quantifier
1546 is automatically "possessified".
1547
1548 24. A recursive reference to a subpattern whose number was greater than 39
1549 went wrong under certain circumstances in UTF-8 mode. This bug could also
1550 have affected the operation of pcre_study().
1551
1552 25. Realized that a little bit of performance could be had by replacing
1553 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1554
1555 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1556
1557 27. Possessive quantifiers such as a++ were previously implemented by turning
1558 them into atomic groups such as ($>a+). Now they have their own opcodes,
1559 which improves performance. This includes the automatically created ones
1560 from 23 above.
1561
1562 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1563 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1564 the first matched character to be a colon. This applied both to named and
1565 numbered groups.
1566
1567 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1568
1569 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1570 building PCRE on Windows easier, so I have included it in the distribution.
1571
1572 31. There is now a check in pcretest against a ridiculously large number being
1573 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1574 loop, the loop is abandoned.
1575
1576 32. Forward references to subpatterns in conditions such as (?(2)...) where
1577 subpattern 2 is defined later cause pcre_compile() to search forwards in
1578 the pattern for the relevant set of parentheses. This search went wrong
1579 when there were unescaped parentheses in a character class, parentheses
1580 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1581
1582 33. "Subroutine" calls and backreferences were previously restricted to
1583 referencing subpatterns earlier in the regex. This restriction has now
1584 been removed.
1585
1586 34. Added a number of extra features that are going to be in Perl 5.10. On the
1587 whole, these are just syntactic alternatives for features that PCRE had
1588 previously implemented using the Python syntax or my own invention. The
1589 other formats are all retained for compatibility.
1590
1591 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1592 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1593 also .NET compatible.
1594
1595 (b) A recursion or subroutine call to a named group can now be defined as
1596 (?&name) as well as (?P>name).
1597
1598 (c) A backreference to a named group can now be defined as \k<name> or
1599 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1600 5.10, are also .NET compatible.
1601
1602 (d) A conditional reference to a named group can now use the syntax
1603 (?(<name>) or (?('name') as well as (?(name).
1604
1605 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1606 groups (named and numbered) that are never evaluated inline, but can be
1607 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1608 is always false. There may be only one alternative in such a group.
1609
1610 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1611 as the simple (?(R). The condition is true only if the most recent
1612 recursion is that of the given number or name. It does not search out
1613 through the entire recursion stack.
1614
1615 (g) The escape \gN or \g{N} has been added, where N is a positive or
1616 negative number, specifying an absolute or relative reference.
1617
1618 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1619 some "unreachable code" warnings.
1620
1621 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1622 things, this adds five new scripts.
1623
1624 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1625 There were also incompatibilities regarding the handling of \Q..\E inside
1626 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1627 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1628
1629 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1630 matches an empty string, and forcibly breaks the loop. There were bugs in
1631 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1632 against aaaa the result was just "a" rather than "aaaa", for example. Two
1633 separate and independent bugs (that affected different cases) have been
1634 fixed.
1635
1636 39. Refactored the code to abolish the use of different opcodes for small
1637 capturing bracket numbers. This is a tidy that I avoided doing when I
1638 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1639 The new approach is not only tidier, it makes it possible to reduce the
1640 memory needed to fix the previous bug (38).
1641
1642 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1643 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1644 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1645 mode.
1646
1647 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1648 report.
1649
1650 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1651 copy construction and assignment in the C++ wrapper.
1652
1653 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1654 couple of bugs that could have given wrong results in the "--newline=crlf"
1655 case.
1656
1657 44. Added a number of casts and did some reorganization of signed/unsigned int
1658 variables following suggestions from Dair Grant. Also renamed the variable
1659 "this" as "item" because it is a C++ keyword.
1660
1661 45. Arranged for dftables to add
1662
1663 #include "pcre_internal.h"
1664
1665 to pcre_chartables.c because without it, gcc 4.x may remove the array
1666 definition from the final binary if PCRE is built into a static library and
1667 dead code stripping is activated.
1668
1669 46. For an unanchored pattern, if a match attempt fails at the start of a
1670 newline sequence, and the newline setting is CRLF or ANY, and the next two
1671 characters are CRLF, advance by two characters instead of one.
1672
1673
1674 Version 6.7 04-Jul-06
1675 ---------------------
1676
1677 1. In order to handle tests when input lines are enormously long, pcretest has
1678 been re-factored so that it automatically extends its buffers when
1679 necessary. The code is crude, but this _is_ just a test program. The
1680 default size has been increased from 32K to 50K.
1681
1682 2. The code in pcre_study() was using the value of the re argument before
1683 testing it for NULL. (Of course, in any sensible call of the function, it
1684 won't be NULL.)
1685
1686 3. The memmove() emulation function in pcre_internal.h, which is used on
1687 systems that lack both memmove() and bcopy() - that is, hardly ever -
1688 was missing a "static" storage class specifier.
1689
1690 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1691 containing an extended class (one that cannot be represented by a bitmap
1692 because it contains high-valued characters or Unicode property items, e.g.
1693 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1694 pattern, but PCRE should not loop if you do not (it no longer does).
1695 [Detail: two cases were found: (a) a repeated subpattern containing an
1696 extended class; (b) a recursive reference to a subpattern that followed a
1697 previous extended class. It wasn't skipping over the extended class
1698 correctly when UTF-8 mode was not set.]
1699
1700 5. A negated single-character class was not being recognized as fixed-length
1701 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1702 compile error "lookbehind assertion is not fixed length".
1703
1704 6. The RunPerlTest auxiliary script was showing an unexpected difference
1705 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1706 write a Perl script that can interpret lines of an input file either as
1707 byte characters or as UTF-8, which is what "perltest" was being required to
1708 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1709 can't do is switch easily at run time between having the "use utf8;" pragma
1710 or not. In the end, I fudged it by using the RunPerlTest script to insert
1711 "use utf8;" explicitly for the UTF-8 tests.
1712
1713 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1714 the end of the subject string, contrary to the documentation and to what
1715 Perl does. This was true of both matching functions. Now it matches only at
1716 the start of the subject and immediately after *internal* newlines.
1717
1718 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1719 a pointer to an int instead of a pointer to an unsigned long int. This
1720 caused problems on 64-bit systems.
1721
1722 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1723 instance of the 'standard' template library not being so standard".
1724
1725 10. There was no check on the number of named subpatterns nor the maximum
1726 length of a subpattern name. The product of these values is used to compute
1727 the size of the memory block for a compiled pattern. By supplying a very
1728 long subpattern name and a large number of named subpatterns, the size
1729 computation could be caused to overflow. This is now prevented by limiting
1730 the length of names to 32 characters, and the number of named subpatterns
1731 to 10,000.
1732
1733 11. Subpatterns that are repeated with specific counts have to be replicated in
1734 the compiled pattern. The size of memory for this was computed from the
1735 length of the subpattern and the repeat count. The latter is limited to
1736 65535, but there was no limit on the former, meaning that integer overflow
1737 could in principle occur. The compiled length of a repeated subpattern is
1738 now limited to 30,000 bytes in order to prevent this.
1739
1740 12. Added the optional facility to have named substrings with the same name.
1741
1742 13. Added the ability to use a named substring as a condition, using the
1743 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1744 are numbers (not recommended). Forward references are permitted.
1745
1746 14. Added forward references in named backreferences (if you see what I mean).
1747
1748 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1749 pattern could run off the end of the subject. For example, the pattern
1750 "(?s)(.{1,5})"8 did this with the subject "ab".
1751
1752 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1753 PCRE_CASELESS was set when matching characters that were quantified with ?
1754 or *.
1755
1756 17. A character class other than a single negated character that had a minimum
1757 but no maximum quantifier - for example [ab]{6,} - was not handled
1758 correctly by pce_dfa_exec(). It would match only one character.
1759
1760 18. A valid (though odd) pattern that looked like a POSIX character
1761 class but used an invalid character after [ (for example [[,abc,]]) caused
1762 pcre_compile() to give the error "Failed: internal error: code overflow" or
1763 in some cases to crash with a glibc free() error. This could even happen if
1764 the pattern terminated after [[ but there just happened to be a sequence of
1765 letters, a binary zero, and a closing ] in the memory that followed.
1766
1767 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1768 over the years. Originally (before any Unicode support), just the bottom 8
1769 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1770 output from "man perlunicode" includes this:
1771
1772 The regular expression compiler produces polymorphic opcodes. That
1773 is, the pattern adapts to the data and automatically switches to
1774 the Unicode character scheme when presented with Unicode data--or
1775 instead uses a traditional byte scheme when presented with byte
1776 data.
1777
1778 Sadly, a wide octal escape does not cause a switch, and in a string with
1779 no other multibyte characters, these octal escapes are treated as before.
1780 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1781 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1782 Unicode string.
1783
1784 I have not perpetrated such confusion in PCRE. Up till now, it took just
1785 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1786 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1787 translate to the appropriate multibyte character.
1788
1789 29. Applied some refactoring to reduce the number of warnings from Microsoft
1790 and Borland compilers. This has included removing the fudge introduced
1791 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1792 a warning about an unused variable.
1793
1794 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1795 characters since release 4.0, because Perl (from release 5.004) does not.
1796 [Or at least, is documented not to: some releases seem to be in conflict
1797 with the documentation.] However, when a pattern was studied with
1798 pcre_study() and all its branches started with \s, PCRE still included VT
1799 as a possible starting character. Of course, this did no harm; it just
1800 caused an unnecessary match attempt.
1801
1802 22. Removed a now-redundant internal flag bit that recorded the fact that case
1803 dependency changed within the pattern. This was once needed for "required
1804 byte" processing, but is no longer used. This recovers a now-scarce options
1805 bit. Also moved the least significant internal flag bit to the most-
1806 significant bit of the word, which was not previously used (hangover from
1807 the days when it was an int rather than a uint) to free up another bit for
1808 the future.
1809
1810 23. Added support for CRLF line endings as well as CR and LF. As well as the
1811 default being selectable at build time, it can now be changed at runtime
1812 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1813 specify that it is scanning data with non-default line endings.
1814
1815 24. Changed the definition of CXXLINK to make it agree with the definition of
1816 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1817
1818 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1819 recursions. This makes a big different to stack usage for some patterns.
1820
1821 26. If a subpattern containing a named recursion or subroutine reference such
1822 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1823 the space required for the compiled pattern went wrong and gave too small a
1824 value. Depending on the environment, this could lead to "Failed: internal
1825 error: code overflow at offset 49" or "glibc detected double free or
1826 corruption" errors.
1827
1828 27. Applied patches from Google (a) to support the new newline modes and (b) to
1829 advance over multibyte UTF-8 characters in GlobalReplace.
1830
1831 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1832 difference for some implementation of PCRE in some Windows version.
1833
1834 29. Added some extra testing facilities to pcretest:
1835
1836 \q<number> in a data line sets the "match limit" value
1837 \Q<number> in a data line sets the "match recursion limt" value
1838 -S <number> sets the stack size, where <number> is in megabytes
1839
1840 The -S option isn't available for Windows.
1841
1842
1843 Version 6.6 06-Feb-06
1844 ---------------------
1845
1846 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1847 in pcreposix.h. I have copied the definition from pcre.h.
1848
1849 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1850 because pcre.h is no longer a built file.
1851
1852 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1853 not normally included in the compiled code.
1854
1855
1856 Version 6.5 01-Feb-06
1857 ---------------------
1858
1859 1. When using the partial match feature with pcre_dfa_exec(), it was not
1860 anchoring the second and subsequent partial matches at the new starting
1861 point. This could lead to incorrect results. For example, with the pattern
1862 /1234/, partially matching against "123" and then "a4" gave a match.
1863
1864 2. Changes to pcregrep:
1865
1866 (a) All non-match returns from pcre_exec() were being treated as failures
1867 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1868 error message is output. Some extra information is given for the
1869 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1870 probably the only errors that are likely to be caused by users (by
1871 specifying a regex that has nested indefinite repeats, for instance).
1872 If there are more than 20 of these errors, pcregrep is abandoned.
1873
1874 (b) A binary zero was treated as data while matching, but terminated the
1875 output line if it was written out. This has been fixed: binary zeroes
1876 are now no different to any other data bytes.
1877
1878 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1879 used to set a locale for matching. The --locale=xxxx long option has
1880 been added (no short equivalent) to specify a locale explicitly on the
1881 pcregrep command, overriding the environment variables.
1882
1883 (d) When -B was used with -n, some line numbers in the output were one less
1884 than they should have been.
1885
1886 (e) Added the -o (--only-matching) option.
1887
1888 (f) If -A or -C was used with -c (count only), some lines of context were
1889 accidentally printed for the final match.
1890
1891 (g) Added the -H (--with-filename) option.
1892
1893 (h) The combination of options -rh failed to suppress file names for files
1894 that were found from directory arguments.
1895
1896 (i) Added the -D (--devices) and -d (--directories) options.
1897
1898 (j) Added the -F (--fixed-strings) option.
1899
1900 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1901
1902 (l) Added the --colo(u)r option.
1903
1904 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1905 is not present by default.
1906
1907 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1908 items such as (?R) or (?1), when the recursion could match a number of
1909 alternatives. If it matched one of the alternatives, but subsequently,
1910 outside the recursion, there was a failure, the code tried to back up into
1911 the recursion. However, because of the way PCRE is implemented, this is not
1912 possible, and the result was an incorrect result from the match.
1913
1914 In order to prevent this happening, the specification of recursion has
1915 been changed so that all such subpatterns are automatically treated as
1916 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1917
1918 4. I had overlooked the fact that, in some locales, there are characters for
1919 which isalpha() is true but neither isupper() nor islower() are true. In
1920 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1921 and ordfeminine) are like this. This affected the treatment of \w and \W
1922 when they appeared in character classes, but not when they appeared outside
1923 a character class. The bit map for "word" characters is now created
1924 separately from the results of isalnum() instead of just taking it from the
1925 upper, lower, and digit maps. (Plus the underscore character, of course.)
1926
1927 5. The above bug also affected the handling of POSIX character classes such as
1928 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1929 permanent tables. Instead, the bit maps for such a class were previously
1930 created as the appropriate unions of the upper, lower, and digit bitmaps.
1931 Now they are created by subtraction from the [[:word:]] class, which has
1932 its own bitmap.
1933
1934 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1935 It is created by subtracting the vertical space characters (\x09, \x0a,
1936 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1937 subtraction was done in the overall bitmap for a character class, meaning
1938 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1939 be recognized. This bug has been fixed.
1940
1941 7. Patches from the folks at Google:
1942
1943 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1944 real life, but is still worth protecting against".
1945
1946 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1947 regular expressions".
1948
1949 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1950 have it.
1951
1952 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1953 "configure" and the latter not, in order to fix a problem somebody had
1954 with compiling the Arg class on HP-UX.
1955
1956 (e) Improve the error-handling of the C++ wrapper a little bit.
1957
1958 (f) New tests for checking recursion limiting.
1959
1960 8. The pcre_memmove() function, which is used only if the environment does not
1961 have a standard memmove() function (and is therefore rarely compiled),
1962 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1963 returning a result (though PCRE never actually uses the result).
1964
1965 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1966 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1967 returned instead of calling malloc() with an overflowing number that would
1968 most likely cause subsequent chaos.
1969
1970 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1971
1972 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1973 with this option is matched, the nmatch and pmatch options of regexec() are
1974 ignored.
1975
1976 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1977 provided in case anyone wants to the the POSIX interface with UTF-8
1978 strings.
1979
1980 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1981 C++ linking (needed for some HP-UX environments).
1982
1983 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1984 (unused parameter) and in the pcre_printint() function (omitted "default"
1985 switch label when the default is to do nothing).
1986
1987 15. Added some code to make it possible, when PCRE is compiled as a C++
1988 library, to replace subject pointers for pcre_exec() with a smart pointer
1989 class, thus making it possible to process discontinuous strings.
1990
1991 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1992 much the same function. They were added by different people who were trying
1993 to make PCRE easy to compile on non-Unix systems. It has been suggested
1994 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1995 for compiling on Windows systems. I have therefore replaced it with
1996 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1997 defaults to "extern" for C or "extern C" for C++, which works fine on
1998 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1999 SCOPE with something explicit in config.h. In addition:
2000
2001 (a) pcreposix.h still had just "extern" instead of either of these macros;
2002 I have replaced it with PCRE_DATA_SCOPE.
2003
2004 (b) Functions such as _pcre_xclass(), which are internal to the library,
2005 but external in the C sense, all had PCRE_EXPORT in their definitions.
2006 This is apparently wrong for the Windows case, so I have removed it.
2007 (It makes no difference on Unix-like systems.)
2008
2009 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2010 of recursive calls to match(). This is different to MATCH_LIMIT because
2011 that limits the total number of calls to match(), not all of which increase
2012 the depth of recursion. Limiting the recursion depth limits the amount of
2013 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2014 when PCRE is compiled, and changed at run time. A patch from Google adds
2015 this functionality to the C++ interface.
2016
2017 18. Changes to the handling of Unicode character properties:
2018
2019 (a) Updated the table to Unicode 4.1.0.
2020
2021 (b) Recognize characters that are not in the table as "Cn" (undefined).
2022
2023 (c) I revised the way the table is implemented to a much improved format
2024 which includes recognition of ranges. It now supports the ranges that
2025 are defined in UnicodeData.txt, and it also amalgamates other
2026 characters into ranges. This has reduced the number of entries in the
2027 table from around 16,000 to around 3,000, thus reducing its size
2028 considerably. I realized I did not need to use a tree structure after
2029 all - a binary chop search is just as efficient. Having reduced the
2030 number of entries, I extended their size from 6 bytes to 8 bytes to
2031 allow for more data.
2032
2033 (d) Added support for Unicode script names via properties such as \p{Han}.
2034
2035 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2036 matching that character.
2037
2038 20. When matching a repeated Unicode property with a minimum greater than zero,
2039 (for example \pL{2,}), PCRE could look past the end of the subject if it
2040 reached it while seeking the minimum number of characters. This could
2041 happen only if some of the characters were more than one byte long, because
2042 there is a check for at least the minimum number of bytes.
2043
2044 21. Refactored the implementation of \p and \P so as to be more general, to
2045 allow for more different types of property in future. This has changed the
2046 compiled form incompatibly. Anybody with saved compiled patterns that use
2047 \p or \P will have to recompile them.
2048
2049 22. Added "Any" and "L&" to the supported property types.
2050
2051 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2052 but give a compile time error if the value is greater than 0xff.
2053
2054 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2055 accidentally not being installed or uninstalled.
2056
2057 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2058 made were to insert the current release number. This seemed silly, because
2059 it made things harder for people building PCRE on systems that don't run
2060 "configure". I have turned pcre.h into a distributed file, no longer built
2061 by "configure", with the version identification directly included. There is
2062 no longer a pcre.h.in file.
2063
2064 However, this change necessitated a change to the pcre-config script as
2065 well. It is built from pcre-config.in, and one of the substitutions was the
2066 release number. I have updated configure.ac so that ./configure now finds
2067 the release number by grepping pcre.h.
2068
2069 26. Added the ability to run the tests under valgrind.
2070
2071
2072 Version 6.4 05-Sep-05
2073 ---------------------
2074
2075 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2076 "--" to be printed when multiple files were scanned, even when none of the
2077 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2078 consider it to be a bug, and have restored the previous behaviour.
2079
2080 2. A couple of code tidies to get rid of compiler warnings.
2081
2082 3. The pcretest program used to cheat by referring to symbols in the library
2083 whose names begin with _pcre_. These are internal symbols that are not
2084 really supposed to be visible externally, and in some environments it is
2085 possible to suppress them. The cheating is now confined to including
2086 certain files from the library's source, which is a bit cleaner.
2087
2088 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2089 file's purpose clearer.
2090
2091 5. Reorganized pcre_ucp_findchar().
2092
2093
2094 Version 6.3 15-Aug-05
2095 ---------------------
2096
2097 1. The file libpcre.pc.in did not have general read permission in the tarball.
2098
2099 2. There were some problems when building without C++ support:
2100
2101 (a) If C++ support was not built, "make install" and "make test" still
2102 tried to test it.
2103
2104 (b) There were problems when the value of CXX was explicitly set. Some
2105 changes have been made to try to fix these, and ...
2106
2107 (c) --disable-cpp can now be used to explicitly disable C++ support.
2108
2109 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2110 backslash in a target when C++ was disabled. This confuses some
2111 versions of "make", apparently. Using an intermediate variable solves
2112 this. (Same for CPP_LOBJ.)
2113
2114 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2115 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2116 necessary on certain architectures.
2117
2118 4. Added a setting of -export-symbols-regex to the link command to remove
2119 those symbols that are exported in the C sense, but actually are local
2120 within the library, and not documented. Their names all begin with
2121 "_pcre_". This is not a perfect job, because (a) we have to except some
2122 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2123 available (and never for static libraries). I have made a note to try to
2124 find a way round (a) in the future.
2125
2126
2127 Version 6.2 01-Aug-05
2128 ---------------------
2129
2130 1. There was no test for integer overflow of quantifier values. A construction
2131 such as {1111111111111111} would give undefined results. What is worse, if
2132 a minimum quantifier for a parenthesized subpattern overflowed and became
2133 negative, the calculation of the memory size went wrong. This could have
2134 led to memory overwriting.
2135
2136 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2137
2138 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2139 operating environments where this matters.
2140
2141 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2142 PCRE options from within the C++ wrapper.
2143
2144 5. Named capturing subpatterns were not being correctly counted when a pattern
2145 was compiled. This caused two problems: (a) If there were more than 100
2146 such subpatterns, the calculation of the memory needed for the whole
2147 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2148 back references of the form \12, where the number was greater than 9, were
2149 not recognized as back references, even though there were sufficient
2150 previous subpatterns.
2151
2152 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2153 versions of gcc, e.g. 2.95.4.
2154
2155
2156 Version 6.1 21-Jun-05
2157 ---------------------
2158
2159 1. There was one reference to the variable "posix" in pcretest.c that was not
2160 surrounded by "#if !defined NOPOSIX".
2161
2162 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2163 the cross-check on the old pcre_info() function, for the benefit of the
2164 cut-down version of PCRE that is currently imported into Exim.
2165
2166 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2167 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2168 patterns that start (?i) but I don't think that matters. The use of (?i) is
2169 just an example; this all applies to the other options as well.
2170
2171 4. Since libtool seems to echo the compile commands it is issuing, the output
2172 from "make" can be reduced a bit by putting "@" in front of each libtool
2173 compile command.
2174
2175 5. Patch from the folks at Google for configure.in to be a bit more thorough
2176 in checking for a suitable C++ installation before trying to compile the
2177 C++ stuff. This should fix a reported problem when a compiler was present,
2178 but no suitable headers.
2179
2180 6. The man pages all had just "PCRE" as their title. I have changed them to
2181 be the relevant file name. I have also arranged that these names are
2182 retained in the file doc/pcre.txt, which is a concatenation in text format
2183 of all the man pages except the little individual ones for each function.
2184
2185 7. The NON-UNIX-USE file had not been updated for the different set of source
2186 files that come with release 6. I also added a few comments about the C++
2187 wrapper.
2188
2189
2190 Version 6.0 07-Jun-05
2191 ---------------------
2192
2193 1. Some minor internal re-organization to help with my DFA experiments.
2194
2195 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2196 didn't matter for the library itself when fully configured, but did matter
2197 when compiling without UCP support, or within Exim, where the ucp files are
2198 not imported.
2199
2200 3. Refactoring of the library code to split up the various functions into
2201 different source modules. The addition of the new DFA matching code (see
2202 below) to a single monolithic source would have made it really too
2203 unwieldy, quite apart from causing all the code to be include in a
2204 statically linked application, when only some functions are used. This is
2205 relevant even without the DFA addition now that patterns can be compiled in
2206 one application and matched in another.
2207
2208 The downside of splitting up is that there have to be some external
2209 functions and data tables that are used internally in different modules of
2210 the library but which are not part of the API. These have all had their
2211 names changed to start with "_pcre_" so that they are unlikely to clash
2212 with other external names.
2213
2214 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2215 a different (DFA) algorithm. Although it is slower than the original
2216 function, it does have some advantages for certain types of matching
2217 problem.
2218
2219 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2220 including restarting after a partial match.
2221
2222 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2223 defined when compiling for Windows was sent to me. I have put it into the
2224 code, though I have no means of testing or verifying it.
2225
2226 7. Added the pcre_refcount() auxiliary function.
2227
2228 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2229 match before or at the first newline in the subject string. In pcretest,
2230 the /f option on a pattern can be used to set this.
2231
2232 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2233 would behave wrongly. This has been present in PCRE since release 4.0.
2234
2235 10. A number of changes to the pcregrep command:
2236
2237 (a) Refactored how -x works; insert ^(...)$ instead of setting
2238 PCRE_ANCHORED and checking the length, in preparation for adding
2239 something similar for -w.
2240
2241 (b) Added the -w (match as a word) option.
2242
2243 (c) Refactored the way lines are read and buffered so as to have more
2244 than one at a time available.
2245
2246 (d) Implemented a pcregrep test script.
2247
2248 (e) Added the -M (multiline match) option. This allows patterns to match
2249 over several lines of the subject. The buffering ensures that at least
2250 8K, or the rest of the document (whichever is the shorter) is available
2251 for matching (and similarly the previous 8K for lookbehind assertions).
2252
2253 (f) Changed the --help output so that it now says
2254
2255 -w, --word-regex(p)
2256
2257 instead of two lines, one with "regex" and the other with "regexp"
2258 because that confused at least one person since the short forms are the
2259 same. (This required a bit of code, as the output is generated
2260 automatically from a table. It wasn't just a text change.)
2261
2262 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2263 option but starts with a hyphen. Could be a pattern or a path name
2264 starting with a hyphen, for instance.
2265
2266 (h) "-" can be given as a file name to represent stdin.
2267
2268 (i) When file names are being printed, "(standard input)" is used for
2269 the standard input, for compatibility with GNU grep. Previously
2270 "<stdin>" was used.
2271
2272 (j) The option --label=xxx can be used to supply a name to be used for
2273 stdin when file names are being printed. There is no short form.
2274
2275 (k) Re-factored the options decoding logic because we are going to add
2276 two more options that take data. Such options can now be given in four
2277 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2278
2279 (l) Added the -A, -B, and -C options for requesting that lines of context
2280 around matches be printed.
2281
2282 (m) Added the -L option to print the names of files that do not contain
2283 any matching lines, that is, the complement of -l.
2284
2285 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2286 continue to scan other files.
2287
2288 (o) The -s option was incorrectly implemented. For compatibility with other
2289 greps, it now suppresses the error message for a non-existent or non-
2290 accessible file (but not the return code). There is a new option called
2291 -q that suppresses the output of matching lines, which was what -s was
2292 previously doing.
2293
2294 (p) Added --include and --exclude options to specify files for inclusion
2295 and exclusion when recursing.
2296
2297 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2298 Hopefully, it now does.
2299
2300 12. Missing cast in pcre_study().
2301
2302 13. Added an "uninstall" target to the makefile.
2303
2304 14. Replaced "extern" in the function prototypes in Makefile.in with
2305 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2306 world, but is set differently for Windows.
2307
2308 15. Added a second compiling function called pcre_compile2(). The only
2309 difference is that it has an extra argument, which is a pointer to an
2310 integer error code. When there is a compile-time failure, this is set
2311 non-zero, in addition to the error test pointer being set to point to an
2312 error message. The new argument may be NULL if no error number is required
2313 (but then you may as well call pcre_compile(), which is now just a
2314 wrapper). This facility is provided because some applications need a
2315 numeric error indication, but it has also enabled me to tidy up the way
2316 compile-time errors are handled in the POSIX wrapper.
2317
2318 16. Added VPATH=.libs to the makefile; this should help when building with one
2319 prefix path and installing with another. (Or so I'm told by someone who
2320 knows more about this stuff than I do.)
2321
2322 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2323 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2324 match everything, including newlines. This is not POSIX-compatible, but
2325 somebody wanted the feature. From pcretest it can be activated by using
2326 both the P and the s flags.
2327
2328 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2329
2330 19. libpcre.pc was being incorrectly installed as executable.
2331
2332 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2333 it now also looks for '\r' so that it will work unmodified on Windows.
2334
2335 21. Added Google's contributed C++ wrapper to the distribution.
2336
2337 22. Added some untidy missing memory free() calls in pcretest, to keep
2338 Electric Fence happy when testing.
2339
2340
2341
2342 Version 5.0 13-Sep-04
2343 ---------------------
2344
2345 1. Internal change: literal characters are no longer packed up into items
2346 containing multiple characters in a single byte-string. Each character
2347 is now matched using a separate opcode. However, there may be more than one
2348 byte in the character in UTF-8 mode.
2349
2350 2. The pcre_callout_block structure has two new fields: pattern_position and
2351 next_item_length. These contain the offset in the pattern to the next match
2352 item, and its length, respectively.
2353
2354 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2355 insertion of callouts before each pattern item. Added the /C option to
2356 pcretest to make use of this.
2357
2358 4. On the advice of a Windows user, the lines
2359
2360 #if defined(_WIN32) || defined(WIN32)
2361 _setmode( _fileno( stdout ), 0x8000 );
2362 #endif /* defined(_WIN32) || defined(WIN32) */
2363
2364 have been added to the source of pcretest. This apparently does useful
2365 magic in relation to line terminators.
2366
2367 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2368 for the benefit of those environments where the "b" makes a difference.
2369
2370 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2371 to know about it. I have put a hack into configure.in that adds in code
2372 to set GCC=yes if CC=icc. This seems to end up at a point in the
2373 generated configure script that is early enough to affect the setting of
2374 compiler options, which is what is needed, but I have no means of testing
2375 whether it really works. (The user who reported this had patched the
2376 generated configure script, which of course I cannot do.)
2377
2378 LATER: After change 22 below (new libtool files), the configure script
2379 seems to know about icc (and also ecc). Therefore, I have commented out
2380 this hack in configure.in.
2381
2382 7. Added support for pkg-config (2 patches were sent in).
2383
2384 8. Negated POSIX character classes that used a combination of internal tables
2385 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2386 [[:^ascii]]. Typically, they would match almost any characters. The other
2387 POSIX classes were not broken in this way.
2388
2389 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2390 to find the match, as PCRE was deluded into thinking that the match had to
2391 start at the start point or following a newline. The same bug applied to
2392 patterns with negative forward assertions or any backward assertions
2393 preceding ".*" at the start, unless the pattern required a fixed first
2394 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2395
2396 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2397 starting at the last subject character, bytes beyond the end of the subject
2398 string were read.
2399
2400 11. Renamed the variable "class" as "classbits" to make life easier for C++
2401 users. (Previously there was a macro definition, but it apparently wasn't
2402 enough.)
2403
2404 12. Added the new field "tables" to the extra data so that tables can be passed
2405 in at exec time, or the internal tables can be re-selected. This allows
2406 a compiled regex to be saved and re-used at a later time by a different
2407 program that might have everything at different addresses.
2408
2409 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2410 -R library as well as a -L library.
2411
2412 14. The debugging options of pcretest (-d on the command line or D on a
2413 pattern) showed incorrect output for anything following an extended class
2414 that contained multibyte characters and which was followed by a quantifier.
2415
2416 15. Added optional support for general category Unicode character properties
2417 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2418 support. It adds about 90K to the size of the library. The meanings of the
2419 inbuilt class escapes such as \d and \s have NOT been changed.
2420
2421 16. Updated pcredemo.c to include calls to free() to release the memory for the
2422 compiled pattern.
2423
2424 17. The generated file chartables.c was being created in the source directory
2425 instead of in the building directory. This caused the build to fail if the
2426 source directory was different from the building directory, and was
2427 read-only.
2428
2429 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2430 file. No doubt somebody will tell me if they don't make sense... Also added
2431 Dan Mooney's comments about building on OpenVMS.
2432
2433 19. Added support for partial matching via the PCRE_PARTIAL option for
2434 pcre_exec() and the \P data escape in pcretest.
2435
2436 20. Extended pcretest with 3 new pattern features:
2437
2438 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2439 write the compiled pattern to the file whose name is "rest-of-line".
2440 This is a straight binary dump of the data, with the saved pointer to
2441 the character tables forced to be NULL. The study data, if any, is
2442 written too. After writing, pcretest reads a new pattern.
2443
2444 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2445 compiled pattern from the given file. There must not be any
2446 occurrences of "<" in the file name (pretty unlikely); if there are,
2447 pcretest will instead treat the initial "<" as a pattern delimiter.
2448 After reading in the pattern, pcretest goes on to read data lines as
2449 usual.
2450
2451 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2452 and 16-bit fields in a compiled pattern, to simulate a pattern that
2453 was compiled on a host of opposite endianness.
2454
2455 21. The pcre-exec() function can now cope with patterns that were compiled on
2456 hosts of opposite endianness, with this restriction:
2457
2458 As for any compiled expression that is saved and used later, the tables
2459 pointer field cannot be preserved; the extra_data field in the arguments
2460 to pcre_exec() should be used to pass in a tables address if a value
2461 other than the default internal tables were used at compile time.
2462
2463 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2464 now diagnosed as an error. Previously, most of the time, a negative number
2465 would have been treated as zero, but if in addition "ovector" was passed as
2466 NULL, a crash could occur.
2467
2468 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2469 new versions from the libtool 1.5 distribution (the last one is a copy of
2470 a file called libtool.m4). This seems to have fixed the need to patch
2471 "configure" to support Darwin 1.3 (which I used to do). However, I still
2472 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2473 workstation).
2474
2475 24. Changed the PCRE licence to be the more standard "BSD" licence.
2476
2477
2478 Version 4.5 01-Dec-03
2479 ---------------------
2480
2481 1. There has been some re-arrangement of the code for the match() function so
2482 that it can be compiled in a version that does not call itself recursively.
2483 Instead, it keeps those local variables that need separate instances for
2484 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2485 needs to "recurse". Keeping track of where control must go is done by means
2486 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2487 hide most of the details from the main code, and operates only if
2488 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2489 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2490 operating.
2491
2492 To make it easier for callers to provide specially tailored get/free
2493 functions for this usage, two new functions, pcre_stack_malloc, and
2494 pcre_stack_free, are used. They are always called in strict stacking order,
2495 and the size of block requested is always the same.
2496
2497 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2498 PCRE has been compiled to use the stack or the heap for recursion. The
2499 -C option of pcretest uses this to show which version is compiled.
2500
2501 A new data escape \S, is added to pcretest; it causes the amounts of store
2502 obtained and freed by both kinds of malloc/free at match time to be added
2503 to the output.
2504
2505 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2506 what's available on my current Linux desktop machine.
2507
2508 3. When matching a UTF-8 string, the test for a valid string at the start has
2509 been extended. If start_offset is not zero, PCRE now checks that it points
2510 to a byte that is the start of a UTF-8 character. If not, it returns
2511 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2512 this is necessary because there may be backward assertions in the pattern.
2513 When matching the same subject several times, it may save resources to use
2514 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2515
2516 4. The code for checking the validity of UTF-8 strings has been tightened so
2517 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2518 containing "overlong sequences".
2519
2520 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2521 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2522 should have been "&", but it just so happened that all the cases this let
2523 through by mistake were picked up later in the function.
2524
2525 6. I had used a variable called "isblank" - this is a C99 function, causing
2526 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2527
2528 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2529 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2530 so the version is shown; (c) stop "make test" echoing "./RunTest".
2531
2532 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2533
2534 9. The prototype for memmove() for systems that don't have it was using
2535 size_t, but the inclusion of the header that defines size_t was later. I've
2536 moved the #includes for the C headers earlier to avoid this.
2537
2538 10. Added some adjustments to the code to make it easier to compiler on certain
2539 special systems:
2540
2541 (a) Some "const" qualifiers were missing.
2542 (b) Added the macro EXPORT before all exported functions; by default this
2543 is defined to be empty.
2544 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2545 that it reads its output file name as an argument instead of writing
2546 to the standard output and assuming this can be redirected.
2547
2548 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2549 class containing characters with values greater than 255, PCRE compilation
2550 went into a loop.
2551
2552 12. A recursive reference to a subpattern that was within another subpattern
2553 that had a minimum quantifier of zero caused PCRE to crash. For example,
2554 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2555 recursion. If the recursively-called subpattern itself had a zero repeat,
2556 that was OK.
2557
2558 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2559 buffer into which it was copied (for escape processing) was still set at
2560 1024, so long lines caused crashes.
2561
2562 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2563 "internal error: code overflow...". This applied to any character class
2564 that was followed by a possessive quantifier.
2565
2566 15. Modified the Makefile to add libpcre.la as a prerequisite for
2567 libpcreposix.la because I was told this is needed for a parallel build to
2568 work.
2569
2570 16. If a pattern that contained .* following optional items at the start was
2571 studied, the wrong optimizing data was generated, leading to matching
2572 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2573 matching string must start with a or b or c. The correct conclusion for
2574 this pattern is that a match can start with any character.
2575
2576
2577 Version 4.4 13-Aug-03
2578 ---------------------
2579
2580 1. In UTF-8 mode, a character class containing characters with values between
2581 127 and 255 was not handled correctly if the compiled pattern was studied.
2582 In fixing this, I have also improved the studying algorithm for such
2583 classes (slightly).
2584
2585 2. Three internal functions had redundant arguments passed to them. Removal
2586 might give a very teeny performance improvement.
2587
2588 3. Documentation bug: the value of the capture_top field in a callout is *one
2589 more than* the number of the hightest numbered captured substring.
2590
2591 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2592 in incorrectly linking with a previously installed version. They now link
2593 explicitly with libpcre.la.
2594
2595 5. configure.in no longer needs to recognize Cygwin specially.
2596
2597 6. A problem in pcre.in for Windows platforms is fixed.
2598
2599 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2600 pcretest, it used to include the size of the study block as part of its
2601 output. Unfortunately, the structure contains a field that has a different
2602 size on different hardware architectures. This meant that the tests that
2603 showed this size failed. As the block is currently always of a fixed size,
2604 this information isn't actually particularly useful in pcretest output, so
2605 I have just removed it.
2606
2607 8. Three pre-processor statements accidentally did not start in column 1.
2608 Sadly, there are *still* compilers around that complain, even though
2609 standard C has not required this for well over a decade. Sigh.
2610
2611 9. In pcretest, the code for checking callouts passed small integers in the
2612 callout_data field, which is a void * field. However, some picky compilers
2613 complained about the casts involved for this on 64-bit systems. Now
2614 pcretest passes the address of the small integer instead, which should get
2615 rid of the warnings.
2616
2617 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2618 both compile and run time, and gives an error if an invalid UTF-8 sequence
2619 is found. There is a option for disabling this check in cases where the
2620 string is known to be correct and/or the maximum performance is wanted.
2621
2622 11. In response to a bug report, I changed one line in Makefile.in from
2623
2624 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2625 to
2626 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2627
2628 to look similar to other lines, but I have no way of telling whether this
2629 is the right thing to do, as I do not use Windows. No doubt I'll get told
2630 if it's wrong...
2631
2632
2633 Version 4.3 21-May-03
2634 ---------------------
2635
2636 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2637 Makefile.
2638
2639 2. Some refactoring to improve the quality of the code:
2640
2641 (i) The utf8_table... variables are now declared "const".
2642
2643 (ii) The code for \cx, which used the "case flipping" table to upper case
2644 lower case letters, now just substracts 32. This is ASCII-specific,
2645 but the whole concept of \cx is ASCII-specific, so it seems
2646 reasonable.
2647
2648 (iii) PCRE was using its character types table to recognize decimal and
2649 hexadecimal digits in the pattern. This is silly, because it handles
2650 only 0-9, a-f, and A-F, but the character types table is locale-
2651 specific, which means strange things might happen. A private
2652 table is now used for this - though it costs 256 bytes, a table is
2653 much faster than multiple explicit tests. Of course, the standard
2654 character types table is still used for matching digits in subject
2655 strings against \d.
2656
2657 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2658 ending in _t are). So I've renamed it as ESC_tee.
2659
2660 3. The first argument for regexec() in the POSIX wrapper should have been
2661 defined as "const".
2662
2663 4. Changed pcretest to use malloc() for its buffers so that they can be
2664 Electric Fenced for debugging.
2665
2666 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2667 to read one or more bytes before the start of the subject string. Often this
2668 had no effect on PCRE's behaviour, but in some circumstances it could
2669 provoke a segmentation fault.
2670
2671 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2672 to try to read one or more bytes before the start of the subject string.
2673
2674 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2675 UTF-8 support could misbehave in various ways if the subject string
2676 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2677 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2678 back over UTF-8 characters.)
2679
2680
2681 Version 4.2 14-Apr-03
2682 ---------------------
2683
2684 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2685
2686 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2687 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2688 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2689 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2690 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2691 and BUILD_EXEEXT
2692 Note: automatic setting of the BUILD variables is not yet working
2693 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2694 compile-time but not at link-time
2695 [LINK]: use for linking executables only
2696 make different versions for Windows and non-Windows
2697 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2698 libraries
2699 [LINK_FOR_BUILD]: new variable
2700 [OBJEXT]: use throughout
2701 [EXEEXT]: use throughout
2702 <winshared>: new target
2703 <wininstall>: new target
2704 <dftables.o>: use native compiler
2705 <dftables>: use native linker
2706 <install>: handle Windows platform correctly
2707 <clean>: ditto
2708 <check>: ditto
2709 copy DLL to top builddir before testing
2710
2711 As part of these changes, -no-undefined was removed again. This was reported
2712 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2713 in any case.
2714
2715 3. Some tidies to get rid of compiler warnings:
2716
2717 . In the match_data structure, match_limit was an unsigned long int, whereas
2718 match_call_count was an int. I've made them both unsigned long ints.
2719
2720 . In pcretest the fact that a const uschar * doesn't automatically cast to
2721 a void * provoked a warning.
2722
2723 . Turning on some more compiler warnings threw up some "shadow" variables
2724 and a few more missing casts.
2725
2726 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2727 option, a class that contained a single character with a value between 128
2728 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2729
2730 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2731 option, a class that contained several characters, but with at least one
2732 whose value was between 128 and 255 caused PCRE to crash.
2733
2734
2735 Version 4.1 12-Mar-03
2736 ---------------------
2737
2738 1. Compiling with gcc -pedantic found a couple of places where casts were
2739 needed, and a string in dftables.c that was longer than standard compilers are
2740 required to support.
2741
2742 2. Compiling with Sun's compiler found a few more places where the code could
2743 be tidied up in order to avoid warnings.
2744
2745 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2746 first of these names is deprecated in the latest Autoconf in favour of the name
2747 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2748 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2749 analogy I have changed it to CFLAGS_FOR_BUILD.
2750
2751 4. Added -no-undefined to the linking command in the Makefile, because this is
2752 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2753 linking step for the pcreposix library.
2754
2755 5. PCRE was failing to diagnose the case of two named groups with the same
2756 name.
2757
2758 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2759 literal character that is needed in the subject for a match, and scans along to
2760 ensure that it is present before embarking on the full matching process. This
2761 saves time in cases of nested unlimited repeats that are never going to match.
2762 Problem: the scan can take a lot of time if the subject is very long (e.g.
2763 megabytes), thus penalizing straightforward matches. It is now done only if the
2764 amount of subject to be scanned is less than 1000 bytes.
2765
2766 7. A lesser problem with the same optimization is that it was recording the
2767 first character of an anchored pattern as "needed", thus provoking a search
2768 right along the subject, even when the first match of the pattern was going to
2769 fail. The "needed" character is now not set for anchored patterns, unless it
2770 follows something in the pattern that is of non-fixed length. Thus, it still
2771 fulfils its original purpose of finding quick non-matches in cases of nested
2772 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2773
2774
2775 Version 4.0 17-Feb-03
2776 ---------------------
2777
2778 1. If a comment in an extended regex that started immediately after a meta-item
2779 extended to the end of string, PCRE compiled incorrect data. This could lead to
2780 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2781
2782 2. Moved to autoconf 2.53 and libtool 1.4.2.
2783
2784 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2785 the special perltest8 script is no longer needed - all the tests can be run
2786 from a single perltest script.
2787
2788 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2789 by \s. It has now been removed in PCRE. This means it isn't recognized as
2790 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2791 class [:space:] *does* include VT, thereby creating a mess.
2792
2793 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2794 space and tab.
2795
2796 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2797 its new features into the main test script, reducing the number of scripts.
2798
2799 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2800 were backward compatible, and made the (?i) apply to the whole pattern, as if
2801 /i were given. Now it behaves more logically, and applies the option setting
2802 only to what follows. PCRE has been changed to follow suit. However, if it
2803 finds options settings right at the start of the pattern, it extracts them into
2804 the global options, as before. Thus, they show up in the info data.
2805
2806 8. Added support for the \Q...\E escape sequence. Characters in between are
2807 treated as literals. This is slightly different from Perl in that $ and @ are
2808 also handled as literals inside the quotes. In Perl, they will cause variable
2809 interpolation. Note the following examples:
2810
2811 Pattern PCRE matches Perl matches
2812
2813 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2814 \Qabc\$xyz\E abc\$xyz abc\$xyz
2815 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2816
2817 For compatibility with Perl, \Q...\E sequences are recognized inside character
2818 classes as well as outside them.
2819
2820 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2821 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2822 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2823 signed/unsigned warnings.
2824
2825 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2826 option for pcretest, so I've replaced it by a simple function that does just
2827 that job.
2828
2829 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2830 "pcregrep -".
2831
2832 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2833 Java package. This provides some syntactic sugar for simple cases of what my
2834 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2835 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2836 item, you can use this simplified notation. Note that only makes sense with
2837 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2838 greediness, whatever the setting of the PCRE_UNGREEDY option.
2839
2840 13. A change of greediness default within a pattern was not taking effect at
2841 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2842 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2843 was abstracted outside.
2844
2845 14. PCRE now supports the \G assertion. It is true when the current matching
2846 position is at the start point of the match. This differs from \A when the
2847 starting offset is non-zero. Used with the /g option of pcretest (or similar
2848 code), it works in the same way as it does for Perl's /g option. If all
2849 alternatives of a regex begin with \G, the expression is anchored to the start
2850 match position, and the "anchored" flag is set in the compiled expression.
2851
2852 15. Some bugs concerning the handling of certain option changes within patterns
2853 have been fixed. These applied to options other than (?ims). For example,
2854 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2855 been the other way round. Some of this was related to change 7 above.
2856
2857 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2858 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2859 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2860 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2861
2862 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2863 mode. Unlike ".", it always matches newline, whatever the setting of
2864 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2865 assertions. Perl allows it, but it doesn't (in general) work because it can't
2866 calculate the length of the lookbehind. At least, that's the case for Perl
2867 5.8.0 - I've been told they are going to document that it doesn't work in
2868 future.
2869
2870 18. Added an error diagnosis for escapes that PCRE does not support: these are
2871 \L, \l, \N, \P, \p, \U, \u, and \X.
2872
2873 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2874 reading past the end of the pattern in cases such as /[abcd/.
2875
2876 20. PCRE was getting more memory than necessary for patterns with classes that
2877 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2878
2879 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2880 compiling PCRE for use with Virtual Pascal.
2881
2882 22. Small fix to the Makefile to make it work properly if the build is done
2883 outside the source tree.
2884
2885 23. Added a new extension: a condition to go with recursion. If a conditional
2886 subpattern starts with (?(R) the "true" branch is used if recursion has
2887 happened, whereas the "false" branch is used only at the top level.
2888
2889 24. When there was a very long string of literal characters (over 255 bytes
2890 without UTF support, over 250 bytes with UTF support), the computation of how
2891 much memory was required could be incorrect, leading to segfaults or other
2892 strange effects.
2893
2894 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2895 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2896 there was a subsequent back reference to those brackets. This meant that, for
2897 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2898 possible to check for precisely this case. All we can do is abandon the
2899 optimization if .* occurs inside capturing brackets when there are any back
2900 references whatsoever. (See below for a better fix that came later.)
2901
2902 26. The handling of the optimization for finding the first character of a
2903 non-anchored pattern, and for finding a character that is required later in the
2904 match were failing in some cases. This didn't break the matching; it just
2905 failed to optimize when it could. The way this is done has been re-implemented.
2906
2907 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2908
2909 28. Added a new feature that provides some of the functionality that Perl
2910 provides with (?{...}). The facility is termed a "callout". The way it is done
2911 in PCRE is for the caller to provide an optional function, by setting
2912 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2913 global variable. By default it is unset, which disables all calling out. To get
2914 the function called, the regex must include (?C) at appropriate points. This
2915 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2916 This provides a means of identifying different callout points. When PCRE
2917 reaches such a point in the regex, if pcre_callout has been set, the external
2918 function is called. It is provided with data in a structure called
2919 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2920 matching continues; if it returns a non-zero value, the match at the current
2921 point fails. However, backtracking will occur if possible. [This was changed
2922 later and other features added - see item 49 below.]
2923
2924 29. pcretest is upgraded to test the callout functionality. It provides a
2925 callout function that displays information. By default, it shows the start of
2926 the match and the current position in the text. There are some new data escapes
2927 to vary what happens:
2928
2929 \C+ in addition, show current contents of captured substrings
2930 \C- do not supply a callout function
2931 \C!n return 1 when callout number n is reached
2932 \C!n!m return 1 when callout number n is reached for the mth time
2933
2934 30. If pcregrep was called with the -l option and just a single file name, it
2935 output "<stdin>" if a match was found, instead of the file name.
2936
2937 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2938 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2939 pcre_exec(). This saves a malloc/free per call. The default value of
2940 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2941 when configuring.
2942
2943 32. The default maximum size of a compiled pattern is 64K. There have been a
2944 few cases of people hitting this limit. The code now uses macros to handle the
2945 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2946 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2947 configuring. Tests 2 and 5 work only with 2-byte links because they output
2948 debugging information about compiled patterns.
2949
2950 33. Internal code re-arrangements:
2951
2952 (a) Moved the debugging function for printing out a compiled regex into
2953 its own source file (printint.c) and used #include to pull it into
2954 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2955 separate copies.
2956
2957 (b) Defined the list of op-code names for debugging as a macro in
2958 internal.h so that it is next to the definition of the opcodes.
2959
2960 (c) Defined a table of op-code lengths for simpler skipping along compiled
2961 code. This is again a macro in internal.h so that it is next to the
2962 definition of the opcodes.
2963
2964 34. Added support for recursive calls to individual subpatterns, along the
2965 lines of Robin Houston's patch (but implemented somewhat differently).
2966
2967 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2968 allow it to read and process whole directories in Win32. This code was
2969 contributed by Lionel Fourquaux; it has not been tested by me.
2970
2971 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2972 used to name a group. Names consist of alphanumerics and underscores, and must
2973 be unique. Back references use the syntax (?P=name) and recursive calls use
2974 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2975 numbers. The function pcre_fullinfo() can be used after compilation to extract
2976 a name/number map. There are three relevant calls:
2977
2978 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2979 PCRE_INFO_NAMECOUNT yields the number of entries
2980 PCRE_INFO_NAMETABLE yields a pointer to the map.
2981
2982 The map is a vector of fixed-size entries. The size of each entry depends on
2983 the length of the longest name used. The first two bytes of each entry are the
2984 group number, most significant byte first. There follows the corresponding
2985 name, zero terminated. The names are in alphabetical order.
2986
2987 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2988 case instead of 255. Making it the same both with and without UTF-8 support
2989 means that the same test output works with both.
2990
2991 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2992 calling malloc() with a zero argument.
2993
2994 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2995 optimization. I've improved things by keeping a bitmap of backreferences with
2996 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2997 fact referenced, the optimization can be applied. It is unlikely that a
2998 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2999 the match to follow \n) will appear inside brackets with a number greater than
3000 31, but if it does, any back reference > 31 suppresses the optimization.
3001
3002 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3003 of disabling numbered capturing parentheses. Any opening parenthesis that is
3004 not followed by ? behaves as if it were followed by ?: but named parentheses
3005 can still be used for capturing (and they will acquire numbers in the usual
3006 way).
3007
3008 41. Redesigned the return codes from the match() function into yes/no/error so
3009 that errors can be passed back from deep inside the nested calls. A malloc
3010 failure while inside a recursive subpattern call now causes the
3011 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3012
3013 42. It is now possible to set a limit on the number of times the match()
3014 function is called in a call to pcre_exec(). This facility makes it possible to
3015 limit the amount of recursion and backtracking, though not in a directly
3016 obvious way, because the match() function is used in a number of different
3017 circumstances. The count starts from zero for each position in the subject
3018 string (for non-anchored patterns). The default limit is, for compatibility, a
3019 large number, namely 10 000 000. You can change this in two ways:
3020
3021 (a) When configuring PCRE before making, you can use --with-match-limit=n
3022 to set a default value for the compiled library.
3023
3024 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3025 a different value is set. See 45 below.
3026
3027 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3028
3029 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3030 of things that can be changed at compile time. The first argument specifies
3031 what is wanted and the second points to where the information is to be placed.
3032 The current list of available information is:
3033
3034 PCRE_CONFIG_UTF8
3035
3036 The output is an integer that is set to one if UTF-8 support is available;
3037 otherwise it is set to zero.
3038
3039 PCRE_CONFIG_NEWLINE
3040
3041 The output is an integer that it set to the value of the code that is used for
3042 newline. It is either LF (10) or CR (13).
3043
3044 PCRE_CONFIG_LINK_SIZE
3045
3046 The output is an integer that contains the number of bytes used for internal
3047 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3048
3049 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3050
3051 The output is an integer that contains the threshold above which the POSIX
3052 interface uses malloc() for output vectors. See item 31 above.
3053
3054 PCRE_CONFIG_MATCH_LIMIT
3055
3056 The output is an unsigned integer that contains the default limit of the number
3057 of match() calls in a pcre_exec() execution. See 42 above.
3058
3059 44. pcretest has been upgraded by the addition of the -C option. This causes it
3060 to extract all the available output from the new pcre_config() function, and to
3061 output it. The program then exits immediately.
3062
3063 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3064 order to support additional features. One way would have been to define
3065 pcre_exec2() (for example) with extra arguments, but this would not have been
3066 extensible, and would also have required all calls to the original function to
3067 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3068 is used for passing in "extra" data from pcre_study().
3069
3070 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3071 contains the following fields:
3072
3073 flags a bitmap indicating which of the following fields are set
3074 study_data opaque data from pcre_study()
3075 match_limit a way of specifying a limit on match() calls for a specific
3076 call to pcre_exec()
3077 callout_data data for callouts (see 49 below)
3078
3079 The flag bits are also defined in pcre.h, and are
3080
3081 PCRE_EXTRA_STUDY_DATA
3082 PCRE_EXTRA_MATCH_LIMIT
3083 PCRE_EXTRA_CALLOUT_DATA
3084
3085 The pcre_study() function now returns one of these new pcre_extra blocks, with
3086 the actual study data pointed to by the study_data field, and the
3087 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3088 before. That is, this change is entirely upwards-compatible and requires no
3089 change to existing code.
3090
3091 If you want to pass in additional data to pcre_exec(), you can either place it
3092 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3093 block.
3094
3095 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3096 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3097 times with different match limits, until it finds the minimum value needed for
3098 pcre_exec() to complete. The value is then output. This can be instructive; for
3099 most simple matches the number is quite small, but for pathological cases it
3100 gets very large very quickly.
3101
3102 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3103 returns the size of the data block pointed to by the study_data field in a
3104 pcre_extra block, that is, the value that was passed as the argument to
3105 pcre_malloc() when PCRE was getting memory in which to place the information
3106 created by pcre_study(). The fourth argument should point to a size_t variable.
3107 pcretest has been extended so that this information is shown after a successful
3108 pcre_study() call when information about the compiled regex is being displayed.
3109
3110 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3111 because what follows is always an absolute path. (Later: it turns out that this
3112 is more than cosmetic for MinGW, because it doesn't like empty path
3113 components.)
3114
3115 49. Some changes have been made to the callout feature (see 28 above):
3116
3117 (i) A callout function now has three choices for what it returns:
3118
3119 0 => success, carry on matching
3120 > 0 => failure at this point, but backtrack if possible
3121 < 0 => serious error, return this value from pcre_exec()
3122
3123 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3124 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3125 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3126 use by callout functions. It will never be used by PCRE itself.
3127
3128 (ii) The pcre_extra structure (see 45 above) has a void * field called
3129 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3130 pcre_callout_block structure has a field of the same name. The contents of
3131 the field passed in the pcre_extra structure are passed to the callout
3132 function in the corresponding field in the callout block. This makes it
3133 easier to use the same callout-containing regex from multiple threads. For
3134 testing, the pcretest program has a new data escape
3135
3136 \C*n pass the number n (may be negative) as callout_data
3137
3138 If the callout function in pcretest receives a non-zero value as
3139 callout_data, it returns that value.
3140
3141 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3142 there were some redundant $(CFLAGS) in commands that are now specified as
3143 $(LINK), which already includes $(CFLAGS).
3144
3145 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3146 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3147 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3148 one-byte characters throughout. Note that case-insensitive matching applies
3149 only to characters whose values are less than 256. PCRE doesn't support the
3150 notion of cases for higher-valued characters.
3151
3152 (i) A character class whose characters are all within 0-255 is handled as
3153 a bit map, and the map is inverted for negative classes. Previously, a
3154 character > 255 always failed to match such a class; however it should
3155 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3156
3157 (ii) A negated character class with a single character < 255 is coded as
3158 "not this character" (OP_NOT). This wasn't working properly when the test
3159 character was multibyte, either singly or repeated.
3160
3161 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3162 mode, for example: \x{100}{2,3}.
3163
3164 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3165 singly or repeated) now correctly test multibyte characters. However,
3166 PCRE doesn't recognize any characters with values greater than 255 as
3167 digits, spaces, or word characters. Such characters always match \D, \S,
3168 and \W, and never match \d, \s, or \w.
3169
3170 (v) Classes may now contain characters and character ranges with values
3171 greater than 255. For example: [ab\x{100}-\x{400}].
3172
3173 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3174 PCRE in UTF-8 mode.
3175
3176 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3177 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3178 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3179 value.)
3180
3181 53. The single man page has become too large. I have therefore split it up into
3182 a number of separate man pages. These also give rise to individual HTML pages;
3183 these are now put in a separate directory, and there is an index.html page that
3184 lists them all. Some hyperlinking between the pages has been installed.
3185
3186 54. Added convenience functions for handling named capturing parentheses.
3187
3188 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3189 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3190 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3191 are faulted.
3192
3193 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3194 calling configure. These values are used when compiling the dftables.c program
3195 which is run to generate the source of the default character tables. They
3196 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3197 you will need to set these values.
3198
3199 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3200
3201
3202 Version 3.9 02-Jan-02
3203 ---------------------
3204
3205 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3206
3207 2. If --disable-static was given, the building process failed when trying to
3208 build pcretest and pcregrep. (For some reason it was using libtool to compile
3209 them, which is not right, as they aren't part of the library.)
3210
3211
3212 Version 3.8 18-Dec-01
3213 ---------------------
3214
3215 1. The experimental UTF-8 code was completely screwed up. It was packing the
3216 bytes in the wrong order. How dumb can you get?
3217
3218
3219 Version 3.7 29-Oct-01
3220 ---------------------
3221
3222 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3223 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3224 this didn't happen under Solaris 8, where I normally test things.
3225
3226 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3227 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3228 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3229 this fix an hour or so after the initial 3.7 release.)
3230
3231
3232 Version 3.6 23-Oct-01
3233 ---------------------
3234
3235 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3236 offsets passed as NULL with zero offset count.
3237
3238 2. The config.guess and config.sub files had not been updated when I moved to
3239 the latest autoconf.
3240
3241
3242 Version 3.5 15-Aug-01
3243 ---------------------
3244
3245 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3246 had been forgotten.
3247
3248 2. By using declared but undefined structures, we can avoid using "void"
3249 definitions in pcre.h while keeping the internal definitions of the structures
3250 private.
3251
3252 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3253 user point of view, this means that both static and shared libraries are built
3254 by default, but this can be individually controlled. More of the work of
3255 handling this static/shared cases is now inside libtool instead of PCRE's make
3256 file.
3257
3258 4. The pcretest utility is now installed along with pcregrep because it is
3259 useful for users (to test regexs) and by doing this, it automatically gets
3260 relinked by libtool. The documentation has been turned into a man page, so
3261 there are now .1, .txt, and .html versions in /doc.
3262
3263 5. Upgrades to pcregrep:
3264 (i) Added long-form option names like gnu grep.
3265 (ii) Added --help to list all options with an explanatory phrase.
3266 (iii) Added -r, --recursive to recurse into sub-directories.
3267 (iv) Added -f, --file to read patterns from a file.
3268
3269 6. pcre_exec() was referring to its "code" argument before testing that
3270 argument for NULL (and giving an error if it was NULL).
3271
3272 7. Upgraded Makefile.in to allow for compiling in a different directory from
3273 the source directory.
3274
3275 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3276 options bits, the pointer it was passed was to an int instead of to an unsigned
3277 long int. This mattered only on 64-bit systems.
3278
3279 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3280 generated) instead of pcre.in, which it its source. Also made the same change
3281 in several of the .c files.
3282
3283 10. A new release of gcc defines printf() as a macro, which broke pcretest
3284 because it had an ifdef in the middle of a string argument for printf(). Fixed
3285 by using separate calls to printf().
3286
3287 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3288 script, to force use of CR or LF instead of \n in the source. On non-Unix
3289 systems, the value can be set in config.h.
3290
3291 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3292 absolute limit. Changed the text of the error message to make this clear, and
3293 likewise updated the man page.
3294
3295 13. The limit of 99 on the number of capturing subpatterns has been removed.
3296 The new limit is 65535, which I hope will not be a "real" limit.
3297
3298
3299 Version 3.4 22-Aug-00
3300 ---------------------
3301
3302 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3303
3304 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3305
3306
3307 Version 3.3 01-Aug-00
3308 ---------------------
3309
3310 1. If an octal character was given, but the value was greater than \377, it
3311 was not getting masked to the least significant bits, as documented. This could
3312 lead to crashes in some systems.
3313
3314 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3315 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3316
3317 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3318 These just pass their arguments on to (pcre_free)(), but they are provided
3319 because some uses of PCRE bind it to non-C systems that can call its functions,
3320 but cannot call free() or pcre_free() directly.
3321
3322 4. Add "make test" as a synonym for "make check". Corrected some comments in
3323 the Makefile.
3324
3325 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3326 Makefile.
3327
3328 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3329 command called pgrep for grepping around the active processes.
3330
3331 7. Added the beginnings of support for UTF-8 character strings.
3332
3333 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3334 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3335 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3336 out for the ar command.)
3337
3338
3339 Version 3.2 12-May-00
3340 ---------------------
3341
3342 This is purely a bug fixing release.
3343
3344 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3345 of ZA. This was just one example of several cases that could provoke this bug,
3346 which was introduced by change 9 of version 2.00. The code for breaking
3347 infinite loops after an iteration that matches an empty string was't working
3348 correctly.
3349
3350 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3351 when matched against abbab (for example). After matching an empty string, it
3352 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3353 caused it to match further down the string than it should.
3354
3355 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3356 was there because it doesn't seem to be needed, and it causes trouble on some
3357 systems, as it is not a Standard C header. It has been removed.
3358
3359 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3360 were reported on the Macintosh. The changes were from
3361
3362 while ((c = *(++ptr)) != 0 && c != '\n');
3363 to
3364 while ((c = *(++ptr)) != 0 && c != '\n') ;
3365
3366 Totally extraordinary, but if that's what it takes...
3367
3368 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3369 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3370 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3371 assumes the way PCRE uses memmove() (always moving upwards).
3372
3373 6. PCRE is being used in one environment where strchr() is not available. There
3374 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3375 faster code anyway.
3376
3377
3378 Version 3.1 09-Feb-00
3379 ---------------------
3380
3381 The only change in this release is the fixing of some bugs in Makefile.in for
3382 the "install" target:
3383
3384 (1) It was failing to install pcreposix.h.
3385
3386 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3387
3388
3389 Version 3.0 01-Feb-00
3390 ---------------------
3391
3392 1. Add support for the /+ modifier to perltest (to output $` like it does in
3393 pcretest).
3394
3395 2. Add support for the /g modifier to perltest.
3396
3397 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3398 matches null strings.
3399
3400 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3401 pattern. Perl treats empty patterns specially - it reuses the most recent
3402 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3403 effect.
3404
3405 5. The POSIX interface was broken in that it was just handing over the POSIX
3406 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3407 required a bigger vector, with some working space on the end. This means that
3408 the POSIX wrapper now has to get and free some memory, and copy the results.
3409
3410 6. Added some simple autoconf support, placing the test data and the
3411 documentation in separate directories, re-organizing some of the
3412 information files, and making it build pcre-config (a GNU standard). Also added
3413 libtool support for building PCRE as a shared library, which is now the
3414 default.
3415
3416 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3417 09 are not valid octal constants. Single digits will be used for minor values
3418 less than 10.
3419
3420 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3421 existing programs that set these in the POSIX interface can use PCRE without
3422 modification.
3423
3424 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3425 return all that pcre_info() returns, plus additional data. The pcre_info()
3426 function is retained for compatibility, but is considered to be obsolete.
3427
3428 10. Added experimental recursion feature (?R) to handle one common case that
3429 Perl 5.6 will be able to do with (?p{...}).
3430
3431 11. Added support for POSIX character classes like [:alpha:], which Perl is
3432 adopting.
3433
3434
3435 Version 2.08 31-Aug-99
3436 ----------------------
3437
3438 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3439 trying to match at the startoffset position, but instead was moving forward to
3440 the next newline as if a previous match had failed.
3441
3442 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3443 and could get into a loop if a null string was matched other than at the start
3444 of the subject.
3445
3446 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3447 be distinguished at compile time, and for completeness also added PCRE_DATE.
3448
3449 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3450 in GnuWin32 environments.
3451
3452
3453 Version 2.07 29-Jul-99
3454 ----------------------
3455
3456 1. The documentation is now supplied in plain text form and HTML as well as in
3457 the form of man page sources.
3458
3459 2. C++ compilers don't like assigning (void *) values to other pointer types.
3460 In particular this affects malloc(). Although there is no problem in Standard
3461 C, I've put in casts to keep C++ compilers happy.
3462
3463 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3464 should be (const char *).
3465
3466 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3467 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3468 However, I haven't made this a standard facility. The documentation doesn't
3469 mention it, and the Makefile doesn't support it.
3470
3471 5. The Makefile now contains an "install" target, with editable destinations at
3472 the top of the file. The pcretest program is not installed.
3473
3474 6. pgrep -V now gives the PCRE version number and date.
3475
3476 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3477 causing the entire string to be ignored, instead of just the last character.
3478
3479 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3480 non-matching string, it can take a very, very long time, even for strings of
3481 quite modest length, because of the nested recursion. PCRE now does better in
3482 some of these cases. It does this by remembering the last required literal
3483 character in the pattern, and pre-searching the subject to ensure it is present
3484 before running the real match. In other words, it applies a heuristic to detect
3485 some types of certain failure quickly, and in the above example, if presented
3486 with a string that has no trailing " it gives "no match" very quickly.
3487
3488 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3489 other alternatives are tried instead.
3490
3491
3492 Version 2.06 09-Jun-99
3493 ----------------------
3494
3495 1. Change pcretest's output for amount of store used to show just the code
3496 space, because the remainder (the data block) varies in size between 32-bit and
3497 64-bit systems.
3498
3499 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3500 start matching at. This allows lookbehinds to work when searching for multiple
3501 occurrences in a string.
3502
3503 3. Added additional options to pcretest for testing multiple occurrences:
3504
3505 /+ outputs the rest of the string that follows a match
3506 /g loops for multiple occurrences, using the new startoffset argument
3507 /G loops for multiple occurrences by passing an incremented pointer
3508
3509 4. PCRE wasn't doing the "first character" optimization for patterns starting
3510 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3511 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3512 the letter 'x'. On long subject strings, this gives a significant speed-up.
3513
3514
3515 Version 2.05 21-Apr-99
3516 ----------------------
3517
3518 1. Changed the type of magic_number from int to long int so that it works
3519 properly on 16-bit systems.
3520
3521 2. Fixed a bug which caused patterns starting with .* not to work correctly
3522 when the subject string contained newline characters. PCRE was assuming
3523 anchoring for such patterns in all cases, which is not correct because .* will
3524 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3525 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3526 must be retried after every newline in the subject.
3527
3528
3529 Version 2.04 18-Feb-99
3530 ----------------------
3531
3532 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3533 computation of the store needed to hold the pattern was incorrect (too large).
3534 If such patterns were nested a few deep, this could multiply and become a real
3535 problem.
3536
3537 2. Added /M option to pcretest to show the memory requirement of a specific
3538 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3539
3540 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3541 compiled in such a way that the backtracking after subsequent failure was
3542 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3543 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3544
3545
3546 Version 2.03 02-Feb-99
3547 ----------------------
3548
3549 1. Fixed typo and small mistake in man page.
3550
3551 2. Added 4th condition (GPL supersedes if conflict) and created separate
3552 LICENCE file containing the conditions.
3553
3554 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3555 Perl, that is the internal \ allows the delimiter to be included in the
3556 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3557 the final delimiter, add \ to the end of the pattern (to test the error).
3558
3559 4. Added the convenience functions for extracting substrings after a successful
3560 match. Updated pcretest to make it able to test these functions.
3561
3562
3563 Version 2.02 14-Jan-99
3564 ----------------------
3565
3566 1. Initialized the working variables associated with each extraction so that
3567 their saving and restoring doesn't refer to uninitialized store.
3568
3569 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3570 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3571 fix the problem.
3572
3573 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3574 calls, and wasn't printing the correct value for compiling calls. Increased the
3575 default value of LOOPREPEAT, and the number of significant figures in the
3576 times.
3577
3578 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3579
3580 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3581 a building problem on Windows NT with a FAT file system.
3582
3583
3584 Version 2.01 21-Oct-98
3585 ----------------------
3586
3587 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3588 to character tables built by pcre_maketables() in the current locale. If NULL
3589 is passed, the default tables are used.
3590
3591
3592 Version 2.00 24-Sep-98
3593 ----------------------
3594
3595 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3596 it any more.
3597
3598 2. Allow quantification of (?>) groups, and make it work correctly.
3599
3600 3. The first character computation wasn't working for (?>) groups.
3601
3602 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3603 end of the subject) and add 5.005's \z, which really does match only at the
3604 very end of the subject.
3605
3606 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3607
3608 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3609 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3610 localized options. All options to pcre_study() were also removed.
3611
3612 7. Add other new features from 5.005:
3613
3614 $(?<= positive lookbehind
3615 $(?<! negative lookbehind
3616 (?imsx-imsx) added the unsetting capability
3617 such a setting is global if at outer level; local otherwise
3618 (?imsx-imsx:) non-capturing groups with option setting
3619 (?(cond)re|re) conditional pattern matching
3620
3621 A backreference to itself in a repeated group matches the previous
3622 captured string.
3623
3624 8. General tidying up of studying (both automatic and via "study")
3625 consequential on the addition of new assertions.
3626
3627 9. As in 5.005, unlimited repeated groups that could match an empty substring
3628 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3629 runtime if any iteration does actually match an empty substring.
3630
3631 10. Include the RunTest script in the distribution.
3632
3633 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3634 discrepancies, some of which were old and were also with respect to 5.004. They
3635 have now been fixed.
3636
3637
3638 Version 1.09 28-Apr-98
3639 ----------------------
3640
3641 1. A negated single character class followed by a quantifier with a minimum
3642 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3643 program crashes, or just wrong answers. This did not apply to negated classes
3644 containing more than one character, or to minima other than one.
3645
3646
3647 Version 1.08 27-Mar-98
3648 ----------------------
3649
3650 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3651
3652 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3653 latter must appear before anything that relies on it in the pattern.
3654
3655
3656 Version 1.07 16-Feb-98
3657 ----------------------
3658
3659 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3660 repeat of a potentially empty string).
3661
3662
3663 Version 1.06 23-Jan-98
3664 ----------------------
3665
3666 1. Added Markus Oberhumer's little patches for C++.
3667
3668 2. Literal strings longer than 255 characters were broken.
3669
3670
3671 Version 1.05 23-Dec-97
3672 ----------------------
3673
3674 1. Negated character classes containing more than one character were failing if
3675 PCRE_CASELESS was set at run time.
3676
3677
3678 Version 1.04 19-Dec-97
3679 ----------------------
3680
3681 1. Corrected the man page, where some "const" qualifiers had been omitted.
3682
3683 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3684 input syntax.
3685
3686 3. Fixed memory leak which occurred when a regex with back references was
3687 matched with an offsets vector that wasn't big enough. The temporary memory
3688 that is used in this case wasn't being freed if the match failed.
3689
3690 4. Tidied pcretest to ensure it frees memory that it gets.
3691
3692 5. Temporary memory was being obtained in the case where the passed offsets
3693 vector was exactly big enough.
3694
3695 6. Corrected definition of offsetof() from change 5 below.
3696
3697 7. I had screwed up change 6 below and broken the rules for the use of
3698 setjmp(). Now fixed.
3699
3700
3701 Version 1.03 18-Dec-97
3702 ----------------------
3703
3704 1. A erroneous regex with a missing opening parenthesis was correctly
3705 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3706 on some systems.
3707
3708 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3709 it was reported that one broken compiler failed on the former because "code" is
3710 also an independent variable.
3711
3712 3. The erroneous regex a[]b caused an array overrun reference.
3713
3714 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3715 fail on data ending with that character. (It was going on too far, and checking
3716 the next character, typically a binary zero.) This was specific to the
3717 optimized code for single-character negative classes.
3718
3719 5. Added a contributed patch from the TIN world which does the following:
3720
3721 + Add an undef for memmove, in case the the system defines a macro for it.
3722
3723 + Add a definition of offsetof(), in case there isn't one. (I don't know
3724 the reason behind this - offsetof() is part of the ANSI standard - but
3725 it does no harm).
3726
3727 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3728 most of the places where whitespace preceded '#'. I have given up and
3729 allowed the remaining 2 cases to be at the margin.
3730
3731 + Rename some variables in pcre to eliminate shadowing. This seems very
3732 pedantic, but does no harm, of course.
3733
3734 6. Moved the call to setjmp() into its own function, to get rid of warnings
3735 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3736
3737 7. Constructs such as \d{8,} were compiling into the equivalent of
3738 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3739 outcome, but in this particular case used more store than had been allocated,
3740 which caused the bug to be discovered because it threw up an internal error.
3741
3742 8. The debugging code in both pcre and pcretest for outputting the compiled
3743 form of a regex was going wrong in the case of back references followed by
3744 curly-bracketed repeats.
3745
3746
3747 Version 1.02 12-Dec-97
3748 ----------------------
3749
3750 1. Typos in pcre.3 and comments in the source fixed.
3751
3752 2. Applied a contributed patch to get rid of places where it used to remove
3753 'const' from variables, and fixed some signed/unsigned and uninitialized
3754 variable warnings.
3755
3756 3. Added the "runtest" target to Makefile.
3757
3758 4. Set default compiler flag to -O2 rather than just -O.
3759
3760
3761 Version 1.01 19-Nov-97
3762 ----------------------
3763
3764 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3765 like /([ab]*)*/, that is, for classes with more than one character in them.
3766
3767 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3768 as /((?>a*))*/ (a PCRE_EXTRA facility).
3769
3770
3771 Version 1.00 18-Nov-97
3772 ----------------------
3773
3774 1. Added compile-time macros to support systems such as SunOS4 which don't have
3775 memmove() or strerror() but have other things that can be used instead.
3776
3777 2. Arranged that "make clean" removes the executables.
3778
3779
3780 Version 0.99 27-Oct-97
3781 ----------------------
3782
3783 1. Fixed bug in code for optimizing classes with only one character. It was
3784 initializing a 32-byte map regardless, which could cause it to run off the end
3785 of the memory it had got.
3786
3787 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3788
3789
3790 Version 0.98 22-Oct-97
3791 ----------------------
3792
3793 1. Fixed bug in code for handling temporary memory usage when there are more
3794 back references than supplied space in the ovector. This could cause segfaults.
3795
3796
3797 Version 0.97 21-Oct-97
3798 ----------------------
3799
3800 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3801
3802 2. Optimized negated single characters not to use a bit map.
3803
3804 3. Brought error texts together as macro definitions; clarified some of them;
3805 fixed one that was wrong - it said "range out of order" when it meant "invalid
3806 escape sequence".
3807
3808 4. Changed some char * arguments to const char *.
3809
3810 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3811
3812 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3813 pcretest.
3814
3815
3816 Version 0.96 16-Oct-97
3817 ----------------------
3818
3819 1. Added a simple "pgrep" utility to the distribution.
3820
3821 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3822 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3823 where "ddd" means "one or more decimal digits".
3824
3825 3. Fixed serious bug. If a pattern had a back reference, but the call to
3826 pcre_exec() didn't supply a large enough ovector to record the related
3827 identifying subpattern, the match always failed. PCRE now remembers the number
3828 of the largest back reference, and gets some temporary memory in which to save
3829 the offsets during matching if necessary, in order to ensure that
3830 backreferences always work.
3831
3832 4. Increased the compatibility with Perl in a number of ways:
3833
3834 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3835 to request this handling. The option can be set at compile or exec time.
3836
3837 (b) $ matches before a terminating newline by default; an option
3838 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3839 mode). The option can be set at compile or exec time.
3840
3841 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3842 the same as Perl's. If the decimal number it represents is less than 10
3843 or there aren't that many previous left capturing parentheses, an octal
3844 escape is read. Inside a character class, it's always an octal escape,
3845 even if it is a single digit.
3846
3847 (d) An escaped but undefined alphabetic character is taken as a literal,
3848 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3849 escapes.
3850
3851 (e) {0} is now permitted. (The previous item is removed from the compiled
3852 pattern).
3853
3854 5. Changed all the names of code files so that the basic parts are no longer
3855 than 10 characters, and abolished the teeny "globals.c" file.
3856
3857 6. Changed the handling of character classes; they are now done with a 32-byte
3858 bit map always.
3859
3860 7. Added the -d and /D options to pcretest to make it possible to look at the
3861 internals of compilation without having to recompile pcre.
3862
3863
3864 Version 0.95 23-Sep-97
3865 ----------------------
3866
3867 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3868 \x20 at the start of a run of normal characters. These were being treated as
3869 real characters, instead of the source characters being re-checked.
3870
3871
3872 Version 0.94 18-Sep-97
3873 ----------------------
3874
3875 1. The functions are now thread-safe, with the caveat that the global variables
3876 containing pointers to malloc() and free() or alternative functions are the
3877 same for all threads.
3878
3879 2. Get pcre_study() to generate a bitmap of initial characters for non-
3880 anchored patterns when this is possible, and use it if passed to pcre_exec().
3881
3882
3883 Version 0.93 15-Sep-97
3884 ----------------------
3885
3886 1. /(b)|(:+)/ was computing an incorrect first character.
3887
3888 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3889 but not actually doing anything yet.
3890
3891 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3892 as Perl does (e.g. [-az] or [az-]).
3893
3894 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3895 all possible positions.
3896
3897 5. Split up into different modules to avoid including unneeded functions in a
3898 compiled binary. However, compile and exec are still in one module. The "study"
3899 function is split off.
3900
3901 6. The character tables are now in a separate module whose source is generated
3902 by an auxiliary program - but can then be edited by hand if required. There are
3903 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3904 toupper() in the code.
3905
3906 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3907 make them global. Abolish the function for setting them, as the caller can now
3908 set them directly.
3909
3910
3911 Version 0.92 11-Sep-97
3912 ----------------------
3913
3914 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3915 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3916
3917 2. Caseless matching was not working in character classes if the characters in
3918 the pattern were in upper case.
3919
3920 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3921
3922 4. Make PCRE_ANCHORED public and accept as a compile option.
3923
3924 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3925 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3926 pass them.
3927
3928 6. Give an error if bad option bits passed at compile or run time.
3929
3930 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3931 pcretest to cause it to pass that flag.
3932
3933 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3934 options, and the first character, if set.
3935
3936 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3937
3938
3939 Version 0.91 10-Sep-97
3940 ----------------------
3941
3942 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3943 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3944
3945 2. PCRE was looping on encountering an indefinitely repeated back reference to
3946 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3947 Perl does - treats the match as successful.
3948
3949 ****

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