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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 521 - (show annotations)
Sat May 22 19:03:35 2010 UTC (5 years, 3 months ago) by ph10
File size: 187086 byte(s)
Error occurred while calculating annotation data.
Remove explicit include of pcre.h from pcreposix.c because pcre_internal.h 
already includes it. This avoids a PCRE_EXP_DECL problem.
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
53
54 Version 8.02 19-Mar-2010
55 ------------------------
56
57 1. The Unicode data tables have been updated to Unicode 5.2.0.
58
59 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
60 configured.
61
62 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
63 original author of that file, following a query about its status.
64
65 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
66 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
67
68 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
69 quantifier applied to a forward-referencing subroutine call, could compile
70 incorrect code or give the error "internal error: previously-checked
71 referenced subpattern not found".
72
73 6. Both MS Visual Studio and Symbian OS have problems with initializing
74 variables to point to external functions. For these systems, therefore,
75 pcre_malloc etc. are now initialized to local functions that call the
76 relevant global functions.
77
78 7. There were two entries missing in the vectors called coptable and poptable
79 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
80 I've fixed the data, and added a kludgy way of testing at compile time that
81 the lengths are correct (equal to the number of opcodes).
82
83 8. Following on from 7, I added a similar kludge to check the length of the
84 eint vector in pcreposix.c.
85
86 9. Error texts for pcre_compile() are held as one long string to avoid too
87 much relocation at load time. To find a text, the string is searched,
88 counting zeros. There was no check for running off the end of the string,
89 which could happen if a new error number was added without updating the
90 string.
91
92 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
93
94 11. \K was not working if it appeared in an atomic group or in a group that
95 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
96 \K is "not well defined" if used in an assertion. PCRE now accepts it if
97 the assertion is positive, but not if it is negative.
98
99 12. Change 11 fortuitously reduced the size of the stack frame used in the
100 "match()" function of pcre_exec.c by one pointer. Forthcoming
101 implementation of support for (*MARK) will need an extra pointer on the
102 stack; I have reserved it now, so that the stack frame size does not
103 decrease.
104
105 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
106 item in branch that calls a recursion is a subroutine call - as in the
107 second branch in the above example - was incorrectly given the compile-
108 time error "recursive call could loop indefinitely" because pcre_compile()
109 was not correctly checking the subroutine for matching a non-empty string.
110
111 14. The checks for overrunning compiling workspace could trigger after an
112 overrun had occurred. This is a "should never occur" error, but it can be
113 triggered by pathological patterns such as hundreds of nested parentheses.
114 The checks now trigger 100 bytes before the end of the workspace.
115
116 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
117
118
119 Version 8.01 19-Jan-2010
120 ------------------------
121
122 1. If a pattern contained a conditional subpattern with only one branch (in
123 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
124 computed the wrong minimum data length (which is of course zero for such
125 subpatterns). This could cause incorrect "no match" results.
126
127 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
128 the pattern is reset in the first branch, pcre_compile() failed with
129 "internal error: code overflow at offset...". This happened only when
130 the reset was to the original external option setting. (An optimization
131 abstracts leading options settings into an external setting, which was the
132 cause of this.)
133
134 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
135 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
136 assertion pattern did not match (meaning that the assertion was true), it
137 was incorrectly treated as false if the SKIP had been reached during the
138 matching. This also applied to assertions used as conditions.
139
140 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
141 assertion subpattern, including such a pattern used as a condition,
142 unpredictable results occurred, instead of the error return
143 PCRE_ERROR_DFA_UITEM.
144
145 5. The C++ GlobalReplace function was not working like Perl for the special
146 situation when an empty string is matched. It now does the fancy magic
147 stuff that is necessary.
148
149 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
150 removed. (These were left over from very, very early versions of PCRE.)
151
152 7. Some cosmetic changes to the code to make life easier when compiling it
153 as part of something else:
154
155 (a) Change DEBUG to PCRE_DEBUG.
156
157 (b) In pcre_compile(), rename the member of the "branch_chain" structure
158 called "current" as "current_branch", to prevent a collision with the
159 Linux macro when compiled as a kernel module.
160
161 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
162 prevent a collision with the Linux macro when compiled as a kernel
163 module.
164
165 8. In pcre_compile() there are some checks for integer overflows that used to
166 cast potentially large values to (double). This has been changed to that
167 when building, a check for int64_t is made, and if it is found, it is used
168 instead, thus avoiding the use of floating point arithmetic. (There is no
169 other use of FP in PCRE.) If int64_t is not found, the fallback is to
170 double.
171
172 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
173 2005 (difference between two addresses compared to an unsigned value).
174
175 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
176 custom one, because of the following reported problem in Windows:
177
178 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
179 under Win32.
180 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
181 therefore missing the function definition.
182 - The compiler thus generates a "C" signature for the test function.
183 - The linker fails to find the "C" function.
184 - PCRE fails to configure if asked to do so against libbz2.
185
186 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
187 messages were output:
188
189 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
190 rerunning libtoolize, to keep the correct libtool macros in-tree.
191 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
192
193 I have done both of these things.
194
195 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
196 most of the time, it *can* run out if it is given a pattern that contains a
197 runaway infinite recursion. I updated the discussion in the pcrestack man
198 page.
199
200 13. Now that we have gone to the x.xx style of version numbers, the minor
201 version may start with zero. Using 08 or 09 is a bad idea because users
202 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
203 interpreted as invalid octal numbers. I've updated the previous comment in
204 configure.ac, and also added a check that gives an error if 08 or 09 are
205 used.
206
207 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
208 causing partial matching to fail when the end of the subject matched \W
209 in a UTF-8 pattern where \W was quantified with a minimum of 3.
210
211 15. There were some discrepancies between the declarations in pcre_internal.h
212 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
213 their definitions. The declarations used "const uschar *" and the
214 definitions used USPTR. Even though USPTR is normally defined as "const
215 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
216 reported that: "This difference in casting confuses some C++ compilers, for
217 example, SunCC recognizes above declarations as different functions and
218 generates broken code for hbpcre." I have changed the declarations to use
219 USPTR.
220
221 16. GNU libtool is named differently on some systems. The autogen.sh script now
222 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
223 (FreeBSD).
224
225 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
226 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
227 comment: "Figure out how to create a longlong from a string: strtoll and
228 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
229 instance, but it only takes 2 args instead of 3!"
230
231 18. A subtle bug concerned with back references has been fixed by a change of
232 specification, with a corresponding code fix. A pattern such as
233 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
234 refers, was giving matches when it shouldn't. For example, xa=xaaa would
235 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
236 same bug. Such groups have to be quantified to be useful, or contained
237 inside another quantified group. (If there's no repetition, the reference
238 can never match.) The problem arises because, having left the group and
239 moved on to the rest of the pattern, a later failure that backtracks into
240 the group uses the captured value from the final iteration of the group
241 rather than the correct earlier one. I have fixed this in PCRE by forcing
242 any group that contains a reference to itself to be an atomic group; that
243 is, there cannot be any backtracking into it once it has completed. This is
244 similar to recursive and subroutine calls.
245
246
247 Version 8.00 19-Oct-09
248 ----------------------
249
250 1. The table for translating pcre_compile() error codes into POSIX error codes
251 was out-of-date, and there was no check on the pcre_compile() error code
252 being within the table. This could lead to an OK return being given in
253 error.
254
255 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
256 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
257 in a Windows environment.
258
259 3. The pcregrep --count option prints the count for each file even when it is
260 zero, as does GNU grep. However, pcregrep was also printing all files when
261 --files-with-matches was added. Now, when both options are given, it prints
262 counts only for those files that have at least one match. (GNU grep just
263 prints the file name in this circumstance, but including the count seems
264 more useful - otherwise, why use --count?) Also ensured that the
265 combination -clh just lists non-zero counts, with no names.
266
267 4. The long form of the pcregrep -F option was incorrectly implemented as
268 --fixed_strings instead of --fixed-strings. This is an incompatible change,
269 but it seems right to fix it, and I didn't think it was worth preserving
270 the old behaviour.
271
272 5. The command line items --regex=pattern and --regexp=pattern were not
273 recognized by pcregrep, which required --regex pattern or --regexp pattern
274 (with a space rather than an '='). The man page documented the '=' forms,
275 which are compatible with GNU grep; these now work.
276
277 6. No libpcreposix.pc file was created for pkg-config; there was just
278 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
279
280 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
281 when UCP support is not needed, by modifying the Python script that
282 generates it from Unicode data files. This should not matter if the module
283 is correctly used as a library, but I received one complaint about 50K of
284 unwanted data. My guess is that the person linked everything into his
285 program rather than using a library. Anyway, it does no harm.
286
287 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
288 was a minimum greater than 1 for a wide character in a possessive
289 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
290 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
291 character. Chaos in the form of incorrect output or a compiling loop could
292 result.
293
294 9. The restrictions on what a pattern can contain when partial matching is
295 requested for pcre_exec() have been removed. All patterns can now be
296 partially matched by this function. In addition, if there are at least two
297 slots in the offset vector, the offset of the earliest inspected character
298 for the match and the offset of the end of the subject are set in them when
299 PCRE_ERROR_PARTIAL is returned.
300
301 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
302 synonymous with PCRE_PARTIAL, for backwards compatibility, and
303 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
304 and may be more useful for multi-segment matching.
305
306 11. Partial matching with pcre_exec() is now more intuitive. A partial match
307 used to be given if ever the end of the subject was reached; now it is
308 given only if matching could not proceed because another character was
309 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
310 string "Z", which now yields "no match" instead of "partial match". In the
311 case of pcre_dfa_exec(), "no match" is given if every matching path for the
312 final character ended with (*FAIL).
313
314 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
315 if the pattern had a "must contain" character that was already found in the
316 earlier partial match, unless partial matching was again requested. For
317 example, with the pattern /dog.(body)?/, the "must contain" character is
318 "g". If the first part-match was for the string "dog", restarting with
319 "sbody" failed. This bug has been fixed.
320
321 13. The string returned by pcre_dfa_exec() after a partial match has been
322 changed so that it starts at the first inspected character rather than the
323 first character of the match. This makes a difference only if the pattern
324 starts with a lookbehind assertion or \b or \B (\K is not supported by
325 pcre_dfa_exec()). It's an incompatible change, but it makes the two
326 matching functions compatible, and I think it's the right thing to do.
327
328 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
329 so that the demonstration program is easily available in environments where
330 PCRE has not been installed from source.
331
332 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
333 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
334 library.
335
336 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
337 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
338 is not the first non-POSIX option to be added. Clearly some people find
339 these options useful.
340
341 17. If a caller to the POSIX matching function regexec() passes a non-zero
342 value for nmatch with a NULL value for pmatch, the value of
343 nmatch is forced to zero.
344
345 18. RunGrepTest did not have a test for the availability of the -u option of
346 the diff command, as RunTest does. It now checks in the same way as
347 RunTest, and also checks for the -b option.
348
349 19. If an odd number of negated classes containing just a single character
350 interposed, within parentheses, between a forward reference to a named
351 subpattern and the definition of the subpattern, compilation crashed with
352 an internal error, complaining that it could not find the referenced
353 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
354 [The bug was that it was starting one character too far in when skipping
355 over the character class, thus treating the ] as data rather than
356 terminating the class. This meant it could skip too much.]
357
358 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
359 /g option in pcretest when the pattern contains \K, which makes it possible
360 to have an empty string match not at the start, even when the pattern is
361 anchored. Updated pcretest and pcredemo to use this option.
362
363 21. If the maximum number of capturing subpatterns in a recursion was greater
364 than the maximum at the outer level, the higher number was returned, but
365 with unset values at the outer level. The correct (outer level) value is
366 now given.
367
368 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
369 PCRE did not set those parentheses (unlike Perl). I have now found a way to
370 make it do so. The string so far is captured, making this feature
371 compatible with Perl.
372
373 23. The tests have been re-organized, adding tests 11 and 12, to make it
374 possible to check the Perl 5.10 features against Perl 5.10.
375
376 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
377 pattern matches a fixed length string. PCRE did not allow this; now it
378 does. Neither allows recursion.
379
380 25. I finally figured out how to implement a request to provide the minimum
381 length of subject string that was needed in order to match a given pattern.
382 (It was back references and recursion that I had previously got hung up
383 on.) This code has now been added to pcre_study(); it finds a lower bound
384 to the length of subject needed. It is not necessarily the greatest lower
385 bound, but using it to avoid searching strings that are too short does give
386 some useful speed-ups. The value is available to calling programs via
387 pcre_fullinfo().
388
389 26. While implementing 25, I discovered to my embarrassment that pcretest had
390 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
391 study optimizations had never been tested with that matching function.
392 Oops. What is worse, even when it was passed study data, there was a bug in
393 pcre_dfa_exec() that meant it never actually used it. Double oops. There
394 were also very few tests of studied patterns with pcre_dfa_exec().
395
396 27. If (?| is used to create subpatterns with duplicate numbers, they are now
397 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
398 on the other side of the coin, they are no longer allowed to have different
399 names, because these cannot be distinguished in PCRE, and this has caused
400 confusion. (This is a difference from Perl.)
401
402 28. When duplicate subpattern names are present (necessarily with different
403 numbers, as required by 27 above), and a test is made by name in a
404 conditional pattern, either for a subpattern having been matched, or for
405 recursion in such a pattern, all the associated numbered subpatterns are
406 tested, and the overall condition is true if the condition is true for any
407 one of them. This is the way Perl works, and is also more like the way
408 testing by number works.
409
410
411 Version 7.9 11-Apr-09
412 ---------------------
413
414 1. When building with support for bzlib/zlib (pcregrep) and/or readline
415 (pcretest), all targets were linked against these libraries. This included
416 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
417 libraries. This caused unwanted dependencies to be created. This problem
418 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
419 pcretest is linked with readline.
420
421 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
422 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
423 moved outside it again, because FALSE and TRUE are already defined in AIX,
424 but BOOL is not.
425
426 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
427 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
428
429 4. The pcregrep documentation said spaces were inserted as well as colons (or
430 hyphens) following file names and line numbers when outputting matching
431 lines. This is not true; no spaces are inserted. I have also clarified the
432 wording for the --colour (or --color) option.
433
434 5. In pcregrep, when --colour was used with -o, the list of matching strings
435 was not coloured; this is different to GNU grep, so I have changed it to be
436 the same.
437
438 6. When --colo(u)r was used in pcregrep, only the first matching substring in
439 each matching line was coloured. Now it goes on to look for further matches
440 of any of the test patterns, which is the same behaviour as GNU grep.
441
442 7. A pattern that could match an empty string could cause pcregrep to loop; it
443 doesn't make sense to accept an empty string match in pcregrep, so I have
444 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
445 seems to be how GNU grep behaves.
446
447 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
448 start or after a newline", because the conditional assertion was not being
449 correctly handled. The rule now is that both the assertion and what follows
450 in the first alternative must satisfy the test.
451
452 9. If auto-callout was enabled in a pattern with a conditional group whose
453 condition was an assertion, PCRE could crash during matching, both with
454 pcre_exec() and pcre_dfa_exec().
455
456 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
457 used for matching.
458
459 11. Unicode property support in character classes was not working for
460 characters (bytes) greater than 127 when not in UTF-8 mode.
461
462 12. Added the -M command line option to pcretest.
463
464 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
465
466 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
467
468 16. Added comments and documentation about mis-use of no_arg in the C++
469 wrapper.
470
471 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
472 from Martin Jerabek that uses macro names for all relevant character and
473 string constants.
474
475 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
476 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
477 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
478 these, but not everybody uses configure.
479
480 19. A conditional group that had only one branch was not being correctly
481 recognized as an item that could match an empty string. This meant that an
482 enclosing group might also not be so recognized, causing infinite looping
483 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
484 with the subject "ab", where knowledge that the repeated group can match
485 nothing is needed in order to break the loop.
486
487 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
488 exec(), but without supplying a callout function, matching went wrong.
489
490 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
491 leak if the size of the offset vector was greater than 30. When the vector
492 is smaller, the saved offsets during recursion go onto a local stack
493 vector, but for larger vectors malloc() is used. It was failing to free
494 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
495 error, in fact).
496
497 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
498 heapframe that is used only when UTF-8 support is enabled. This caused no
499 problem, but was untidy.
500
501 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
502 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
503 included within another project.
504
505 24. Steven Van Ingelgem's patches to add more options to the CMake support,
506 slightly modified by me:
507
508 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
509 not building pcregrep.
510
511 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
512 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
513
514 25. Forward references, both numeric and by name, in patterns that made use of
515 duplicate group numbers, could behave incorrectly or give incorrect errors,
516 because when scanning forward to find the reference group, PCRE was not
517 taking into account the duplicate group numbers. A pattern such as
518 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
519
520 26. Changed a few more instances of "const unsigned char *" to USPTR, making
521 the feature of a custom pointer more persuasive (as requested by a user).
522
523 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
524 pcretest.c, inside #ifndefs, because it seems they are sometimes already
525 pre-defined.
526
527 28. Added support for (*UTF8) at the start of a pattern.
528
529 29. Arrange for flags added by the "release type" setting in CMake to be shown
530 in the configuration summary.
531
532
533 Version 7.8 05-Sep-08
534 ---------------------
535
536 1. Replaced UCP searching code with optimized version as implemented for Ad
537 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
538 stage table and inline lookup instead of a function, giving speed ups of 2
539 to 5 times on some simple patterns that I tested. Permission was given to
540 distribute the MultiStage2.py script that generates the tables (it's not in
541 the tarball, but is in the Subversion repository).
542
543 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
544 scripts.
545
546 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
547 a group with a zero qualifier. The result of the study could be incorrect,
548 or the function might crash, depending on the pattern.
549
550 4. Caseless matching was not working for non-ASCII characters in back
551 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
552 It now works when Unicode Property Support is available.
553
554 5. In pcretest, an escape such as \x{de} in the data was always generating
555 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
556 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
557 truncation.
558
559 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
560
561 7. Added two (int) casts to pcregrep when printing the difference of two
562 pointers, in case they are 64-bit values.
563
564 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
565 test 2 if it fails.
566
567 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
568 and a #define of that name to empty if it is not externally set. This is to
569 allow users of MSVC to set it if necessary.
570
571 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
572 the convenience functions in the pcre_get.c source file.
573
574 11. An option change at the start of a pattern that had top-level alternatives
575 could cause overwriting and/or a crash. This command provoked a crash in
576 some environments:
577
578 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
579
580 This potential security problem was recorded as CVE-2008-2371.
581
582 12. For a pattern where the match had to start at the beginning or immediately
583 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
584 pcre_dfa_exec() could read past the end of the passed subject if there was
585 no match. To help with detecting such bugs (e.g. with valgrind), I modified
586 pcretest so that it places the subject at the end of its malloc-ed buffer.
587
588 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
589 exec() might read past the end of the data buffer in UTF-8 mode.
590
591 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
592 the data contained the byte 0x85 as part of a UTF-8 character within its
593 first line. This applied both to normal and DFA matching.
594
595 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
596 /^[^d]*?$/8 failed to match "abc".
597
598 16. Added a missing copyright notice to pcrecpp_internal.h.
599
600 17. Make it more clear in the documentation that values returned from
601 pcre_exec() in ovector are byte offsets, not character counts.
602
603 18. Tidied a few places to stop certain compilers from issuing warnings.
604
605 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
606 supplied by Stefan Weber. I made a further small update for 7.8 because
607 there is a change of source arrangements: the pcre_searchfuncs.c module is
608 replaced by pcre_ucd.c.
609
610
611 Version 7.7 07-May-08
612 ---------------------
613
614 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
615 a string to a long long, pretend we don't even have a long long." This is
616 done by checking for the strtoq, strtoll, and _strtoi64 functions.
617
618 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
619 pre-7.6 versions, which defined a global no_arg variable instead of putting
620 it in the RE class. (See also #8 below.)
621
622 3. Remove a line of dead code, identified by coverity and reported by Nuno
623 Lopes.
624
625 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
626
627 (1) The include/exclude patterns were being applied to the whole pathnames
628 of files, instead of just to the final components.
629
630 (2) If there was more than one level of directory, the subdirectories were
631 skipped unless they satisfied the include/exclude conditions. This is
632 inconsistent with GNU grep (and could even be seen as contrary to the
633 pcregrep specification - which I improved to make it absolutely clear).
634 The action now is always to scan all levels of directory, and just
635 apply the include/exclude patterns to regular files.
636
637 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
638 --exclude_dir in the tests to avoid scanning .svn directories.
639
640 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
641 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
642 doesn't support NULs in patterns.
643
644 7. Added some missing "const"s to declarations of static tables in
645 pcre_compile.c and pcre_dfa_exec.c.
646
647 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
648 caused by fix #2 above. (Subsequently also a second patch to fix the
649 first patch. And a third patch - this was a messy problem.)
650
651 9. Applied Craig's patch to remove the use of push_back().
652
653 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
654 matching function regexec().
655
656 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
657 which, however, unlike Perl's \g{...}, are subroutine calls, not back
658 references. PCRE supports relative numbers with this syntax (I don't think
659 Oniguruma does).
660
661 12. Previously, a group with a zero repeat such as (...){0} was completely
662 omitted from the compiled regex. However, this means that if the group
663 was called as a subroutine from elsewhere in the pattern, things went wrong
664 (an internal error was given). Such groups are now left in the compiled
665 pattern, with a new opcode that causes them to be skipped at execution
666 time.
667
668 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
669 to the way PCRE behaves:
670
671 (a) A lone ] character is dis-allowed (Perl treats it as data).
672
673 (b) A back reference to an unmatched subpattern matches an empty string
674 (Perl fails the current match path).
675
676 (c) A data ] in a character class must be notated as \] because if the
677 first data character in a class is ], it defines an empty class. (In
678 Perl it is not possible to have an empty class.) The empty class []
679 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
680 The negative empty class [^] matches any one character, independently
681 of the DOTALL setting.
682
683 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
684 non-existent subpattern following a character class starting with ']' and
685 containing () gave an internal compiling error instead of "reference to
686 non-existent subpattern". Fortunately, when the pattern did exist, the
687 compiled code was correct. (When scanning forwards to check for the
688 existencd of the subpattern, it was treating the data ']' as terminating
689 the class, so got the count wrong. When actually compiling, the reference
690 was subsequently set up correctly.)
691
692 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
693 it was being rejected as not supported by pcre_dfa_exec(), even though
694 other assertions are supported. I have made pcre_dfa_exec() support
695 (*FAIL).
696
697 16. The implementation of 13c above involved the invention of a new opcode,
698 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
699 cannot be changed at match time, I realized I could make a small
700 improvement to matching performance by compiling OP_ALLANY instead of
701 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
702 on the OP_ANY path.
703
704 17. Compiling pcretest on Windows with readline support failed without the
705 following two fixes: (1) Make the unistd.h include conditional on
706 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
707
708 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
709 ncurses library to be included for pcretest when ReadLine support is
710 requested, but also to allow for it to be overridden. This patch came from
711 Daniel Bergström.
712
713 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
714 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
715 any errors with the current Unicode tables. Thanks to Peter Kankowski for
716 spotting this.
717
718
719 Version 7.6 28-Jan-08
720 ---------------------
721
722 1. A character class containing a very large number of characters with
723 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
724 overflow.
725
726 2. Patch to cut out the "long long" test in pcrecpp_unittest when
727 HAVE_LONG_LONG is not defined.
728
729 3. Applied Christian Ehrlicher's patch to update the CMake build files to
730 bring them up to date and include new features. This patch includes:
731
732 - Fixed PH's badly added libz and libbz2 support.
733 - Fixed a problem with static linking.
734 - Added pcredemo. [But later removed - see 7 below.]
735 - Fixed dftables problem and added an option.
736 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
737 HAVE_LONG_LONG.
738 - Added readline support for pcretest.
739 - Added an listing of the option settings after cmake has run.
740
741 4. A user submitted a patch to Makefile that makes it easy to create
742 "pcre.dll" under mingw when using Configure/Make. I added stuff to
743 Makefile.am that cause it to include this special target, without
744 affecting anything else. Note that the same mingw target plus all
745 the other distribution libraries and programs are now supported
746 when configuring with CMake (see 6 below) instead of with
747 Configure/Make.
748
749 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
750 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
751 exported in the Windows port". It has not yet been confirmed that the patch
752 solves the problem, but it does no harm.
753
754 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
755 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
756 with CMake, and also correct the comment about stack recursion.
757
758 7. Remove the automatic building of pcredemo from the ./configure system and
759 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
760 of a program that users should build themselves after PCRE is installed, so
761 building it automatically is not really right. What is more, it gave
762 trouble in some build environments.
763
764 8. Further tidies to CMakeLists.txt from Sheri and Christian.
765
766
767 Version 7.5 10-Jan-08
768 ---------------------
769
770 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
771 values in parens when parsing an RE using the C++ wrapper."
772
773 2. Negative specials like \S did not work in character classes in UTF-8 mode.
774 Characters greater than 255 were excluded from the class instead of being
775 included.
776
777 3. The same bug as (2) above applied to negated POSIX classes such as
778 [:^space:].
779
780 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
781 defined or documented. It seems to have been a typo for PCRE_STATIC, so
782 I have changed it.
783
784 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
785 first named subpattern) and a construct such as (?&a) would reference the
786 first named subpattern whose name started with "a" (in other words, the
787 length check was missing). Both these problems are fixed. "Subpattern name
788 expected" is now given for (?&) (a zero-length name), and this patch also
789 makes it give the same error for \k'' (previously it complained that that
790 was a reference to a non-existent subpattern).
791
792 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
793 this is right because (?- can be followed by option settings as well as by
794 digits. I have, however, made the messages clearer.
795
796 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
797 than the number used in the conditional) now cause a compile-time error.
798 This is actually not compatible with Perl, which accepts such patterns, but
799 treats the conditional as always being FALSE (as PCRE used to), but it
800 seems to me that giving a diagnostic is better.
801
802 8. Change "alphameric" to the more common word "alphanumeric" in comments
803 and messages.
804
805 9. Fix two occurrences of "backslash" in comments that should have been
806 "backspace".
807
808 10. Remove two redundant lines of code that can never be obeyed (their function
809 was moved elsewhere).
810
811 11. The program that makes PCRE's Unicode character property table had a bug
812 which caused it to generate incorrect table entries for sequences of
813 characters that have the same character type, but are in different scripts.
814 It amalgamated them into a single range, with the script of the first of
815 them. In other words, some characters were in the wrong script. There were
816 thirteen such cases, affecting characters in the following ranges:
817
818 U+002b0 - U+002c1
819 U+0060c - U+0060d
820 U+0061e - U+00612
821 U+0064b - U+0065e
822 U+0074d - U+0076d
823 U+01800 - U+01805
824 U+01d00 - U+01d77
825 U+01d9b - U+01dbf
826 U+0200b - U+0200f
827 U+030fc - U+030fe
828 U+03260 - U+0327f
829 U+0fb46 - U+0fbb1
830 U+10450 - U+1049d
831
832 12. The -o option (show only the matching part of a line) for pcregrep was not
833 compatible with GNU grep in that, if there was more than one match in a
834 line, it showed only the first of them. It now behaves in the same way as
835 GNU grep.
836
837 13. If the -o and -v options were combined for pcregrep, it printed a blank
838 line for every non-matching line. GNU grep prints nothing, and pcregrep now
839 does the same. The return code can be used to tell if there were any
840 non-matching lines.
841
842 14. Added --file-offsets and --line-offsets to pcregrep.
843
844 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
845 infinitely looping recursion. The bug was that positive lookaheads were not
846 being skipped when checking for a possible empty match (negative lookaheads
847 and both kinds of lookbehind were skipped).
848
849 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
850 inclusion of <windows.h> to before rather than after the definition of
851 INVALID_FILE_ATTRIBUTES (patch from David Byron).
852
853 17. Specifying a possessive quantifier with a specific limit for a Unicode
854 character property caused pcre_compile() to compile bad code, which led at
855 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
856 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
857 caused the error; without that there was no problem.
858
859 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
860
861 19. Added --enable-pcretest-libreadline.
862
863 20. In pcrecpp.cc, the variable 'count' was incremented twice in
864 RE::GlobalReplace(). As a result, the number of replacements returned was
865 double what it should be. I removed one of the increments, but Craig sent a
866 later patch that removed the other one (the right fix) and added unit tests
867 that check the return values (which was not done before).
868
869 21. Several CMake things:
870
871 (1) Arranged that, when cmake is used on Unix, the libraries end up with
872 the names libpcre and libpcreposix, not just pcre and pcreposix.
873
874 (2) The above change means that pcretest and pcregrep are now correctly
875 linked with the newly-built libraries, not previously installed ones.
876
877 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
878
879 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
880 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
881 UTF-8 newline character). The key issue is that the pattern starts .*;
882 this means that the match must be either at the beginning, or after a
883 newline. The bug was in the code for advancing after a failed match and
884 checking that the new position followed a newline. It was not taking
885 account of UTF-8 characters correctly.
886
887 23. PCRE was behaving differently from Perl in the way it recognized POSIX
888 character classes. PCRE was not treating the sequence [:...:] as a
889 character class unless the ... were all letters. Perl, however, seems to
890 allow any characters between [: and :], though of course it rejects as
891 unknown any "names" that contain non-letters, because all the known class
892 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
893 for example, whereas PCRE did not - it did not recognize a POSIX character
894 class. This seemed a bit dangerous, so the code has been changed to be
895 closer to Perl. The behaviour is not identical to Perl, because PCRE will
896 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
897 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
898 Perl does, and where it didn't before.
899
900 24. Rewrite so as to remove the single use of %n from pcregrep because in some
901 Windows environments %n is disabled by default.
902
903
904 Version 7.4 21-Sep-07
905 ---------------------
906
907 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
908 means that a class such as [\s] counted as "explicit reference to CR or
909 LF". That isn't really right - the whole point of the change was to try to
910 help when there was an actual mention of one of the two characters. So now
911 the change happens only if \r or \n (or a literal CR or LF) character is
912 encountered.
913
914 2. The 32-bit options word was also used for 6 internal flags, but the numbers
915 of both had grown to the point where there were only 3 bits left.
916 Fortunately, there was spare space in the data structure, and so I have
917 moved the internal flags into a new 16-bit field to free up more option
918 bits.
919
920 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
921 but did not set the internal JCHANGED flag - either of these is enough to
922 control the way the "get" function works - but the PCRE_INFO_JCHANGED
923 facility is supposed to tell if (?J) was ever used, so now (?J) at the
924 start sets both bits.
925
926 4. Added options (at build time, compile time, exec time) to change \R from
927 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
928
929 5. doc/pcresyntax.html was missing from the distribution.
930
931 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
932 compatibility, even though it is no longer used.
933
934 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
935 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
936 windows.h file is present (where different names are used). [This was
937 reversed later after testing - see 16 below.]
938
939 8. Changed all #include <config.h> to #include "config.h". There were also
940 some further <pcre.h> cases that I changed to "pcre.h".
941
942 9. When pcregrep was used with the --colour option, it missed the line ending
943 sequence off the lines that it output.
944
945 10. It was pointed out to me that arrays of string pointers cause lots of
946 relocations when a shared library is dynamically loaded. A technique of
947 using a single long string with a table of offsets can drastically reduce
948 these. I have refactored PCRE in four places to do this. The result is
949 dramatic:
950
951 Originally: 290
952 After changing UCP table: 187
953 After changing error message table: 43
954 After changing table of "verbs" 36
955 After changing table of Posix names 22
956
957 Thanks to the folks working on Gregex for glib for this insight.
958
959 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
960 unicode-properties was also set.
961
962 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
963
964 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
965 checked only for CRLF.
966
967 14. Added casts to pcretest.c to avoid compiler warnings.
968
969 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
970
971 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
972 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
973 entirely. This removes changes made in 7 above.
974
975 17. The CMake files have been updated, and there is now more information about
976 building with CMake in the NON-UNIX-USE document.
977
978
979 Version 7.3 28-Aug-07
980 ---------------------
981
982 1. In the rejigging of the build system that eventually resulted in 7.1, the
983 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
984 brackets there is not right, since it causes compilers to look for an
985 installed pcre.h, not the version that is in the source that is being
986 compiled (which of course may be different). I have changed it back to:
987
988 #include "pcre.h"
989
990 I have a vague recollection that the change was concerned with compiling in
991 different directories, but in the new build system, that is taken care of
992 by the VPATH setting the Makefile.
993
994 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
995 when the subject happened to end in the byte 0x85 (e.g. if the last
996 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
997 characters but of course it shouldn't be taken as a newline when it is part
998 of another character. The bug was that, for an unlimited repeat of . in
999 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1000 characters when looking for a newline.
1001
1002 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1003
1004 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1005 in debug output.
1006
1007 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1008 long printing in the pcrecpp unittest when running under MinGW.
1009
1010 6. ESC_K was left out of the EBCDIC table.
1011
1012 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1013 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1014 limit also applies to "virtual nesting" when a pattern is recursive, and in
1015 this case 1000 isn't so big. I have been able to remove this limit at the
1016 expense of backing off one optimization in certain circumstances. Normally,
1017 when pcre_exec() would call its internal match() function recursively and
1018 immediately return the result unconditionally, it uses a "tail recursion"
1019 feature to save stack. However, when a subpattern that can match an empty
1020 string has an unlimited repetition quantifier, it no longer makes this
1021 optimization. That gives it a stack frame in which to save the data for
1022 checking that an empty string has been matched. Previously this was taken
1023 from the 1000-entry workspace that had been reserved. So now there is no
1024 explicit limit, but more stack is used.
1025
1026 8. Applied Daniel's patches to solve problems with the import/export magic
1027 syntax that is required for Windows, and which was going wrong for the
1028 pcreposix and pcrecpp parts of the library. These were overlooked when this
1029 problem was solved for the main library.
1030
1031 9. There were some crude static tests to avoid integer overflow when computing
1032 the size of patterns that contain repeated groups with explicit upper
1033 limits. As the maximum quantifier is 65535, the maximum group length was
1034 set at 30,000 so that the product of these two numbers did not overflow a
1035 32-bit integer. However, it turns out that people want to use groups that
1036 are longer than 30,000 bytes (though not repeat them that many times).
1037 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1038 made it possible to implement the integer overflow checks in a much more
1039 dynamic way, which I have now done. The artificial limitation on group
1040 length has been removed - we now have only the limit on the total length of
1041 the compiled pattern, which depends on the LINK_SIZE setting.
1042
1043 10. Fixed a bug in the documentation for get/copy named substring when
1044 duplicate names are permitted. If none of the named substrings are set, the
1045 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1046 empty string.
1047
1048 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1049 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1050 because the ] is interpreted as the first data character and the
1051 terminating ] is not found. PCRE has been made compatible with Perl in this
1052 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1053 cause memory overwriting.
1054
1055 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1056 string has been matched (to stop an infinite loop). It was not recognizing
1057 a conditional subpattern that could match an empty string if that
1058 subpattern was within another subpattern. For example, it looped when
1059 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1060 condition was not nested. This bug has been fixed.
1061
1062 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1063 past the start of the subject in the presence of bytes with the top bit
1064 set, for example "\x8aBCD".
1065
1066 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1067 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1068
1069 14. Optimized (?!) to (*FAIL).
1070
1071 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1072 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1073 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1074 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1075 does: it's just the validity check that is more restrictive.
1076
1077 16. Inserted checks for integer overflows during escape sequence (backslash)
1078 processing, and also fixed erroneous offset values for syntax errors during
1079 backslash processing.
1080
1081 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1082 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1083
1084 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1085 caused an overrun.
1086
1087 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1088 something other than just ASCII characters) inside a group that had an
1089 unlimited repeat caused a loop at compile time (while checking to see
1090 whether the group could match an empty string).
1091
1092 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1093 [\P{Any}] did so. (Error in the code for printing property names.)
1094
1095 21. An orphan \E inside a character class could cause a crash.
1096
1097 22. A repeated capturing bracket such as (A)? could cause a wild memory
1098 reference during compilation.
1099
1100 23. There are several functions in pcre_compile() that scan along a compiled
1101 expression for various reasons (e.g. to see if it's fixed length for look
1102 behind). There were bugs in these functions when a repeated \p or \P was
1103 present in the pattern. These operators have additional parameters compared
1104 with \d, etc, and these were not being taken into account when moving along
1105 the compiled data. Specifically:
1106
1107 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1108 length.
1109
1110 (b) An item such as \pL+ within a repeated group could cause crashes or
1111 loops.
1112
1113 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1114 "reference to non-existent subpattern" error.
1115
1116 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1117
1118 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1119 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1120
1121 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1122
1123 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1124 character were causing crashes (broken optimization).
1125
1126 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1127 \p or \P) caused a compile-time loop.
1128
1129 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1130 break. For example, the unstudied pattern [\r\n]A does not match the string
1131 "\r\nA" because change 7.0/46 below moves the current point on by two
1132 characters after failing to match at the start. However, the pattern \nA
1133 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1134 the same is true. There doesn't seem any very clean way out of this, but
1135 what I have chosen to do makes the common cases work: PCRE now takes note
1136 of whether there can be an explicit match for \r or \n anywhere in the
1137 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1138 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1139 pattern has explicit CR or LF references.
1140
1141 29. Added (*CR) etc for changing newline setting at start of pattern.
1142
1143
1144 Version 7.2 19-Jun-07
1145 ---------------------
1146
1147 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1148 which is apparently normally available under Windows.
1149
1150 2. Re-jig the pcregrep tests with different newline settings in an attempt
1151 to make them independent of the local environment's newline setting.
1152
1153 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1154
1155 4. Some of the "internals" tests were previously cut out when the link size
1156 was not 2, because the output contained actual offsets. The recent new
1157 "Z" feature of pcretest means that these can be cut out, making the tests
1158 usable with all link sizes.
1159
1160 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1161 stack recursion. This gives a massive performance boost under BSD, but just
1162 a small improvement under Linux. However, it saves one field in the frame
1163 in all cases.
1164
1165 6. Added more features from the forthcoming Perl 5.10:
1166
1167 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1168 recursion call. It refers to the nth most recently opened parentheses.
1169
1170 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1171 to be opened parentheses.
1172
1173 (c) Conditions that refer to capturing parentheses can be specified
1174 relatively, for example, (?(-2)... or (?(+3)...
1175
1176 (d) \K resets the start of the current match so that everything before
1177 is not part of it.
1178
1179 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1180
1181 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1182 reference syntax.
1183
1184 (g) (?| introduces a group in which the numbering of parentheses in each
1185 alternative starts with the same number.
1186
1187 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1188
1189 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1190 PCRE_INFO_JCHANGED.
1191
1192 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1193 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1194 for detecting groups that can match an empty string.
1195
1196 9. A pattern with a very large number of alternatives (more than several
1197 hundred) was running out of internal workspace during the pre-compile
1198 phase, where pcre_compile() figures out how much memory will be needed. A
1199 bit of new cunning has reduced the workspace needed for groups with
1200 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1201 workspace instead of running out of the 4096 that are available.
1202
1203 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1204
1205 11. Applied patch from Google to remove an optimization that didn't quite work.
1206 The report of the bug said:
1207
1208 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1209 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1210 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1211
1212 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1213 it matched the wrong number of bytes.
1214
1215
1216 Version 7.1 24-Apr-07
1217 ---------------------
1218
1219 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1220 that is more "standard", making use of automake and other Autotools. There
1221 is some re-arrangement of the files and adjustment of comments consequent
1222 on this.
1223
1224 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1225 for recursive directory scanning broke on some systems because the files
1226 are not scanned in any specific order and on different systems the order
1227 was different. A call to "sort" has been inserted into RunGrepTest for the
1228 approprate test as a short-term fix. In the longer term there may be an
1229 alternative.
1230
1231 3. I had an email from Eric Raymond about problems translating some of PCRE's
1232 man pages to HTML (despite the fact that I distribute HTML pages, some
1233 people do their own conversions for various reasons). The problems
1234 concerned the use of low-level troff macros .br and .in. I have therefore
1235 removed all such uses from the man pages (some were redundant, some could
1236 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1237 HTML has been updated to handle .nf/.fi and to complain if it encounters
1238 .br or .in.
1239
1240 4. Updated comments in configure.ac that get placed in config.h.in and also
1241 arranged for config.h to be included in the distribution, with the name
1242 config.h.generic, for the benefit of those who have to compile without
1243 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1244
1245 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1246 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1247 makevp.bat for the current PCRE, using the additional files
1248 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1249
1250 6. A Windows user reported a minor discrepancy with test 2, which turned out
1251 to be caused by a trailing space on an input line that had got lost in his
1252 copy. The trailing space was an accident, so I've just removed it.
1253
1254 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1255 that is needed.
1256
1257 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1258 as "const" (a) because they are and (b) because it helps the PHP
1259 maintainers who have recently made a script to detect big data structures
1260 in the php code that should be moved to the .rodata section. I remembered
1261 to update Builducptable as well, so it won't revert if ucptable.h is ever
1262 re-created.
1263
1264 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1265 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1266 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1267 support is not required. This saves 1.5-2K of code, which is important in
1268 some applications.
1269
1270 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1271 so as not to refer to the tables, even though these functions will never be
1272 called when UTF-8 support is disabled. Otherwise there are problems with a
1273 shared library.
1274
1275 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1276
1277 (a) It was defining its arguments as char * instead of void *.
1278
1279 (b) It was assuming that all moves were upwards in memory; this was true
1280 a long time ago when I wrote it, but is no longer the case.
1281
1282 The emulated memove() is provided for those environments that have neither
1283 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1284 is clearly not the case, as these two bugs were recently reported.
1285
1286 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1287 and Detrail to create the HTML documentation, the .txt form of the man
1288 pages, and it removes trailing spaces from listed files. It also creates
1289 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1290 case, it wraps all the #defines with #ifndefs. This script should be run
1291 before "make dist".
1292
1293 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1294 with Unicode property support.
1295
1296 (a) For a maximizing quantifier, if the two different cases of the
1297 character were of different lengths in their UTF-8 codings (there are
1298 some cases like this - I found 11), and the matching function had to
1299 back up over a mixture of the two cases, it incorrectly assumed they
1300 were both the same length.
1301
1302 (b) When PCRE was configured to use the heap rather than the stack for
1303 recursion during matching, it was not correctly preserving the data for
1304 the other case of a UTF-8 character when checking ahead for a match
1305 while processing a minimizing repeat. If the check also involved
1306 matching a wide character, but failed, corruption could cause an
1307 erroneous result when trying to check for a repeat of the original
1308 character.
1309
1310 13. Some tidying changes to the testing mechanism:
1311
1312 (a) The RunTest script now detects the internal link size and whether there
1313 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1314 values substituted by "configure". (The RunGrepTest script already did
1315 this for UTF-8.) The configure.ac script no longer substitutes the
1316 relevant variables.
1317
1318 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1319 with length and offset values. This means that the output is different
1320 for different internal link sizes. Test 2 is skipped for link sizes
1321 other than 2 because of this, bypassing the problem. Unfortunately,
1322 there was also a test in test 3 (the locale tests) that used /B and
1323 failed for link sizes other than 2. Rather than cut the whole test out,
1324 I have added a new /Z option to pcretest that replaces the length and
1325 offset values with spaces. This is now used to make test 3 independent
1326 of link size. (Test 2 will be tidied up later.)
1327
1328 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1329 segmentation fault instead of returning the appropriate error message.
1330
1331 15. In multiline mode when the newline sequence was set to "any", the pattern
1332 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1333 This doesn't seem right; it now treats the CRLF combination as the line
1334 ending, and so does not match in that case. It's only a pattern such as ^$
1335 that would hit this one: something like ^ABC$ would have failed after \r
1336 and then tried again after \r\n.
1337
1338 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1339 in an attempt to make files that differ only in their line terminators
1340 compare equal. This works on Linux.
1341
1342 17. Under certain error circumstances pcregrep might try to free random memory
1343 as it exited. This is now fixed, thanks to valgrind.
1344
1345 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1346 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1347 was because its rules for how to advance for /g after matching an empty
1348 string at the end of a line did not allow for this case. They now check for
1349 it specially.
1350
1351 20. pcretest is supposed to handle patterns and data of any length, by
1352 extending its buffers when necessary. It was getting this wrong when the
1353 buffer for a data line had to be extended.
1354
1355 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1356 CRLF as a newline sequence.
1357
1358 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1359 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1360 I have nevertheless tidied it up.
1361
1362 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1363
1364 24. Added a man page for pcre-config.
1365
1366
1367 Version 7.0 19-Dec-06
1368 ---------------------
1369
1370 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1371 moving to gcc 4.1.1.
1372
1373 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1374 sys/time.h, which is documented as needed for this function. It doesn't
1375 seem to matter on Linux, but it showed up on some releases of OS X.
1376
1377 3. It seems that there are systems where bytes whose values are greater than
1378 127 match isprint() in the "C" locale. The "C" locale should be the
1379 default when a C program starts up. In most systems, only ASCII printing
1380 characters match isprint(). This difference caused the output from pcretest
1381 to vary, making some of the tests fail. I have changed pcretest so that:
1382
1383 (a) When it is outputting text in the compiled version of a pattern, bytes
1384 other than 32-126 are always shown as hex escapes.
1385
1386 (b) When it is outputting text that is a matched part of a subject string,
1387 it does the same, unless a different locale has been set for the match
1388 (using the /L modifier). In this case, it uses isprint() to decide.
1389
1390 4. Fixed a major bug that caused incorrect computation of the amount of memory
1391 required for a compiled pattern when options that changed within the
1392 pattern affected the logic of the preliminary scan that determines the
1393 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1394 that the computed length was too small. The symptoms of this bug were
1395 either the PCRE error "internal error: code overflow" from pcre_compile(),
1396 or a glibc crash with a message such as "pcretest: free(): invalid next
1397 size (fast)". Examples of patterns that provoked this bug (shown in
1398 pcretest format) are:
1399
1400 /(?-x: )/x
1401 /(?x)(?-x: \s*#\s*)/
1402 /((?i)[\x{c0}])/8
1403 /(?i:[\x{c0}])/8
1404
1405 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1406 is now done differently.
1407
1408 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1409 wrapper classes; (b) implement a new function in the C++ scanner that is
1410 more efficient than the old way of doing things because it avoids levels of
1411 recursion in the regex matching; (c) add a paragraph to the documentation
1412 for the FullMatch() function.
1413
1414 6. The escape sequence \n was being treated as whatever was defined as
1415 "newline". Not only was this contrary to the documentation, which states
1416 that \n is character 10 (hex 0A), but it also went horribly wrong when
1417 "newline" was defined as CRLF. This has been fixed.
1418
1419 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1420 was being set to -1 for the "end of line" case (supposedly a value that no
1421 character can have). Though this value is never used (the check for end of
1422 line is "zero bytes in current character"), it caused compiler complaints.
1423 I've changed it to 0xffffffff.
1424
1425 8. In pcre_version.c, the version string was being built by a sequence of
1426 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1427 string (as it is for production releases) called a macro with an empty
1428 argument. The C standard says the result of this is undefined. The gcc
1429 compiler treats it as an empty string (which was what was wanted) but it is
1430 reported that Visual C gives an error. The source has been hacked around to
1431 avoid this problem.
1432
1433 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1434 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1435 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1436 of them did).
1437
1438 10. Originally, pcretest opened its input and output without "b"; then I was
1439 told that "b" was needed in some environments, so it was added for release
1440 5.0 to both the input and output. (It makes no difference on Unix-like
1441 systems.) Later I was told that it is wrong for the input on Windows. I've
1442 now abstracted the modes into two macros, to make it easier to fiddle with
1443 them, and removed "b" from the input mode under Windows.
1444
1445 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1446
1447 12. Added -help and --help to pcretest as an official way of being reminded
1448 of the options.
1449
1450 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1451 and pcrecpp.cc because they annoy compilers at high warning levels.
1452
1453 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1454
1455 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1456 scripts are not C programs :-) and which was not noticed because it works
1457 on Linux.
1458
1459 16. pcretest is supposed to handle any length of pattern and data line (as one
1460 line or as a continued sequence of lines) by extending its input buffer if
1461 necessary. This feature was broken for very long pattern lines, leading to
1462 a string of junk being passed to pcre_compile() if the pattern was longer
1463 than about 50K.
1464
1465 17. I have done a major re-factoring of the way pcre_compile() computes the
1466 amount of memory needed for a compiled pattern. Previously, there was code
1467 that made a preliminary scan of the pattern in order to do this. That was
1468 OK when PCRE was new, but as the facilities have expanded, it has become
1469 harder and harder to keep it in step with the real compile phase, and there
1470 have been a number of bugs (see for example, 4 above). I have now found a
1471 cunning way of running the real compile function in a "fake" mode that
1472 enables it to compute how much memory it would need, while actually only
1473 ever using a few hundred bytes of working memory and without too many
1474 tests of the mode. This should make future maintenance and development
1475 easier. A side effect of this work is that the limit of 200 on the nesting
1476 depth of parentheses has been removed (though this was never a serious
1477 limitation, I suspect). However, there is a downside: pcre_compile() now
1478 runs more slowly than before (30% or more, depending on the pattern). I
1479 hope this isn't a big issue. There is no effect on runtime performance.
1480
1481 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1482 newline (only possible for the last line of a file) and it was a
1483 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1484
1485 19. Added additional timing features to pcretest. (1) The -tm option now times
1486 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1487 separate command line item, by a number that specifies the number of
1488 repeats to use when timing. The default is 50000; this gives better
1489 precision, but takes uncomfortably long for very large patterns.
1490
1491 20. Extended pcre_study() to be more clever in cases where a branch of a
1492 subpattern has no definite first character. For example, (a*|b*)[cd] would
1493 previously give no result from pcre_study(). Now it recognizes that the
1494 first character must be a, b, c, or d.
1495
1496 21. There was an incorrect error "recursive call could loop indefinitely" if
1497 a subpattern (or the entire pattern) that was being tested for matching an
1498 empty string contained only one non-empty item after a nested subpattern.
1499 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1500 incorrectly, because the \d was being skipped in the check.
1501
1502 22. The pcretest program now has a new pattern option /B and a command line
1503 option -b, which is equivalent to adding /B to every pattern. This causes
1504 it to show the compiled bytecode, without the additional information that
1505 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1506 is the same as /B/I).
1507
1508 23. A new optimization is now able automatically to treat some sequences such
1509 as a*b as a*+b. More specifically, if something simple (such as a character
1510 or a simple class like \d) has an unlimited quantifier, and is followed by
1511 something that cannot possibly match the quantified thing, the quantifier
1512 is automatically "possessified".
1513
1514 24. A recursive reference to a subpattern whose number was greater than 39
1515 went wrong under certain circumstances in UTF-8 mode. This bug could also
1516 have affected the operation of pcre_study().
1517
1518 25. Realized that a little bit of performance could be had by replacing
1519 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1520
1521 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1522
1523 27. Possessive quantifiers such as a++ were previously implemented by turning
1524 them into atomic groups such as ($>a+). Now they have their own opcodes,
1525 which improves performance. This includes the automatically created ones
1526 from 23 above.
1527
1528 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1529 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1530 the first matched character to be a colon. This applied both to named and
1531 numbered groups.
1532
1533 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1534
1535 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1536 building PCRE on Windows easier, so I have included it in the distribution.
1537
1538 31. There is now a check in pcretest against a ridiculously large number being
1539 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1540 loop, the loop is abandoned.
1541
1542 32. Forward references to subpatterns in conditions such as (?(2)...) where
1543 subpattern 2 is defined later cause pcre_compile() to search forwards in
1544 the pattern for the relevant set of parentheses. This search went wrong
1545 when there were unescaped parentheses in a character class, parentheses
1546 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1547
1548 33. "Subroutine" calls and backreferences were previously restricted to
1549 referencing subpatterns earlier in the regex. This restriction has now
1550 been removed.
1551
1552 34. Added a number of extra features that are going to be in Perl 5.10. On the
1553 whole, these are just syntactic alternatives for features that PCRE had
1554 previously implemented using the Python syntax or my own invention. The
1555 other formats are all retained for compatibility.
1556
1557 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1558 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1559 also .NET compatible.
1560
1561 (b) A recursion or subroutine call to a named group can now be defined as
1562 (?&name) as well as (?P>name).
1563
1564 (c) A backreference to a named group can now be defined as \k<name> or
1565 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1566 5.10, are also .NET compatible.
1567
1568 (d) A conditional reference to a named group can now use the syntax
1569 (?(<name>) or (?('name') as well as (?(name).
1570
1571 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1572 groups (named and numbered) that are never evaluated inline, but can be
1573 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1574 is always false. There may be only one alternative in such a group.
1575
1576 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1577 as the simple (?(R). The condition is true only if the most recent
1578 recursion is that of the given number or name. It does not search out
1579 through the entire recursion stack.
1580
1581 (g) The escape \gN or \g{N} has been added, where N is a positive or
1582 negative number, specifying an absolute or relative reference.
1583
1584 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1585 some "unreachable code" warnings.
1586
1587 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1588 things, this adds five new scripts.
1589
1590 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1591 There were also incompatibilities regarding the handling of \Q..\E inside
1592 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1593 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1594
1595 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1596 matches an empty string, and forcibly breaks the loop. There were bugs in
1597 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1598 against aaaa the result was just "a" rather than "aaaa", for example. Two
1599 separate and independent bugs (that affected different cases) have been
1600 fixed.
1601
1602 39. Refactored the code to abolish the use of different opcodes for small
1603 capturing bracket numbers. This is a tidy that I avoided doing when I
1604 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1605 The new approach is not only tidier, it makes it possible to reduce the
1606 memory needed to fix the previous bug (38).
1607
1608 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1609 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1610 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1611 mode.
1612
1613 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1614 report.
1615
1616 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1617 copy construction and assignment in the C++ wrapper.
1618
1619 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1620 couple of bugs that could have given wrong results in the "--newline=crlf"
1621 case.
1622
1623 44. Added a number of casts and did some reorganization of signed/unsigned int
1624 variables following suggestions from Dair Grant. Also renamed the variable
1625 "this" as "item" because it is a C++ keyword.
1626
1627 45. Arranged for dftables to add
1628
1629 #include "pcre_internal.h"
1630
1631 to pcre_chartables.c because without it, gcc 4.x may remove the array
1632 definition from the final binary if PCRE is built into a static library and
1633 dead code stripping is activated.
1634
1635 46. For an unanchored pattern, if a match attempt fails at the start of a
1636 newline sequence, and the newline setting is CRLF or ANY, and the next two
1637 characters are CRLF, advance by two characters instead of one.
1638
1639
1640 Version 6.7 04-Jul-06
1641 ---------------------
1642
1643 1. In order to handle tests when input lines are enormously long, pcretest has
1644 been re-factored so that it automatically extends its buffers when
1645 necessary. The code is crude, but this _is_ just a test program. The
1646 default size has been increased from 32K to 50K.
1647
1648 2. The code in pcre_study() was using the value of the re argument before
1649 testing it for NULL. (Of course, in any sensible call of the function, it
1650 won't be NULL.)
1651
1652 3. The memmove() emulation function in pcre_internal.h, which is used on
1653 systems that lack both memmove() and bcopy() - that is, hardly ever -
1654 was missing a "static" storage class specifier.
1655
1656 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1657 containing an extended class (one that cannot be represented by a bitmap
1658 because it contains high-valued characters or Unicode property items, e.g.
1659 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1660 pattern, but PCRE should not loop if you do not (it no longer does).
1661 [Detail: two cases were found: (a) a repeated subpattern containing an
1662 extended class; (b) a recursive reference to a subpattern that followed a
1663 previous extended class. It wasn't skipping over the extended class
1664 correctly when UTF-8 mode was not set.]
1665
1666 5. A negated single-character class was not being recognized as fixed-length
1667 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1668 compile error "lookbehind assertion is not fixed length".
1669
1670 6. The RunPerlTest auxiliary script was showing an unexpected difference
1671 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1672 write a Perl script that can interpret lines of an input file either as
1673 byte characters or as UTF-8, which is what "perltest" was being required to
1674 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1675 can't do is switch easily at run time between having the "use utf8;" pragma
1676 or not. In the end, I fudged it by using the RunPerlTest script to insert
1677 "use utf8;" explicitly for the UTF-8 tests.
1678
1679 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1680 the end of the subject string, contrary to the documentation and to what
1681 Perl does. This was true of both matching functions. Now it matches only at
1682 the start of the subject and immediately after *internal* newlines.
1683
1684 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1685 a pointer to an int instead of a pointer to an unsigned long int. This
1686 caused problems on 64-bit systems.
1687
1688 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1689 instance of the 'standard' template library not being so standard".
1690
1691 10. There was no check on the number of named subpatterns nor the maximum
1692 length of a subpattern name. The product of these values is used to compute
1693 the size of the memory block for a compiled pattern. By supplying a very
1694 long subpattern name and a large number of named subpatterns, the size
1695 computation could be caused to overflow. This is now prevented by limiting
1696 the length of names to 32 characters, and the number of named subpatterns
1697 to 10,000.
1698
1699 11. Subpatterns that are repeated with specific counts have to be replicated in
1700 the compiled pattern. The size of memory for this was computed from the
1701 length of the subpattern and the repeat count. The latter is limited to
1702 65535, but there was no limit on the former, meaning that integer overflow
1703 could in principle occur. The compiled length of a repeated subpattern is
1704 now limited to 30,000 bytes in order to prevent this.
1705
1706 12. Added the optional facility to have named substrings with the same name.
1707
1708 13. Added the ability to use a named substring as a condition, using the
1709 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1710 are numbers (not recommended). Forward references are permitted.
1711
1712 14. Added forward references in named backreferences (if you see what I mean).
1713
1714 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1715 pattern could run off the end of the subject. For example, the pattern
1716 "(?s)(.{1,5})"8 did this with the subject "ab".
1717
1718 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1719 PCRE_CASELESS was set when matching characters that were quantified with ?
1720 or *.
1721
1722 17. A character class other than a single negated character that had a minimum
1723 but no maximum quantifier - for example [ab]{6,} - was not handled
1724 correctly by pce_dfa_exec(). It would match only one character.
1725
1726 18. A valid (though odd) pattern that looked like a POSIX character
1727 class but used an invalid character after [ (for example [[,abc,]]) caused
1728 pcre_compile() to give the error "Failed: internal error: code overflow" or
1729 in some cases to crash with a glibc free() error. This could even happen if
1730 the pattern terminated after [[ but there just happened to be a sequence of
1731 letters, a binary zero, and a closing ] in the memory that followed.
1732
1733 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1734 over the years. Originally (before any Unicode support), just the bottom 8
1735 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1736 output from "man perlunicode" includes this:
1737
1738 The regular expression compiler produces polymorphic opcodes. That
1739 is, the pattern adapts to the data and automatically switches to
1740 the Unicode character scheme when presented with Unicode data--or
1741 instead uses a traditional byte scheme when presented with byte
1742 data.
1743
1744 Sadly, a wide octal escape does not cause a switch, and in a string with
1745 no other multibyte characters, these octal escapes are treated as before.
1746 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1747 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1748 Unicode string.
1749
1750 I have not perpetrated such confusion in PCRE. Up till now, it took just
1751 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1752 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1753 translate to the appropriate multibyte character.
1754
1755 29. Applied some refactoring to reduce the number of warnings from Microsoft
1756 and Borland compilers. This has included removing the fudge introduced
1757 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1758 a warning about an unused variable.
1759
1760 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1761 characters since release 4.0, because Perl (from release 5.004) does not.
1762 [Or at least, is documented not to: some releases seem to be in conflict
1763 with the documentation.] However, when a pattern was studied with
1764 pcre_study() and all its branches started with \s, PCRE still included VT
1765 as a possible starting character. Of course, this did no harm; it just
1766 caused an unnecessary match attempt.
1767
1768 22. Removed a now-redundant internal flag bit that recorded the fact that case
1769 dependency changed within the pattern. This was once needed for "required
1770 byte" processing, but is no longer used. This recovers a now-scarce options
1771 bit. Also moved the least significant internal flag bit to the most-
1772 significant bit of the word, which was not previously used (hangover from
1773 the days when it was an int rather than a uint) to free up another bit for
1774 the future.
1775
1776 23. Added support for CRLF line endings as well as CR and LF. As well as the
1777 default being selectable at build time, it can now be changed at runtime
1778 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1779 specify that it is scanning data with non-default line endings.
1780
1781 24. Changed the definition of CXXLINK to make it agree with the definition of
1782 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1783
1784 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1785 recursions. This makes a big different to stack usage for some patterns.
1786
1787 26. If a subpattern containing a named recursion or subroutine reference such
1788 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1789 the space required for the compiled pattern went wrong and gave too small a
1790 value. Depending on the environment, this could lead to "Failed: internal
1791 error: code overflow at offset 49" or "glibc detected double free or
1792 corruption" errors.
1793
1794 27. Applied patches from Google (a) to support the new newline modes and (b) to
1795 advance over multibyte UTF-8 characters in GlobalReplace.
1796
1797 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1798 difference for some implementation of PCRE in some Windows version.
1799
1800 29. Added some extra testing facilities to pcretest:
1801
1802 \q<number> in a data line sets the "match limit" value
1803 \Q<number> in a data line sets the "match recursion limt" value
1804 -S <number> sets the stack size, where <number> is in megabytes
1805
1806 The -S option isn't available for Windows.
1807
1808
1809 Version 6.6 06-Feb-06
1810 ---------------------
1811
1812 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1813 in pcreposix.h. I have copied the definition from pcre.h.
1814
1815 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1816 because pcre.h is no longer a built file.
1817
1818 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1819 not normally included in the compiled code.
1820
1821
1822 Version 6.5 01-Feb-06
1823 ---------------------
1824
1825 1. When using the partial match feature with pcre_dfa_exec(), it was not
1826 anchoring the second and subsequent partial matches at the new starting
1827 point. This could lead to incorrect results. For example, with the pattern
1828 /1234/, partially matching against "123" and then "a4" gave a match.
1829
1830 2. Changes to pcregrep:
1831
1832 (a) All non-match returns from pcre_exec() were being treated as failures
1833 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1834 error message is output. Some extra information is given for the
1835 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1836 probably the only errors that are likely to be caused by users (by
1837 specifying a regex that has nested indefinite repeats, for instance).
1838 If there are more than 20 of these errors, pcregrep is abandoned.
1839
1840 (b) A binary zero was treated as data while matching, but terminated the
1841 output line if it was written out. This has been fixed: binary zeroes
1842 are now no different to any other data bytes.
1843
1844 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1845 used to set a locale for matching. The --locale=xxxx long option has
1846 been added (no short equivalent) to specify a locale explicitly on the
1847 pcregrep command, overriding the environment variables.
1848
1849 (d) When -B was used with -n, some line numbers in the output were one less
1850 than they should have been.
1851
1852 (e) Added the -o (--only-matching) option.
1853
1854 (f) If -A or -C was used with -c (count only), some lines of context were
1855 accidentally printed for the final match.
1856
1857 (g) Added the -H (--with-filename) option.
1858
1859 (h) The combination of options -rh failed to suppress file names for files
1860 that were found from directory arguments.
1861
1862 (i) Added the -D (--devices) and -d (--directories) options.
1863
1864 (j) Added the -F (--fixed-strings) option.
1865
1866 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1867
1868 (l) Added the --colo(u)r option.
1869
1870 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1871 is not present by default.
1872
1873 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1874 items such as (?R) or (?1), when the recursion could match a number of
1875 alternatives. If it matched one of the alternatives, but subsequently,
1876 outside the recursion, there was a failure, the code tried to back up into
1877 the recursion. However, because of the way PCRE is implemented, this is not
1878 possible, and the result was an incorrect result from the match.
1879
1880 In order to prevent this happening, the specification of recursion has
1881 been changed so that all such subpatterns are automatically treated as
1882 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1883
1884 4. I had overlooked the fact that, in some locales, there are characters for
1885 which isalpha() is true but neither isupper() nor islower() are true. In
1886 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1887 and ordfeminine) are like this. This affected the treatment of \w and \W
1888 when they appeared in character classes, but not when they appeared outside
1889 a character class. The bit map for "word" characters is now created
1890 separately from the results of isalnum() instead of just taking it from the
1891 upper, lower, and digit maps. (Plus the underscore character, of course.)
1892
1893 5. The above bug also affected the handling of POSIX character classes such as
1894 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1895 permanent tables. Instead, the bit maps for such a class were previously
1896 created as the appropriate unions of the upper, lower, and digit bitmaps.
1897 Now they are created by subtraction from the [[:word:]] class, which has
1898 its own bitmap.
1899
1900 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1901 It is created by subtracting the vertical space characters (\x09, \x0a,
1902 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1903 subtraction was done in the overall bitmap for a character class, meaning
1904 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1905 be recognized. This bug has been fixed.
1906
1907 7. Patches from the folks at Google:
1908
1909 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1910 real life, but is still worth protecting against".
1911
1912 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1913 regular expressions".
1914
1915 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1916 have it.
1917
1918 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1919 "configure" and the latter not, in order to fix a problem somebody had
1920 with compiling the Arg class on HP-UX.
1921
1922 (e) Improve the error-handling of the C++ wrapper a little bit.
1923
1924 (f) New tests for checking recursion limiting.
1925
1926 8. The pcre_memmove() function, which is used only if the environment does not
1927 have a standard memmove() function (and is therefore rarely compiled),
1928 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1929 returning a result (though PCRE never actually uses the result).
1930
1931 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1932 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1933 returned instead of calling malloc() with an overflowing number that would
1934 most likely cause subsequent chaos.
1935
1936 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1937
1938 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1939 with this option is matched, the nmatch and pmatch options of regexec() are
1940 ignored.
1941
1942 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1943 provided in case anyone wants to the the POSIX interface with UTF-8
1944 strings.
1945
1946 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1947 C++ linking (needed for some HP-UX environments).
1948
1949 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1950 (unused parameter) and in the pcre_printint() function (omitted "default"
1951 switch label when the default is to do nothing).
1952
1953 15. Added some code to make it possible, when PCRE is compiled as a C++
1954 library, to replace subject pointers for pcre_exec() with a smart pointer
1955 class, thus making it possible to process discontinuous strings.
1956
1957 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1958 much the same function. They were added by different people who were trying
1959 to make PCRE easy to compile on non-Unix systems. It has been suggested
1960 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1961 for compiling on Windows systems. I have therefore replaced it with
1962 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1963 defaults to "extern" for C or "extern C" for C++, which works fine on
1964 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1965 SCOPE with something explicit in config.h. In addition:
1966
1967 (a) pcreposix.h still had just "extern" instead of either of these macros;
1968 I have replaced it with PCRE_DATA_SCOPE.
1969
1970 (b) Functions such as _pcre_xclass(), which are internal to the library,
1971 but external in the C sense, all had PCRE_EXPORT in their definitions.
1972 This is apparently wrong for the Windows case, so I have removed it.
1973 (It makes no difference on Unix-like systems.)
1974
1975 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1976 of recursive calls to match(). This is different to MATCH_LIMIT because
1977 that limits the total number of calls to match(), not all of which increase
1978 the depth of recursion. Limiting the recursion depth limits the amount of
1979 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1980 when PCRE is compiled, and changed at run time. A patch from Google adds
1981 this functionality to the C++ interface.
1982
1983 18. Changes to the handling of Unicode character properties:
1984
1985 (a) Updated the table to Unicode 4.1.0.
1986
1987 (b) Recognize characters that are not in the table as "Cn" (undefined).
1988
1989 (c) I revised the way the table is implemented to a much improved format
1990 which includes recognition of ranges. It now supports the ranges that
1991 are defined in UnicodeData.txt, and it also amalgamates other
1992 characters into ranges. This has reduced the number of entries in the
1993 table from around 16,000 to around 3,000, thus reducing its size
1994 considerably. I realized I did not need to use a tree structure after
1995 all - a binary chop search is just as efficient. Having reduced the
1996 number of entries, I extended their size from 6 bytes to 8 bytes to
1997 allow for more data.
1998
1999 (d) Added support for Unicode script names via properties such as \p{Han}.
2000
2001 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2002 matching that character.
2003
2004 20. When matching a repeated Unicode property with a minimum greater than zero,
2005 (for example \pL{2,}), PCRE could look past the end of the subject if it
2006 reached it while seeking the minimum number of characters. This could
2007 happen only if some of the characters were more than one byte long, because
2008 there is a check for at least the minimum number of bytes.
2009
2010 21. Refactored the implementation of \p and \P so as to be more general, to
2011 allow for more different types of property in future. This has changed the
2012 compiled form incompatibly. Anybody with saved compiled patterns that use
2013 \p or \P will have to recompile them.
2014
2015 22. Added "Any" and "L&" to the supported property types.
2016
2017 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2018 but give a compile time error if the value is greater than 0xff.
2019
2020 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2021 accidentally not being installed or uninstalled.
2022
2023 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2024 made were to insert the current release number. This seemed silly, because
2025 it made things harder for people building PCRE on systems that don't run
2026 "configure". I have turned pcre.h into a distributed file, no longer built
2027 by "configure", with the version identification directly included. There is
2028 no longer a pcre.h.in file.
2029
2030 However, this change necessitated a change to the pcre-config script as
2031 well. It is built from pcre-config.in, and one of the substitutions was the
2032 release number. I have updated configure.ac so that ./configure now finds
2033 the release number by grepping pcre.h.
2034
2035 26. Added the ability to run the tests under valgrind.
2036
2037
2038 Version 6.4 05-Sep-05
2039 ---------------------
2040
2041 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2042 "--" to be printed when multiple files were scanned, even when none of the
2043 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2044 consider it to be a bug, and have restored the previous behaviour.
2045
2046 2. A couple of code tidies to get rid of compiler warnings.
2047
2048 3. The pcretest program used to cheat by referring to symbols in the library
2049 whose names begin with _pcre_. These are internal symbols that are not
2050 really supposed to be visible externally, and in some environments it is
2051 possible to suppress them. The cheating is now confined to including
2052 certain files from the library's source, which is a bit cleaner.
2053
2054 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2055 file's purpose clearer.
2056
2057 5. Reorganized pcre_ucp_findchar().
2058
2059
2060 Version 6.3 15-Aug-05
2061 ---------------------
2062
2063 1. The file libpcre.pc.in did not have general read permission in the tarball.
2064
2065 2. There were some problems when building without C++ support:
2066
2067 (a) If C++ support was not built, "make install" and "make test" still
2068 tried to test it.
2069
2070 (b) There were problems when the value of CXX was explicitly set. Some
2071 changes have been made to try to fix these, and ...
2072
2073 (c) --disable-cpp can now be used to explicitly disable C++ support.
2074
2075 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2076 backslash in a target when C++ was disabled. This confuses some
2077 versions of "make", apparently. Using an intermediate variable solves
2078 this. (Same for CPP_LOBJ.)
2079
2080 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2081 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2082 necessary on certain architectures.
2083
2084 4. Added a setting of -export-symbols-regex to the link command to remove
2085 those symbols that are exported in the C sense, but actually are local
2086 within the library, and not documented. Their names all begin with
2087 "_pcre_". This is not a perfect job, because (a) we have to except some
2088 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2089 available (and never for static libraries). I have made a note to try to
2090 find a way round (a) in the future.
2091
2092
2093 Version 6.2 01-Aug-05
2094 ---------------------
2095
2096 1. There was no test for integer overflow of quantifier values. A construction
2097 such as {1111111111111111} would give undefined results. What is worse, if
2098 a minimum quantifier for a parenthesized subpattern overflowed and became
2099 negative, the calculation of the memory size went wrong. This could have
2100 led to memory overwriting.
2101
2102 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2103
2104 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2105 operating environments where this matters.
2106
2107 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2108 PCRE options from within the C++ wrapper.
2109
2110 5. Named capturing subpatterns were not being correctly counted when a pattern
2111 was compiled. This caused two problems: (a) If there were more than 100
2112 such subpatterns, the calculation of the memory needed for the whole
2113 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2114 back references of the form \12, where the number was greater than 9, were
2115 not recognized as back references, even though there were sufficient
2116 previous subpatterns.
2117
2118 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2119 versions of gcc, e.g. 2.95.4.
2120
2121
2122 Version 6.1 21-Jun-05
2123 ---------------------
2124
2125 1. There was one reference to the variable "posix" in pcretest.c that was not
2126 surrounded by "#if !defined NOPOSIX".
2127
2128 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2129 the cross-check on the old pcre_info() function, for the benefit of the
2130 cut-down version of PCRE that is currently imported into Exim.
2131
2132 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2133 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2134 patterns that start (?i) but I don't think that matters. The use of (?i) is
2135 just an example; this all applies to the other options as well.
2136
2137 4. Since libtool seems to echo the compile commands it is issuing, the output
2138 from "make" can be reduced a bit by putting "@" in front of each libtool
2139 compile command.
2140
2141 5. Patch from the folks at Google for configure.in to be a bit more thorough
2142 in checking for a suitable C++ installation before trying to compile the
2143 C++ stuff. This should fix a reported problem when a compiler was present,
2144 but no suitable headers.
2145
2146 6. The man pages all had just "PCRE" as their title. I have changed them to
2147 be the relevant file name. I have also arranged that these names are
2148 retained in the file doc/pcre.txt, which is a concatenation in text format
2149 of all the man pages except the little individual ones for each function.
2150
2151 7. The NON-UNIX-USE file had not been updated for the different set of source
2152 files that come with release 6. I also added a few comments about the C++
2153 wrapper.
2154
2155
2156 Version 6.0 07-Jun-05
2157 ---------------------
2158
2159 1. Some minor internal re-organization to help with my DFA experiments.
2160
2161 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2162 didn't matter for the library itself when fully configured, but did matter
2163 when compiling without UCP support, or within Exim, where the ucp files are
2164 not imported.
2165
2166 3. Refactoring of the library code to split up the various functions into
2167 different source modules. The addition of the new DFA matching code (see
2168 below) to a single monolithic source would have made it really too
2169 unwieldy, quite apart from causing all the code to be include in a
2170 statically linked application, when only some functions are used. This is
2171 relevant even without the DFA addition now that patterns can be compiled in
2172 one application and matched in another.
2173
2174 The downside of splitting up is that there have to be some external
2175 functions and data tables that are used internally in different modules of
2176 the library but which are not part of the API. These have all had their
2177 names changed to start with "_pcre_" so that they are unlikely to clash
2178 with other external names.
2179
2180 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2181 a different (DFA) algorithm. Although it is slower than the original
2182 function, it does have some advantages for certain types of matching
2183 problem.
2184
2185 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2186 including restarting after a partial match.
2187
2188 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2189 defined when compiling for Windows was sent to me. I have put it into the
2190 code, though I have no means of testing or verifying it.
2191
2192 7. Added the pcre_refcount() auxiliary function.
2193
2194 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2195 match before or at the first newline in the subject string. In pcretest,
2196 the /f option on a pattern can be used to set this.
2197
2198 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2199 would behave wrongly. This has been present in PCRE since release 4.0.
2200
2201 10. A number of changes to the pcregrep command:
2202
2203 (a) Refactored how -x works; insert ^(...)$ instead of setting
2204 PCRE_ANCHORED and checking the length, in preparation for adding
2205 something similar for -w.
2206
2207 (b) Added the -w (match as a word) option.
2208
2209 (c) Refactored the way lines are read and buffered so as to have more
2210 than one at a time available.
2211
2212 (d) Implemented a pcregrep test script.
2213
2214 (e) Added the -M (multiline match) option. This allows patterns to match
2215 over several lines of the subject. The buffering ensures that at least
2216 8K, or the rest of the document (whichever is the shorter) is available
2217 for matching (and similarly the previous 8K for lookbehind assertions).
2218
2219 (f) Changed the --help output so that it now says
2220
2221 -w, --word-regex(p)
2222
2223 instead of two lines, one with "regex" and the other with "regexp"
2224 because that confused at least one person since the short forms are the
2225 same. (This required a bit of code, as the output is generated
2226 automatically from a table. It wasn't just a text change.)
2227
2228 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2229 option but starts with a hyphen. Could be a pattern or a path name
2230 starting with a hyphen, for instance.
2231
2232 (h) "-" can be given as a file name to represent stdin.
2233
2234 (i) When file names are being printed, "(standard input)" is used for
2235 the standard input, for compatibility with GNU grep. Previously
2236 "<stdin>" was used.
2237
2238 (j) The option --label=xxx can be used to supply a name to be used for
2239 stdin when file names are being printed. There is no short form.
2240
2241 (k) Re-factored the options decoding logic because we are going to add
2242 two more options that take data. Such options can now be given in four
2243 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2244
2245 (l) Added the -A, -B, and -C options for requesting that lines of context
2246 around matches be printed.
2247
2248 (m) Added the -L option to print the names of files that do not contain
2249 any matching lines, that is, the complement of -l.
2250
2251 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2252 continue to scan other files.
2253
2254 (o) The -s option was incorrectly implemented. For compatibility with other
2255 greps, it now suppresses the error message for a non-existent or non-
2256 accessible file (but not the return code). There is a new option called
2257 -q that suppresses the output of matching lines, which was what -s was
2258 previously doing.
2259
2260 (p) Added --include and --exclude options to specify files for inclusion
2261 and exclusion when recursing.
2262
2263 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2264 Hopefully, it now does.
2265
2266 12. Missing cast in pcre_study().
2267
2268 13. Added an "uninstall" target to the makefile.
2269
2270 14. Replaced "extern" in the function prototypes in Makefile.in with
2271 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2272 world, but is set differently for Windows.
2273
2274 15. Added a second compiling function called pcre_compile2(). The only
2275 difference is that it has an extra argument, which is a pointer to an
2276 integer error code. When there is a compile-time failure, this is set
2277 non-zero, in addition to the error test pointer being set to point to an
2278 error message. The new argument may be NULL if no error number is required
2279 (but then you may as well call pcre_compile(), which is now just a
2280 wrapper). This facility is provided because some applications need a
2281 numeric error indication, but it has also enabled me to tidy up the way
2282 compile-time errors are handled in the POSIX wrapper.
2283
2284 16. Added VPATH=.libs to the makefile; this should help when building with one
2285 prefix path and installing with another. (Or so I'm told by someone who
2286 knows more about this stuff than I do.)
2287
2288 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2289 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2290 match everything, including newlines. This is not POSIX-compatible, but
2291 somebody wanted the feature. From pcretest it can be activated by using
2292 both the P and the s flags.
2293
2294 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2295
2296 19. libpcre.pc was being incorrectly installed as executable.
2297
2298 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2299 it now also looks for '\r' so that it will work unmodified on Windows.
2300
2301 21. Added Google's contributed C++ wrapper to the distribution.
2302
2303 22. Added some untidy missing memory free() calls in pcretest, to keep
2304 Electric Fence happy when testing.
2305
2306
2307
2308 Version 5.0 13-Sep-04
2309 ---------------------
2310
2311 1. Internal change: literal characters are no longer packed up into items
2312 containing multiple characters in a single byte-string. Each character
2313 is now matched using a separate opcode. However, there may be more than one
2314 byte in the character in UTF-8 mode.
2315
2316 2. The pcre_callout_block structure has two new fields: pattern_position and
2317 next_item_length. These contain the offset in the pattern to the next match
2318 item, and its length, respectively.
2319
2320 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2321 insertion of callouts before each pattern item. Added the /C option to
2322 pcretest to make use of this.
2323
2324 4. On the advice of a Windows user, the lines
2325
2326 #if defined(_WIN32) || defined(WIN32)
2327 _setmode( _fileno( stdout ), 0x8000 );
2328 #endif /* defined(_WIN32) || defined(WIN32) */
2329
2330 have been added to the source of pcretest. This apparently does useful
2331 magic in relation to line terminators.
2332
2333 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2334 for the benefit of those environments where the "b" makes a difference.
2335
2336 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2337 to know about it. I have put a hack into configure.in that adds in code
2338 to set GCC=yes if CC=icc. This seems to end up at a point in the
2339 generated configure script that is early enough to affect the setting of
2340 compiler options, which is what is needed, but I have no means of testing
2341 whether it really works. (The user who reported this had patched the
2342 generated configure script, which of course I cannot do.)
2343
2344 LATER: After change 22 below (new libtool files), the configure script
2345 seems to know about icc (and also ecc). Therefore, I have commented out
2346 this hack in configure.in.
2347
2348 7. Added support for pkg-config (2 patches were sent in).
2349
2350 8. Negated POSIX character classes that used a combination of internal tables
2351 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2352 [[:^ascii]]. Typically, they would match almost any characters. The other
2353 POSIX classes were not broken in this way.
2354
2355 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2356 to find the match, as PCRE was deluded into thinking that the match had to
2357 start at the start point or following a newline. The same bug applied to
2358 patterns with negative forward assertions or any backward assertions
2359 preceding ".*" at the start, unless the pattern required a fixed first
2360 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2361
2362 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2363 starting at the last subject character, bytes beyond the end of the subject
2364 string were read.
2365
2366 11. Renamed the variable "class" as "classbits" to make life easier for C++
2367 users. (Previously there was a macro definition, but it apparently wasn't
2368 enough.)
2369
2370 12. Added the new field "tables" to the extra data so that tables can be passed
2371 in at exec time, or the internal tables can be re-selected. This allows
2372 a compiled regex to be saved and re-used at a later time by a different
2373 program that might have everything at different addresses.
2374
2375 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2376 -R library as well as a -L library.
2377
2378 14. The debugging options of pcretest (-d on the command line or D on a
2379 pattern) showed incorrect output for anything following an extended class
2380 that contained multibyte characters and which was followed by a quantifier.
2381
2382 15. Added optional support for general category Unicode character properties
2383 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2384 support. It adds about 90K to the size of the library. The meanings of the
2385 inbuilt class escapes such as \d and \s have NOT been changed.
2386
2387 16. Updated pcredemo.c to include calls to free() to release the memory for the
2388 compiled pattern.
2389
2390 17. The generated file chartables.c was being created in the source directory
2391 instead of in the building directory. This caused the build to fail if the
2392 source directory was different from the building directory, and was
2393 read-only.
2394
2395 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2396 file. No doubt somebody will tell me if they don't make sense... Also added
2397 Dan Mooney's comments about building on OpenVMS.
2398
2399 19. Added support for partial matching via the PCRE_PARTIAL option for
2400 pcre_exec() and the \P data escape in pcretest.
2401
2402 20. Extended pcretest with 3 new pattern features:
2403
2404 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2405 write the compiled pattern to the file whose name is "rest-of-line".
2406 This is a straight binary dump of the data, with the saved pointer to
2407 the character tables forced to be NULL. The study data, if any, is
2408 written too. After writing, pcretest reads a new pattern.
2409
2410 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2411 compiled pattern from the given file. There must not be any
2412 occurrences of "<" in the file name (pretty unlikely); if there are,
2413 pcretest will instead treat the initial "<" as a pattern delimiter.
2414 After reading in the pattern, pcretest goes on to read data lines as
2415 usual.
2416
2417 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2418 and 16-bit fields in a compiled pattern, to simulate a pattern that
2419 was compiled on a host of opposite endianness.
2420
2421 21. The pcre-exec() function can now cope with patterns that were compiled on
2422 hosts of opposite endianness, with this restriction:
2423
2424 As for any compiled expression that is saved and used later, the tables
2425 pointer field cannot be preserved; the extra_data field in the arguments
2426 to pcre_exec() should be used to pass in a tables address if a value
2427 other than the default internal tables were used at compile time.
2428
2429 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2430 now diagnosed as an error. Previously, most of the time, a negative number
2431 would have been treated as zero, but if in addition "ovector" was passed as
2432 NULL, a crash could occur.
2433
2434 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2435 new versions from the libtool 1.5 distribution (the last one is a copy of
2436 a file called libtool.m4). This seems to have fixed the need to patch
2437 "configure" to support Darwin 1.3 (which I used to do). However, I still
2438 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2439 workstation).
2440
2441 24. Changed the PCRE licence to be the more standard "BSD" licence.
2442
2443
2444 Version 4.5 01-Dec-03
2445 ---------------------
2446
2447 1. There has been some re-arrangement of the code for the match() function so
2448 that it can be compiled in a version that does not call itself recursively.
2449 Instead, it keeps those local variables that need separate instances for
2450 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2451 needs to "recurse". Keeping track of where control must go is done by means
2452 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2453 hide most of the details from the main code, and operates only if
2454 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2455 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2456 operating.
2457
2458 To make it easier for callers to provide specially tailored get/free
2459 functions for this usage, two new functions, pcre_stack_malloc, and
2460 pcre_stack_free, are used. They are always called in strict stacking order,
2461 and the size of block requested is always the same.
2462
2463 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2464 PCRE has been compiled to use the stack or the heap for recursion. The
2465 -C option of pcretest uses this to show which version is compiled.
2466
2467 A new data escape \S, is added to pcretest; it causes the amounts of store
2468 obtained and freed by both kinds of malloc/free at match time to be added
2469 to the output.
2470
2471 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2472 what's available on my current Linux desktop machine.
2473
2474 3. When matching a UTF-8 string, the test for a valid string at the start has
2475 been extended. If start_offset is not zero, PCRE now checks that it points
2476 to a byte that is the start of a UTF-8 character. If not, it returns
2477 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2478 this is necessary because there may be backward assertions in the pattern.
2479 When matching the same subject several times, it may save resources to use
2480 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2481
2482 4. The code for checking the validity of UTF-8 strings has been tightened so
2483 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2484 containing "overlong sequences".
2485
2486 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2487 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2488 should have been "&", but it just so happened that all the cases this let
2489 through by mistake were picked up later in the function.
2490
2491 6. I had used a variable called "isblank" - this is a C99 function, causing
2492 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2493
2494 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2495 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2496 so the version is shown; (c) stop "make test" echoing "./RunTest".
2497
2498 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2499
2500 9. The prototype for memmove() for systems that don't have it was using
2501 size_t, but the inclusion of the header that defines size_t was later. I've
2502 moved the #includes for the C headers earlier to avoid this.
2503
2504 10. Added some adjustments to the code to make it easier to compiler on certain
2505 special systems:
2506
2507 (a) Some "const" qualifiers were missing.
2508 (b) Added the macro EXPORT before all exported functions; by default this
2509 is defined to be empty.
2510 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2511 that it reads its output file name as an argument instead of writing
2512 to the standard output and assuming this can be redirected.
2513
2514 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2515 class containing characters with values greater than 255, PCRE compilation
2516 went into a loop.
2517
2518 12. A recursive reference to a subpattern that was within another subpattern
2519 that had a minimum quantifier of zero caused PCRE to crash. For example,
2520 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2521 recursion. If the recursively-called subpattern itself had a zero repeat,
2522 that was OK.
2523
2524 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2525 buffer into which it was copied (for escape processing) was still set at
2526 1024, so long lines caused crashes.
2527
2528 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2529 "internal error: code overflow...". This applied to any character class
2530 that was followed by a possessive quantifier.
2531
2532 15. Modified the Makefile to add libpcre.la as a prerequisite for
2533 libpcreposix.la because I was told this is needed for a parallel build to
2534 work.
2535
2536 16. If a pattern that contained .* following optional items at the start was
2537 studied, the wrong optimizing data was generated, leading to matching
2538 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2539 matching string must start with a or b or c. The correct conclusion for
2540 this pattern is that a match can start with any character.
2541
2542
2543 Version 4.4 13-Aug-03
2544 ---------------------
2545
2546 1. In UTF-8 mode, a character class containing characters with values between
2547 127 and 255 was not handled correctly if the compiled pattern was studied.
2548 In fixing this, I have also improved the studying algorithm for such
2549 classes (slightly).
2550
2551 2. Three internal functions had redundant arguments passed to them. Removal
2552 might give a very teeny performance improvement.
2553
2554 3. Documentation bug: the value of the capture_top field in a callout is *one
2555 more than* the number of the hightest numbered captured substring.
2556
2557 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2558 in incorrectly linking with a previously installed version. They now link
2559 explicitly with libpcre.la.
2560
2561 5. configure.in no longer needs to recognize Cygwin specially.
2562
2563 6. A problem in pcre.in for Windows platforms is fixed.
2564
2565 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2566 pcretest, it used to include the size of the study block as part of its
2567 output. Unfortunately, the structure contains a field that has a different
2568 size on different hardware architectures. This meant that the tests that
2569 showed this size failed. As the block is currently always of a fixed size,
2570 this information isn't actually particularly useful in pcretest output, so
2571 I have just removed it.
2572
2573 8. Three pre-processor statements accidentally did not start in column 1.
2574 Sadly, there are *still* compilers around that complain, even though
2575 standard C has not required this for well over a decade. Sigh.
2576
2577 9. In pcretest, the code for checking callouts passed small integers in the
2578 callout_data field, which is a void * field. However, some picky compilers
2579 complained about the casts involved for this on 64-bit systems. Now
2580 pcretest passes the address of the small integer instead, which should get
2581 rid of the warnings.
2582
2583 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2584 both compile and run time, and gives an error if an invalid UTF-8 sequence
2585 is found. There is a option for disabling this check in cases where the
2586 string is known to be correct and/or the maximum performance is wanted.
2587
2588 11. In response to a bug report, I changed one line in Makefile.in from
2589
2590 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2591 to
2592 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2593
2594 to look similar to other lines, but I have no way of telling whether this
2595 is the right thing to do, as I do not use Windows. No doubt I'll get told
2596 if it's wrong...
2597
2598
2599 Version 4.3 21-May-03
2600 ---------------------
2601
2602 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2603 Makefile.
2604
2605 2. Some refactoring to improve the quality of the code:
2606
2607 (i) The utf8_table... variables are now declared "const".
2608
2609 (ii) The code for \cx, which used the "case flipping" table to upper case
2610 lower case letters, now just substracts 32. This is ASCII-specific,
2611 but the whole concept of \cx is ASCII-specific, so it seems
2612 reasonable.
2613
2614 (iii) PCRE was using its character types table to recognize decimal and
2615 hexadecimal digits in the pattern. This is silly, because it handles
2616 only 0-9, a-f, and A-F, but the character types table is locale-
2617 specific, which means strange things might happen. A private
2618 table is now used for this - though it costs 256 bytes, a table is
2619 much faster than multiple explicit tests. Of course, the standard
2620 character types table is still used for matching digits in subject
2621 strings against \d.
2622
2623 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2624 ending in _t are). So I've renamed it as ESC_tee.
2625
2626 3. The first argument for regexec() in the POSIX wrapper should have been
2627 defined as "const".
2628
2629 4. Changed pcretest to use malloc() for its buffers so that they can be
2630 Electric Fenced for debugging.
2631
2632 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2633 to read one or more bytes before the start of the subject string. Often this
2634 had no effect on PCRE's behaviour, but in some circumstances it could
2635 provoke a segmentation fault.
2636
2637 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2638 to try to read one or more bytes before the start of the subject string.
2639
2640 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2641 UTF-8 support could misbehave in various ways if the subject string
2642 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2643 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2644 back over UTF-8 characters.)
2645
2646
2647 Version 4.2 14-Apr-03
2648 ---------------------
2649
2650 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2651
2652 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2653 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2654 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2655 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2656 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2657 and BUILD_EXEEXT
2658 Note: automatic setting of the BUILD variables is not yet working
2659 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2660 compile-time but not at link-time
2661 [LINK]: use for linking executables only
2662 make different versions for Windows and non-Windows
2663 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2664 libraries
2665 [LINK_FOR_BUILD]: new variable
2666 [OBJEXT]: use throughout
2667 [EXEEXT]: use throughout
2668 <winshared>: new target
2669 <wininstall>: new target
2670 <dftables.o>: use native compiler
2671 <dftables>: use native linker
2672 <install>: handle Windows platform correctly
2673 <clean>: ditto
2674 <check>: ditto
2675 copy DLL to top builddir before testing
2676
2677 As part of these changes, -no-undefined was removed again. This was reported
2678 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2679 in any case.
2680
2681 3. Some tidies to get rid of compiler warnings:
2682
2683 . In the match_data structure, match_limit was an unsigned long int, whereas
2684 match_call_count was an int. I've made them both unsigned long ints.
2685
2686 . In pcretest the fact that a const uschar * doesn't automatically cast to
2687 a void * provoked a warning.
2688
2689 . Turning on some more compiler warnings threw up some "shadow" variables
2690 and a few more missing casts.
2691
2692 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2693 option, a class that contained a single character with a value between 128
2694 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2695
2696 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2697 option, a class that contained several characters, but with at least one
2698 whose value was between 128 and 255 caused PCRE to crash.
2699
2700
2701 Version 4.1 12-Mar-03
2702 ---------------------
2703
2704 1. Compiling with gcc -pedantic found a couple of places where casts were
2705 needed, and a string in dftables.c that was longer than standard compilers are
2706 required to support.
2707
2708 2. Compiling with Sun's compiler found a few more places where the code could
2709 be tidied up in order to avoid warnings.
2710
2711 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2712 first of these names is deprecated in the latest Autoconf in favour of the name
2713 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2714 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2715 analogy I have changed it to CFLAGS_FOR_BUILD.
2716
2717 4. Added -no-undefined to the linking command in the Makefile, because this is
2718 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2719 linking step for the pcreposix library.
2720
2721 5. PCRE was failing to diagnose the case of two named groups with the same
2722 name.
2723
2724 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2725 literal character that is needed in the subject for a match, and scans along to
2726 ensure that it is present before embarking on the full matching process. This
2727 saves time in cases of nested unlimited repeats that are never going to match.
2728 Problem: the scan can take a lot of time if the subject is very long (e.g.
2729 megabytes), thus penalizing straightforward matches. It is now done only if the
2730 amount of subject to be scanned is less than 1000 bytes.
2731
2732 7. A lesser problem with the same optimization is that it was recording the
2733 first character of an anchored pattern as "needed", thus provoking a search
2734 right along the subject, even when the first match of the pattern was going to
2735 fail. The "needed" character is now not set for anchored patterns, unless it
2736 follows something in the pattern that is of non-fixed length. Thus, it still
2737 fulfils its original purpose of finding quick non-matches in cases of nested
2738 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2739
2740
2741 Version 4.0 17-Feb-03
2742 ---------------------
2743
2744 1. If a comment in an extended regex that started immediately after a meta-item
2745 extended to the end of string, PCRE compiled incorrect data. This could lead to
2746 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2747
2748 2. Moved to autoconf 2.53 and libtool 1.4.2.
2749
2750 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2751 the special perltest8 script is no longer needed - all the tests can be run
2752 from a single perltest script.
2753
2754 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2755 by \s. It has now been removed in PCRE. This means it isn't recognized as
2756 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2757 class [:space:] *does* include VT, thereby creating a mess.
2758
2759 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2760 space and tab.
2761
2762 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2763 its new features into the main test script, reducing the number of scripts.
2764
2765 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2766 were backward compatible, and made the (?i) apply to the whole pattern, as if
2767 /i were given. Now it behaves more logically, and applies the option setting
2768 only to what follows. PCRE has been changed to follow suit. However, if it
2769 finds options settings right at the start of the pattern, it extracts them into
2770 the global options, as before. Thus, they show up in the info data.
2771
2772 8. Added support for the \Q...\E escape sequence. Characters in between are
2773 treated as literals. This is slightly different from Perl in that $ and @ are
2774 also handled as literals inside the quotes. In Perl, they will cause variable
2775 interpolation. Note the following examples:
2776
2777 Pattern PCRE matches Perl matches
2778
2779 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2780 \Qabc\$xyz\E abc\$xyz abc\$xyz
2781 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2782
2783 For compatibility with Perl, \Q...\E sequences are recognized inside character
2784 classes as well as outside them.
2785
2786 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2787 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2788 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2789 signed/unsigned warnings.
2790
2791 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2792 option for pcretest, so I've replaced it by a simple function that does just
2793 that job.
2794
2795 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2796 "pcregrep -".
2797
2798 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2799 Java package. This provides some syntactic sugar for simple cases of what my
2800 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2801 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2802 item, you can use this simplified notation. Note that only makes sense with
2803 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2804 greediness, whatever the setting of the PCRE_UNGREEDY option.
2805
2806 13. A change of greediness default within a pattern was not taking effect at
2807 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2808 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2809 was abstracted outside.
2810
2811 14. PCRE now supports the \G assertion. It is true when the current matching
2812 position is at the start point of the match. This differs from \A when the
2813 starting offset is non-zero. Used with the /g option of pcretest (or similar
2814 code), it works in the same way as it does for Perl's /g option. If all
2815 alternatives of a regex begin with \G, the expression is anchored to the start
2816 match position, and the "anchored" flag is set in the compiled expression.
2817
2818 15. Some bugs concerning the handling of certain option changes within patterns
2819 have been fixed. These applied to options other than (?ims). For example,
2820 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2821 been the other way round. Some of this was related to change 7 above.
2822
2823 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2824 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2825 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2826 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2827
2828 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2829 mode. Unlike ".", it always matches newline, whatever the setting of
2830 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2831 assertions. Perl allows it, but it doesn't (in general) work because it can't
2832 calculate the length of the lookbehind. At least, that's the case for Perl
2833 5.8.0 - I've been told they are going to document that it doesn't work in
2834 future.
2835
2836 18. Added an error diagnosis for escapes that PCRE does not support: these are
2837 \L, \l, \N, \P, \p, \U, \u, and \X.
2838
2839 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2840 reading past the end of the pattern in cases such as /[abcd/.
2841
2842 20. PCRE was getting more memory than necessary for patterns with classes that
2843 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2844
2845 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2846 compiling PCRE for use with Virtual Pascal.
2847
2848 22. Small fix to the Makefile to make it work properly if the build is done
2849 outside the source tree.
2850
2851 23. Added a new extension: a condition to go with recursion. If a conditional
2852 subpattern starts with (?(R) the "true" branch is used if recursion has
2853 happened, whereas the "false" branch is used only at the top level.
2854
2855 24. When there was a very long string of literal characters (over 255 bytes
2856 without UTF support, over 250 bytes with UTF support), the computation of how
2857 much memory was required could be incorrect, leading to segfaults or other
2858 strange effects.
2859
2860 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2861 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2862 there was a subsequent back reference to those brackets. This meant that, for
2863 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2864 possible to check for precisely this case. All we can do is abandon the
2865 optimization if .* occurs inside capturing brackets when there are any back
2866 references whatsoever. (See below for a better fix that came later.)
2867
2868 26. The handling of the optimization for finding the first character of a
2869 non-anchored pattern, and for finding a character that is required later in the
2870 match were failing in some cases. This didn't break the matching; it just
2871 failed to optimize when it could. The way this is done has been re-implemented.
2872
2873 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2874
2875 28. Added a new feature that provides some of the functionality that Perl
2876 provides with (?{...}). The facility is termed a "callout". The way it is done
2877 in PCRE is for the caller to provide an optional function, by setting
2878 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2879 global variable. By default it is unset, which disables all calling out. To get
2880 the function called, the regex must include (?C) at appropriate points. This
2881 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2882 This provides a means of identifying different callout points. When PCRE
2883 reaches such a point in the regex, if pcre_callout has been set, the external
2884 function is called. It is provided with data in a structure called
2885 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2886 matching continues; if it returns a non-zero value, the match at the current
2887 point fails. However, backtracking will occur if possible. [This was changed
2888 later and other features added - see item 49 below.]
2889
2890 29. pcretest is upgraded to test the callout functionality. It provides a
2891 callout function that displays information. By default, it shows the start of
2892 the match and the current position in the text. There are some new data escapes
2893 to vary what happens:
2894
2895 \C+ in addition, show current contents of captured substrings
2896 \C- do not supply a callout function
2897 \C!n return 1 when callout number n is reached
2898 \C!n!m return 1 when callout number n is reached for the mth time
2899
2900 30. If pcregrep was called with the -l option and just a single file name, it
2901 output "<stdin>" if a match was found, instead of the file name.
2902
2903 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2904 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2905 pcre_exec(). This saves a malloc/free per call. The default value of
2906 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2907 when configuring.
2908
2909 32. The default maximum size of a compiled pattern is 64K. There have been a
2910 few cases of people hitting this limit. The code now uses macros to handle the
2911 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2912 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2913 configuring. Tests 2 and 5 work only with 2-byte links because they output
2914 debugging information about compiled patterns.
2915
2916 33. Internal code re-arrangements:
2917
2918 (a) Moved the debugging function for printing out a compiled regex into
2919 its own source file (printint.c) and used #include to pull it into
2920 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2921 separate copies.
2922
2923 (b) Defined the list of op-code names for debugging as a macro in
2924 internal.h so that it is next to the definition of the opcodes.
2925
2926 (c) Defined a table of op-code lengths for simpler skipping along compiled
2927 code. This is again a macro in internal.h so that it is next to the
2928 definition of the opcodes.
2929
2930 34. Added support for recursive calls to individual subpatterns, along the
2931 lines of Robin Houston's patch (but implemented somewhat differently).
2932
2933 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2934 allow it to read and process whole directories in Win32. This code was
2935 contributed by Lionel Fourquaux; it has not been tested by me.
2936
2937 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2938 used to name a group. Names consist of alphanumerics and underscores, and must
2939 be unique. Back references use the syntax (?P=name) and recursive calls use
2940 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2941 numbers. The function pcre_fullinfo() can be used after compilation to extract
2942 a name/number map. There are three relevant calls:
2943
2944 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2945 PCRE_INFO_NAMECOUNT yields the number of entries
2946 PCRE_INFO_NAMETABLE yields a pointer to the map.
2947
2948 The map is a vector of fixed-size entries. The size of each entry depends on
2949 the length of the longest name used. The first two bytes of each entry are the
2950 group number, most significant byte first. There follows the corresponding
2951 name, zero terminated. The names are in alphabetical order.
2952
2953 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2954 case instead of 255. Making it the same both with and without UTF-8 support
2955 means that the same test output works with both.
2956
2957 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2958 calling malloc() with a zero argument.
2959
2960 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2961 optimization. I've improved things by keeping a bitmap of backreferences with
2962 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2963 fact referenced, the optimization can be applied. It is unlikely that a
2964 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2965 the match to follow \n) will appear inside brackets with a number greater than
2966 31, but if it does, any back reference > 31 suppresses the optimization.
2967
2968 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2969 of disabling numbered capturing parentheses. Any opening parenthesis that is
2970 not followed by ? behaves as if it were followed by ?: but named parentheses
2971 can still be used for capturing (and they will acquire numbers in the usual
2972 way).
2973
2974 41. Redesigned the return codes from the match() function into yes/no/error so
2975 that errors can be passed back from deep inside the nested calls. A malloc
2976 failure while inside a recursive subpattern call now causes the
2977 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2978
2979 42. It is now possible to set a limit on the number of times the match()
2980 function is called in a call to pcre_exec(). This facility makes it possible to
2981 limit the amount of recursion and backtracking, though not in a directly
2982 obvious way, because the match() function is used in a number of different
2983 circumstances. The count starts from zero for each position in the subject
2984 string (for non-anchored patterns). The default limit is, for compatibility, a
2985 large number, namely 10 000 000. You can change this in two ways:
2986
2987 (a) When configuring PCRE before making, you can use --with-match-limit=n
2988 to set a default value for the compiled library.
2989
2990 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2991 a different value is set. See 45 below.
2992
2993 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2994
2995 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2996 of things that can be changed at compile time. The first argument specifies
2997 what is wanted and the second points to where the information is to be placed.
2998 The current list of available information is:
2999
3000 PCRE_CONFIG_UTF8
3001
3002 The output is an integer that is set to one if UTF-8 support is available;
3003 otherwise it is set to zero.
3004
3005 PCRE_CONFIG_NEWLINE
3006
3007 The output is an integer that it set to the value of the code that is used for
3008 newline. It is either LF (10) or CR (13).
3009
3010 PCRE_CONFIG_LINK_SIZE
3011
3012 The output is an integer that contains the number of bytes used for internal
3013 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3014
3015 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3016
3017 The output is an integer that contains the threshold above which the POSIX
3018 interface uses malloc() for output vectors. See item 31 above.
3019
3020 PCRE_CONFIG_MATCH_LIMIT
3021
3022 The output is an unsigned integer that contains the default limit of the number
3023 of match() calls in a pcre_exec() execution. See 42 above.
3024
3025 44. pcretest has been upgraded by the addition of the -C option. This causes it
3026 to extract all the available output from the new pcre_config() function, and to
3027 output it. The program then exits immediately.
3028
3029 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3030 order to support additional features. One way would have been to define
3031 pcre_exec2() (for example) with extra arguments, but this would not have been
3032 extensible, and would also have required all calls to the original function to
3033 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3034 is used for passing in "extra" data from pcre_study().
3035
3036 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3037 contains the following fields:
3038
3039 flags a bitmap indicating which of the following fields are set
3040 study_data opaque data from pcre_study()
3041 match_limit a way of specifying a limit on match() calls for a specific
3042 call to pcre_exec()
3043 callout_data data for callouts (see 49 below)
3044
3045 The flag bits are also defined in pcre.h, and are
3046
3047 PCRE_EXTRA_STUDY_DATA
3048 PCRE_EXTRA_MATCH_LIMIT
3049 PCRE_EXTRA_CALLOUT_DATA
3050
3051 The pcre_study() function now returns one of these new pcre_extra blocks, with
3052 the actual study data pointed to by the study_data field, and the
3053 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3054 before. That is, this change is entirely upwards-compatible and requires no
3055 change to existing code.
3056
3057 If you want to pass in additional data to pcre_exec(), you can either place it
3058 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3059 block.
3060
3061 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3062 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3063 times with different match limits, until it finds the minimum value needed for
3064 pcre_exec() to complete. The value is then output. This can be instructive; for
3065 most simple matches the number is quite small, but for pathological cases it
3066 gets very large very quickly.
3067
3068 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3069 returns the size of the data block pointed to by the study_data field in a
3070 pcre_extra block, that is, the value that was passed as the argument to
3071 pcre_malloc() when PCRE was getting memory in which to place the information
3072 created by pcre_study(). The fourth argument should point to a size_t variable.
3073 pcretest has been extended so that this information is shown after a successful
3074 pcre_study() call when information about the compiled regex is being displayed.
3075
3076 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3077 because what follows is always an absolute path. (Later: it turns out that this
3078 is more than cosmetic for MinGW, because it doesn't like empty path
3079 components.)
3080
3081 49. Some changes have been made to the callout feature (see 28 above):
3082
3083 (i) A callout function now has three choices for what it returns:
3084
3085 0 => success, carry on matching
3086 > 0 => failure at this point, but backtrack if possible
3087 < 0 => serious error, return this value from pcre_exec()
3088
3089 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3090 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3091 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3092 use by callout functions. It will never be used by PCRE itself.
3093
3094 (ii) The pcre_extra structure (see 45 above) has a void * field called
3095 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3096 pcre_callout_block structure has a field of the same name. The contents of
3097 the field passed in the pcre_extra structure are passed to the callout
3098 function in the corresponding field in the callout block. This makes it
3099 easier to use the same callout-containing regex from multiple threads. For
3100 testing, the pcretest program has a new data escape
3101
3102 \C*n pass the number n (may be negative) as callout_data
3103
3104 If the callout function in pcretest receives a non-zero value as
3105 callout_data, it returns that value.
3106
3107 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3108 there were some redundant $(CFLAGS) in commands that are now specified as
3109 $(LINK), which already includes $(CFLAGS).
3110
3111 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3112 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3113 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3114 one-byte characters throughout. Note that case-insensitive matching applies
3115 only to characters whose values are less than 256. PCRE doesn't support the
3116 notion of cases for higher-valued characters.
3117
3118 (i) A character class whose characters are all within 0-255 is handled as
3119 a bit map, and the map is inverted for negative classes. Previously, a
3120 character > 255 always failed to match such a class; however it should
3121 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3122
3123 (ii) A negated character class with a single character < 255 is coded as
3124 "not this character" (OP_NOT). This wasn't working properly when the test
3125 character was multibyte, either singly or repeated.
3126
3127 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3128 mode, for example: \x{100}{2,3}.
3129
3130 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3131 singly or repeated) now correctly test multibyte characters. However,
3132 PCRE doesn't recognize any characters with values greater than 255 as
3133 digits, spaces, or word characters. Such characters always match \D, \S,
3134 and \W, and never match \d, \s, or \w.
3135
3136 (v) Classes may now contain characters and character ranges with values
3137 greater than 255. For example: [ab\x{100}-\x{400}].
3138
3139 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3140 PCRE in UTF-8 mode.
3141
3142 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3143 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3144 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3145 value.)
3146
3147 53. The single man page has become too large. I have therefore split it up into
3148 a number of separate man pages. These also give rise to individual HTML pages;
3149 these are now put in a separate directory, and there is an index.html page that
3150 lists them all. Some hyperlinking between the pages has been installed.
3151
3152 54. Added convenience functions for handling named capturing parentheses.
3153
3154 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3155 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3156 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3157 are faulted.
3158
3159 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3160 calling configure. These values are used when compiling the dftables.c program
3161 which is run to generate the source of the default character tables. They
3162 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3163 you will need to set these values.
3164
3165 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3166
3167
3168 Version 3.9 02-Jan-02
3169 ---------------------
3170
3171 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3172
3173 2. If --disable-static was given, the building process failed when trying to
3174 build pcretest and pcregrep. (For some reason it was using libtool to compile
3175 them, which is not right, as they aren't part of the library.)
3176
3177
3178 Version 3.8 18-Dec-01
3179 ---------------------
3180
3181 1. The experimental UTF-8 code was completely screwed up. It was packing the
3182 bytes in the wrong order. How dumb can you get?
3183
3184
3185 Version 3.7 29-Oct-01
3186 ---------------------
3187
3188 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3189 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3190 this didn't happen under Solaris 8, where I normally test things.
3191
3192 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3193 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3194 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3195 this fix an hour or so after the initial 3.7 release.)
3196
3197
3198 Version 3.6 23-Oct-01
3199 ---------------------
3200
3201 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3202 offsets passed as NULL with zero offset count.
3203
3204 2. The config.guess and config.sub files had not been updated when I moved to
3205 the latest autoconf.
3206
3207
3208 Version 3.5 15-Aug-01
3209 ---------------------
3210
3211 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3212 had been forgotten.
3213
3214 2. By using declared but undefined structures, we can avoid using "void"
3215 definitions in pcre.h while keeping the internal definitions of the structures
3216 private.
3217
3218 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3219 user point of view, this means that both static and shared libraries are built
3220 by default, but this can be individually controlled. More of the work of
3221 handling this static/shared cases is now inside libtool instead of PCRE's make
3222 file.
3223
3224 4. The pcretest utility is now installed along with pcregrep because it is
3225 useful for users (to test regexs) and by doing this, it automatically gets
3226 relinked by libtool. The documentation has been turned into a man page, so
3227 there are now .1, .txt, and .html versions in /doc.
3228
3229 5. Upgrades to pcregrep:
3230 (i) Added long-form option names like gnu grep.
3231 (ii) Added --help to list all options with an explanatory phrase.
3232 (iii) Added -r, --recursive to recurse into sub-directories.
3233 (iv) Added -f, --file to read patterns from a file.
3234
3235 6. pcre_exec() was referring to its "code" argument before testing that
3236 argument for NULL (and giving an error if it was NULL).
3237
3238 7. Upgraded Makefile.in to allow for compiling in a different directory from
3239 the source directory.
3240
3241 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3242 options bits, the pointer it was passed was to an int instead of to an unsigned
3243 long int. This mattered only on 64-bit systems.
3244
3245 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3246 generated) instead of pcre.in, which it its source. Also made the same change
3247 in several of the .c files.
3248
3249 10. A new release of gcc defines printf() as a macro, which broke pcretest
3250 because it had an ifdef in the middle of a string argument for printf(). Fixed
3251 by using separate calls to printf().
3252
3253 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3254 script, to force use of CR or LF instead of \n in the source. On non-Unix
3255 systems, the value can be set in config.h.
3256
3257 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3258 absolute limit. Changed the text of the error message to make this clear, and
3259 likewise updated the man page.
3260
3261 13. The limit of 99 on the number of capturing subpatterns has been removed.
3262 The new limit is 65535, which I hope will not be a "real" limit.
3263
3264
3265 Version 3.4 22-Aug-00
3266 ---------------------
3267
3268 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3269
3270 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3271
3272
3273 Version 3.3 01-Aug-00
3274 ---------------------
3275
3276 1. If an octal character was given, but the value was greater than \377, it
3277 was not getting masked to the least significant bits, as documented. This could
3278 lead to crashes in some systems.
3279
3280 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3281 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3282
3283 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3284 These just pass their arguments on to (pcre_free)(), but they are provided
3285 because some uses of PCRE bind it to non-C systems that can call its functions,
3286 but cannot call free() or pcre_free() directly.
3287
3288 4. Add "make test" as a synonym for "make check". Corrected some comments in
3289 the Makefile.
3290
3291 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3292 Makefile.
3293
3294 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3295 command called pgrep for grepping around the active processes.
3296
3297 7. Added the beginnings of support for UTF-8 character strings.
3298
3299 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3300 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3301 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3302 out for the ar command.)
3303
3304
3305 Version 3.2 12-May-00
3306 ---------------------
3307
3308 This is purely a bug fixing release.
3309
3310 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3311 of ZA. This was just one example of several cases that could provoke this bug,
3312 which was introduced by change 9 of version 2.00. The code for breaking
3313 infinite loops after an iteration that matches an empty string was't working
3314 correctly.
3315
3316 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3317 when matched against abbab (for example). After matching an empty string, it
3318 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3319 caused it to match further down the string than it should.
3320
3321 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3322 was there because it doesn't seem to be needed, and it causes trouble on some
3323 systems, as it is not a Standard C header. It has been removed.
3324
3325 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3326 were reported on the Macintosh. The changes were from
3327
3328 while ((c = *(++ptr)) != 0 && c != '\n');
3329 to
3330 while ((c = *(++ptr)) != 0 && c != '\n') ;
3331
3332 Totally extraordinary, but if that's what it takes...
3333
3334 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3335 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3336 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3337 assumes the way PCRE uses memmove() (always moving upwards).
3338
3339 6. PCRE is being used in one environment where strchr() is not available. There
3340 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3341 faster code anyway.
3342
3343
3344 Version 3.1 09-Feb-00
3345 ---------------------
3346
3347 The only change in this release is the fixing of some bugs in Makefile.in for
3348 the "install" target:
3349
3350 (1) It was failing to install pcreposix.h.
3351
3352 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3353
3354
3355 Version 3.0 01-Feb-00
3356 ---------------------
3357
3358 1. Add support for the /+ modifier to perltest (to output $` like it does in
3359 pcretest).
3360
3361 2. Add support for the /g modifier to perltest.
3362
3363 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3364 matches null strings.
3365
3366 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3367 pattern. Perl treats empty patterns specially - it reuses the most recent
3368 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3369 effect.
3370
3371 5. The POSIX interface was broken in that it was just handing over the POSIX
3372 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3373 required a bigger vector, with some working space on the end. This means that
3374 the POSIX wrapper now has to get and free some memory, and copy the results.
3375
3376 6. Added some simple autoconf support, placing the test data and the
3377 documentation in separate directories, re-organizing some of the
3378 information files, and making it build pcre-config (a GNU standard). Also added
3379 libtool support for building PCRE as a shared library, which is now the
3380 default.
3381
3382 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3383 09 are not valid octal constants. Single digits will be used for minor values
3384 less than 10.
3385
3386 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3387 existing programs that set these in the POSIX interface can use PCRE without
3388 modification.
3389
3390 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3391 return all that pcre_info() returns, plus additional data. The pcre_info()
3392 function is retained for compatibility, but is considered to be obsolete.
3393
3394 10. Added experimental recursion feature (?R) to handle one common case that
3395 Perl 5.6 will be able to do with (?p{...}).
3396
3397 11. Added support for POSIX character classes like [:alpha:], which Perl is
3398 adopting.
3399
3400
3401 Version 2.08 31-Aug-99
3402 ----------------------
3403
3404 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3405 trying to match at the startoffset position, but instead was moving forward to
3406 the next newline as if a previous match had failed.
3407
3408 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3409 and could get into a loop if a null string was matched other than at the start
3410 of the subject.
3411
3412 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3413 be distinguished at compile time, and for completeness also added PCRE_DATE.
3414
3415 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3416 in GnuWin32 environments.
3417
3418
3419 Version 2.07 29-Jul-99
3420 ----------------------
3421
3422 1. The documentation is now supplied in plain text form and HTML as well as in
3423 the form of man page sources.
3424
3425 2. C++ compilers don't like assigning (void *) values to other pointer types.
3426 In particular this affects malloc(). Although there is no problem in Standard
3427 C, I've put in casts to keep C++ compilers happy.
3428
3429 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3430 should be (const char *).
3431
3432 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3433 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3434 However, I haven't made this a standard facility. The documentation doesn't
3435 mention it, and the Makefile doesn't support it.
3436
3437 5. The Makefile now contains an "install" target, with editable destinations at
3438 the top of the file. The pcretest program is not installed.
3439
3440 6. pgrep -V now gives the PCRE version number and date.
3441
3442 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3443 causing the entire string to be ignored, instead of just the last character.
3444
3445 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3446 non-matching string, it can take a very, very long time, even for strings of
3447 quite modest length, because of the nested recursion. PCRE now does better in
3448 some of these cases. It does this by remembering the last required literal
3449 character in the pattern, and pre-searching the subject to ensure it is present
3450 before running the real match. In other words, it applies a heuristic to detect
3451 some types of certain failure quickly, and in the above example, if presented
3452 with a string that has no trailing " it gives "no match" very quickly.
3453
3454 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3455 other alternatives are tried instead.
3456
3457
3458 Version 2.06 09-Jun-99
3459 ----------------------
3460
3461 1. Change pcretest's output for amount of store used to show just the code
3462 space, because the remainder (the data block) varies in size between 32-bit and
3463 64-bit systems.
3464
3465 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3466 start matching at. This allows lookbehinds to work when searching for multiple
3467 occurrences in a string.
3468
3469 3. Added additional options to pcretest for testing multiple occurrences:
3470
3471 /+ outputs the rest of the string that follows a match
3472 /g loops for multiple occurrences, using the new startoffset argument
3473 /G loops for multiple occurrences by passing an incremented pointer
3474
3475 4. PCRE wasn't doing the "first character" optimization for patterns starting
3476 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3477 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3478 the letter 'x'. On long subject strings, this gives a significant speed-up.
3479
3480
3481 Version 2.05 21-Apr-99
3482 ----------------------
3483
3484 1. Changed the type of magic_number from int to long int so that it works
3485 properly on 16-bit systems.
3486
3487 2. Fixed a bug which caused patterns starting with .* not to work correctly
3488 when the subject string contained newline characters. PCRE was assuming
3489 anchoring for such patterns in all cases, which is not correct because .* will
3490 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3491 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3492 must be retried after every newline in the subject.
3493
3494
3495 Version 2.04 18-Feb-99
3496 ----------------------
3497
3498 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3499 computation of the store needed to hold the pattern was incorrect (too large).
3500 If such patterns were nested a few deep, this could multiply and become a real
3501 problem.
3502
3503 2. Added /M option to pcretest to show the memory requirement of a specific
3504 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3505
3506 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3507 compiled in such a way that the backtracking after subsequent failure was
3508 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3509 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3510
3511
3512 Version 2.03 02-Feb-99
3513 ----------------------
3514
3515 1. Fixed typo and small mistake in man page.
3516
3517 2. Added 4th condition (GPL supersedes if conflict) and created separate
3518 LICENCE file containing the conditions.
3519
3520 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3521 Perl, that is the internal \ allows the delimiter to be included in the
3522 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3523 the final delimiter, add \ to the end of the pattern (to test the error).
3524
3525 4. Added the convenience functions for extracting substrings after a successful
3526 match. Updated pcretest to make it able to test these functions.
3527
3528
3529 Version 2.02 14-Jan-99
3530 ----------------------
3531
3532 1. Initialized the working variables associated with each extraction so that
3533 their saving and restoring doesn't refer to uninitialized store.
3534
3535 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3536 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3537 fix the problem.
3538
3539 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3540 calls, and wasn't printing the correct value for compiling calls. Increased the
3541 default value of LOOPREPEAT, and the number of significant figures in the
3542 times.
3543
3544 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3545
3546 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3547 a building problem on Windows NT with a FAT file system.
3548
3549
3550 Version 2.01 21-Oct-98
3551 ----------------------
3552
3553 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3554 to character tables built by pcre_maketables() in the current locale. If NULL
3555 is passed, the default tables are used.
3556
3557
3558 Version 2.00 24-Sep-98
3559 ----------------------
3560
3561 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3562 it any more.
3563
3564 2. Allow quantification of (?>) groups, and make it work correctly.
3565
3566 3. The first character computation wasn't working for (?>) groups.
3567
3568 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3569 end of the subject) and add 5.005's \z, which really does match only at the
3570 very end of the subject.
3571
3572 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3573
3574 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3575 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3576 localized options. All options to pcre_study() were also removed.
3577
3578 7. Add other new features from 5.005:
3579
3580 $(?<= positive lookbehind
3581 $(?<! negative lookbehind
3582 (?imsx-imsx) added the unsetting capability
3583 such a setting is global if at outer level; local otherwise
3584 (?imsx-imsx:) non-capturing groups with option setting
3585 (?(cond)re|re) conditional pattern matching
3586
3587 A backreference to itself in a repeated group matches the previous
3588 captured string.
3589
3590 8. General tidying up of studying (both automatic and via "study")
3591 consequential on the addition of new assertions.
3592
3593 9. As in 5.005, unlimited repeated groups that could match an empty substring
3594 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3595 runtime if any iteration does actually match an empty substring.
3596
3597 10. Include the RunTest script in the distribution.
3598
3599 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3600 discrepancies, some of which were old and were also with respect to 5.004. They
3601 have now been fixed.
3602
3603
3604 Version 1.09 28-Apr-98
3605 ----------------------
3606
3607 1. A negated single character class followed by a quantifier with a minimum
3608 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3609 program crashes, or just wrong answers. This did not apply to negated classes
3610 containing more than one character, or to minima other than one.
3611
3612
3613 Version 1.08 27-Mar-98
3614 ----------------------
3615
3616 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3617
3618 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3619 latter must appear before anything that relies on it in the pattern.
3620
3621
3622 Version 1.07 16-Feb-98
3623 ----------------------
3624
3625 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3626 repeat of a potentially empty string).
3627
3628
3629 Version 1.06 23-Jan-98
3630 ----------------------
3631
3632 1. Added Markus Oberhumer's little patches for C++.
3633
3634 2. Literal strings longer than 255 characters were broken.
3635
3636
3637 Version 1.05 23-Dec-97
3638 ----------------------
3639
3640 1. Negated character classes containing more than one character were failing if
3641 PCRE_CASELESS was set at run time.
3642
3643
3644 Version 1.04 19-Dec-97
3645 ----------------------
3646
3647 1. Corrected the man page, where some "const" qualifiers had been omitted.
3648
3649 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3650 input syntax.
3651
3652 3. Fixed memory leak which occurred when a regex with back references was
3653 matched with an offsets vector that wasn't big enough. The temporary memory
3654 that is used in this case wasn't being freed if the match failed.
3655
3656 4. Tidied pcretest to ensure it frees memory that it gets.
3657
3658 5. Temporary memory was being obtained in the case where the passed offsets
3659 vector was exactly big enough.
3660
3661 6. Corrected definition of offsetof() from change 5 below.
3662
3663 7. I had screwed up change 6 below and broken the rules for the use of
3664 setjmp(). Now fixed.
3665
3666
3667 Version 1.03 18-Dec-97
3668 ----------------------
3669
3670 1. A erroneous regex with a missing opening parenthesis was correctly
3671 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3672 on some systems.
3673
3674 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3675 it was reported that one broken compiler failed on the former because "code" is
3676 also an independent variable.
3677
3678 3. The erroneous regex a[]b caused an array overrun reference.
3679
3680 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3681 fail on data ending with that character. (It was going on too far, and checking
3682 the next character, typically a binary zero.) This was specific to the
3683 optimized code for single-character negative classes.
3684
3685 5. Added a contributed patch from the TIN world which does the following:
3686
3687 + Add an undef for memmove, in case the the system defines a macro for it.
3688
3689 + Add a definition of offsetof(), in case there isn't one. (I don't know
3690 the reason behind this - offsetof() is part of the ANSI standard - but
3691 it does no harm).
3692
3693 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3694 most of the places where whitespace preceded '#'. I have given up and
3695 allowed the remaining 2 cases to be at the margin.
3696
3697 + Rename some variables in pcre to eliminate shadowing. This seems very
3698 pedantic, but does no harm, of course.
3699
3700 6. Moved the call to setjmp() into its own function, to get rid of warnings
3701 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3702
3703 7. Constructs such as \d{8,} were compiling into the equivalent of
3704 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3705 outcome, but in this particular case used more store than had been allocated,
3706 which caused the bug to be discovered because it threw up an internal error.
3707
3708 8. The debugging code in both pcre and pcretest for outputting the compiled
3709 form of a regex was going wrong in the case of back references followed by
3710 curly-bracketed repeats.
3711
3712
3713 Version 1.02 12-Dec-97
3714 ----------------------
3715
3716 1. Typos in pcre.3 and comments in the source fixed.
3717
3718 2. Applied a contributed patch to get rid of places where it used to remove
3719 'const' from variables, and fixed some signed/unsigned and uninitialized
3720 variable warnings.
3721
3722 3. Added the "runtest" target to Makefile.
3723
3724 4. Set default compiler flag to -O2 rather than just -O.
3725
3726
3727 Version 1.01 19-Nov-97
3728 ----------------------
3729
3730 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3731 like /([ab]*)*/, that is, for classes with more than one character in them.
3732
3733 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3734 as /((?>a*))*/ (a PCRE_EXTRA facility).
3735
3736
3737 Version 1.00 18-Nov-97
3738 ----------------------
3739
3740 1. Added compile-time macros to support systems such as SunOS4 which don't have
3741 memmove() or strerror() but have other things that can be used instead.
3742
3743 2. Arranged that "make clean" removes the executables.
3744
3745
3746 Version 0.99 27-Oct-97
3747 ----------------------
3748
3749 1. Fixed bug in code for optimizing classes with only one character. It was
3750 initializing a 32-byte map regardless, which could cause it to run off the end
3751 of the memory it had got.
3752
3753 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3754
3755
3756 Version 0.98 22-Oct-97
3757 ----------------------
3758
3759 1. Fixed bug in code for handling temporary memory usage when there are more
3760 back references than supplied space in the ovector. This could cause segfaults.
3761
3762
3763 Version 0.97 21-Oct-97
3764 ----------------------
3765
3766 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3767
3768 2. Optimized negated single characters not to use a bit map.
3769
3770 3. Brought error texts together as macro definitions; clarified some of them;
3771 fixed one that was wrong - it said "range out of order" when it meant "invalid
3772 escape sequence".
3773
3774 4. Changed some char * arguments to const char *.
3775
3776 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3777
3778 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3779 pcretest.
3780
3781
3782 Version 0.96 16-Oct-97
3783 ----------------------
3784
3785 1. Added a simple "pgrep" utility to the distribution.
3786
3787 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3788 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3789 where "ddd" means "one or more decimal digits".
3790
3791 3. Fixed serious bug. If a pattern had a back reference, but the call to
3792 pcre_exec() didn't supply a large enough ovector to record the related
3793 identifying subpattern, the match always failed. PCRE now remembers the number
3794 of the largest back reference, and gets some temporary memory in which to save
3795 the offsets during matching if necessary, in order to ensure that
3796 backreferences always work.
3797
3798 4. Increased the compatibility with Perl in a number of ways:
3799
3800 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3801 to request this handling. The option can be set at compile or exec time.
3802
3803 (b) $ matches before a terminating newline by default; an option
3804 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3805 mode). The option can be set at compile or exec time.
3806
3807 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3808 the same as Perl's. If the decimal number it represents is less than 10
3809 or there aren't that many previous left capturing parentheses, an octal
3810 escape is read. Inside a character class, it's always an octal escape,
3811 even if it is a single digit.
3812
3813 (d) An escaped but undefined alphabetic character is taken as a literal,
3814 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3815 escapes.
3816
3817 (e) {0} is now permitted. (The previous item is removed from the compiled
3818 pattern).
3819
3820 5. Changed all the names of code files so that the basic parts are no longer
3821 than 10 characters, and abolished the teeny "globals.c" file.
3822
3823 6. Changed the handling of character classes; they are now done with a 32-byte
3824 bit map always.
3825
3826 7. Added the -d and /D options to pcretest to make it possible to look at the
3827 internals of compilation without having to recompile pcre.
3828
3829
3830 Version 0.95 23-Sep-97
3831 ----------------------
3832
3833 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3834 \x20 at the start of a run of normal characters. These were being treated as
3835 real characters, instead of the source characters being re-checked.
3836
3837
3838 Version 0.94 18-Sep-97
3839 ----------------------
3840
3841 1. The functions are now thread-safe, with the caveat that the global variables
3842 containing pointers to malloc() and free() or alternative functions are the
3843 same for all threads.
3844
3845 2. Get pcre_study() to generate a bitmap of initial characters for non-
3846 anchored patterns when this is possible, and use it if passed to pcre_exec().
3847
3848
3849 Version 0.93 15-Sep-97
3850 ----------------------
3851
3852 1. /(b)|(:+)/ was computing an incorrect first character.
3853
3854 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3855 but not actually doing anything yet.
3856
3857 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3858 as Perl does (e.g. [-az] or [az-]).
3859
3860 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3861 all possible positions.
3862
3863 5. Split up into different modules to avoid including unneeded functions in a
3864 compiled binary. However, compile and exec are still in one module. The "study"
3865 function is split off.
3866
3867 6. The character tables are now in a separate module whose source is generated
3868 by an auxiliary program - but can then be edited by hand if required. There are
3869 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3870 toupper() in the code.
3871
3872 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3873 make them global. Abolish the function for setting them, as the caller can now
3874 set them directly.
3875
3876
3877 Version 0.92 11-Sep-97
3878 ----------------------
3879
3880 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3881 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3882
3883 2. Caseless matching was not working in character classes if the characters in
3884 the pattern were in upper case.
3885
3886 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3887
3888 4. Make PCRE_ANCHORED public and accept as a compile option.
3889
3890 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3891 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3892 pass them.
3893
3894 6. Give an error if bad option bits passed at compile or run time.
3895
3896 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3897 pcretest to cause it to pass that flag.
3898
3899 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3900 options, and the first character, if set.
3901
3902 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3903
3904
3905 Version 0.91 10-Sep-97
3906 ----------------------
3907
3908 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3909 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3910
3911 2. PCRE was looping on encountering an indefinitely repeated back reference to
3912 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3913 Perl does - treats the match as successful.
3914
3915 ****

Properties

Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

  ViewVC Help
Powered by ViewVC 1.1.5