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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 530 - (show annotations) (download)
Tue Jun 1 13:42:06 2010 UTC (3 years, 10 months ago) by ph10
File size: 188064 byte(s)
Added a lot of (int) casts to avoid compiler warnings in systems where      
size_t is 64-bit.

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

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