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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 537 - (show annotations)
Tue Jun 8 15:28:26 2010 UTC (5 years, 2 months ago) by ph10
File size: 188117 byte(s)
Error occurred while calculating annotation data.
Add newline in pcretest output if the last data line doesn't have one.
1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.10 03-Jun-2010
5 ------------------------
6
7 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
8 THEN.
9
10 2. (*ACCEPT) was not working when inside an atomic group.
11
12 3. Inside a character class, \B is treated as a literal by default, but
13 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14 causes the error). The code is unchanged, but I tidied the documentation.
15
16 4. Inside a character class, PCRE always treated \R and \X as literals,
17 whereas Perl faults them if its -w option is set. I have changed PCRE so
18 that it faults them when PCRE_EXTRA is set.
19
20 5. Added support for \N, which always matches any character other than
21 newline. (It is the same as "." when PCRE_DOTALL is not set.)
22
23 6. When compiling pcregrep with newer versions of gcc which may have
24 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25 declared with attribute warn_unused_result" were given. Just casting the
26 result to (void) does not stop the warnings; a more elaborate fudge is
27 needed. I've used a macro to implement this.
28
29 7. Minor change to pcretest.c to avoid a compiler warning.
30
31 8. Added four artifical Unicode properties to help with an option to make
32 \s etc use properties (see next item). The new properties are: Xan
33 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
34
35 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
36 use Unicode properties. (*UCP) at the start of a pattern can be used to set
37 this option. Modified pcretest to add /W to test this facility. Added
38 REG_UCP to make it available via the POSIX interface.
39
40 10. Added --line-buffered to pcregrep.
41
42 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
43 studied, and the match started with a letter with a code point greater than
44 127 whose first byte was different to the first byte of the other case of
45 the letter, the other case of this starting letter was not recognized
46 (#976).
47
48 12. If a pattern that was studied started with a repeated Unicode property
49 test, for example, \p{Nd}+, there was the theoretical possibility of
50 setting up an incorrect bitmap of starting bytes, but fortunately it could
51 not have actually happened in practice until change 8 above was made (it
52 added property types that matched character-matching opcodes).
53
54 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
55 possible starting bytes for non-anchored patterns.
56
57 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
58 \R, and also a number of cases that involve Unicode properties, both
59 explicit and implicit when PCRE_UCP is set.
60
61 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
62 input, it could crash or give wrong results if characters with values
63 greater than 0xc0 were present in the subject string. (Detail: it assumed
64 UTF-8 input when processing these items.)
65
66 16. Added a lot of (int) casts to avoid compiler warnings in systems where
67 size_t is 64-bit (#991).
68
69 17. Added a check for running out of memory when PCRE is compiled with
70 --disable-stack-for-recursion (#990).
71
72 18. If the last data line in a file for pcretest does not have a newline on
73 the end, a newline was missing in the output.
74
75
76
77 Version 8.02 19-Mar-2010
78 ------------------------
79
80 1. The Unicode data tables have been updated to Unicode 5.2.0.
81
82 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
83 configured.
84
85 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
86 original author of that file, following a query about its status.
87
88 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
89 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
90
91 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
92 quantifier applied to a forward-referencing subroutine call, could compile
93 incorrect code or give the error "internal error: previously-checked
94 referenced subpattern not found".
95
96 6. Both MS Visual Studio and Symbian OS have problems with initializing
97 variables to point to external functions. For these systems, therefore,
98 pcre_malloc etc. are now initialized to local functions that call the
99 relevant global functions.
100
101 7. There were two entries missing in the vectors called coptable and poptable
102 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
103 I've fixed the data, and added a kludgy way of testing at compile time that
104 the lengths are correct (equal to the number of opcodes).
105
106 8. Following on from 7, I added a similar kludge to check the length of the
107 eint vector in pcreposix.c.
108
109 9. Error texts for pcre_compile() are held as one long string to avoid too
110 much relocation at load time. To find a text, the string is searched,
111 counting zeros. There was no check for running off the end of the string,
112 which could happen if a new error number was added without updating the
113 string.
114
115 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
116
117 11. \K was not working if it appeared in an atomic group or in a group that
118 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
119 \K is "not well defined" if used in an assertion. PCRE now accepts it if
120 the assertion is positive, but not if it is negative.
121
122 12. Change 11 fortuitously reduced the size of the stack frame used in the
123 "match()" function of pcre_exec.c by one pointer. Forthcoming
124 implementation of support for (*MARK) will need an extra pointer on the
125 stack; I have reserved it now, so that the stack frame size does not
126 decrease.
127
128 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
129 item in branch that calls a recursion is a subroutine call - as in the
130 second branch in the above example - was incorrectly given the compile-
131 time error "recursive call could loop indefinitely" because pcre_compile()
132 was not correctly checking the subroutine for matching a non-empty string.
133
134 14. The checks for overrunning compiling workspace could trigger after an
135 overrun had occurred. This is a "should never occur" error, but it can be
136 triggered by pathological patterns such as hundreds of nested parentheses.
137 The checks now trigger 100 bytes before the end of the workspace.
138
139 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
140
141
142 Version 8.01 19-Jan-2010
143 ------------------------
144
145 1. If a pattern contained a conditional subpattern with only one branch (in
146 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
147 computed the wrong minimum data length (which is of course zero for such
148 subpatterns). This could cause incorrect "no match" results.
149
150 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
151 the pattern is reset in the first branch, pcre_compile() failed with
152 "internal error: code overflow at offset...". This happened only when
153 the reset was to the original external option setting. (An optimization
154 abstracts leading options settings into an external setting, which was the
155 cause of this.)
156
157 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
158 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
159 assertion pattern did not match (meaning that the assertion was true), it
160 was incorrectly treated as false if the SKIP had been reached during the
161 matching. This also applied to assertions used as conditions.
162
163 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
164 assertion subpattern, including such a pattern used as a condition,
165 unpredictable results occurred, instead of the error return
166 PCRE_ERROR_DFA_UITEM.
167
168 5. The C++ GlobalReplace function was not working like Perl for the special
169 situation when an empty string is matched. It now does the fancy magic
170 stuff that is necessary.
171
172 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
173 removed. (These were left over from very, very early versions of PCRE.)
174
175 7. Some cosmetic changes to the code to make life easier when compiling it
176 as part of something else:
177
178 (a) Change DEBUG to PCRE_DEBUG.
179
180 (b) In pcre_compile(), rename the member of the "branch_chain" structure
181 called "current" as "current_branch", to prevent a collision with the
182 Linux macro when compiled as a kernel module.
183
184 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
185 prevent a collision with the Linux macro when compiled as a kernel
186 module.
187
188 8. In pcre_compile() there are some checks for integer overflows that used to
189 cast potentially large values to (double). This has been changed to that
190 when building, a check for int64_t is made, and if it is found, it is used
191 instead, thus avoiding the use of floating point arithmetic. (There is no
192 other use of FP in PCRE.) If int64_t is not found, the fallback is to
193 double.
194
195 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
196 2005 (difference between two addresses compared to an unsigned value).
197
198 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
199 custom one, because of the following reported problem in Windows:
200
201 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
202 under Win32.
203 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
204 therefore missing the function definition.
205 - The compiler thus generates a "C" signature for the test function.
206 - The linker fails to find the "C" function.
207 - PCRE fails to configure if asked to do so against libbz2.
208
209 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
210 messages were output:
211
212 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
213 rerunning libtoolize, to keep the correct libtool macros in-tree.
214 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
215
216 I have done both of these things.
217
218 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
219 most of the time, it *can* run out if it is given a pattern that contains a
220 runaway infinite recursion. I updated the discussion in the pcrestack man
221 page.
222
223 13. Now that we have gone to the x.xx style of version numbers, the minor
224 version may start with zero. Using 08 or 09 is a bad idea because users
225 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
226 interpreted as invalid octal numbers. I've updated the previous comment in
227 configure.ac, and also added a check that gives an error if 08 or 09 are
228 used.
229
230 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
231 causing partial matching to fail when the end of the subject matched \W
232 in a UTF-8 pattern where \W was quantified with a minimum of 3.
233
234 15. There were some discrepancies between the declarations in pcre_internal.h
235 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
236 their definitions. The declarations used "const uschar *" and the
237 definitions used USPTR. Even though USPTR is normally defined as "const
238 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
239 reported that: "This difference in casting confuses some C++ compilers, for
240 example, SunCC recognizes above declarations as different functions and
241 generates broken code for hbpcre." I have changed the declarations to use
242 USPTR.
243
244 16. GNU libtool is named differently on some systems. The autogen.sh script now
245 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
246 (FreeBSD).
247
248 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
249 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
250 comment: "Figure out how to create a longlong from a string: strtoll and
251 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
252 instance, but it only takes 2 args instead of 3!"
253
254 18. A subtle bug concerned with back references has been fixed by a change of
255 specification, with a corresponding code fix. A pattern such as
256 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
257 refers, was giving matches when it shouldn't. For example, xa=xaaa would
258 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
259 same bug. Such groups have to be quantified to be useful, or contained
260 inside another quantified group. (If there's no repetition, the reference
261 can never match.) The problem arises because, having left the group and
262 moved on to the rest of the pattern, a later failure that backtracks into
263 the group uses the captured value from the final iteration of the group
264 rather than the correct earlier one. I have fixed this in PCRE by forcing
265 any group that contains a reference to itself to be an atomic group; that
266 is, there cannot be any backtracking into it once it has completed. This is
267 similar to recursive and subroutine calls.
268
269
270 Version 8.00 19-Oct-09
271 ----------------------
272
273 1. The table for translating pcre_compile() error codes into POSIX error codes
274 was out-of-date, and there was no check on the pcre_compile() error code
275 being within the table. This could lead to an OK return being given in
276 error.
277
278 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
279 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
280 in a Windows environment.
281
282 3. The pcregrep --count option prints the count for each file even when it is
283 zero, as does GNU grep. However, pcregrep was also printing all files when
284 --files-with-matches was added. Now, when both options are given, it prints
285 counts only for those files that have at least one match. (GNU grep just
286 prints the file name in this circumstance, but including the count seems
287 more useful - otherwise, why use --count?) Also ensured that the
288 combination -clh just lists non-zero counts, with no names.
289
290 4. The long form of the pcregrep -F option was incorrectly implemented as
291 --fixed_strings instead of --fixed-strings. This is an incompatible change,
292 but it seems right to fix it, and I didn't think it was worth preserving
293 the old behaviour.
294
295 5. The command line items --regex=pattern and --regexp=pattern were not
296 recognized by pcregrep, which required --regex pattern or --regexp pattern
297 (with a space rather than an '='). The man page documented the '=' forms,
298 which are compatible with GNU grep; these now work.
299
300 6. No libpcreposix.pc file was created for pkg-config; there was just
301 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
302
303 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
304 when UCP support is not needed, by modifying the Python script that
305 generates it from Unicode data files. This should not matter if the module
306 is correctly used as a library, but I received one complaint about 50K of
307 unwanted data. My guess is that the person linked everything into his
308 program rather than using a library. Anyway, it does no harm.
309
310 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
311 was a minimum greater than 1 for a wide character in a possessive
312 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
313 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
314 character. Chaos in the form of incorrect output or a compiling loop could
315 result.
316
317 9. The restrictions on what a pattern can contain when partial matching is
318 requested for pcre_exec() have been removed. All patterns can now be
319 partially matched by this function. In addition, if there are at least two
320 slots in the offset vector, the offset of the earliest inspected character
321 for the match and the offset of the end of the subject are set in them when
322 PCRE_ERROR_PARTIAL is returned.
323
324 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
325 synonymous with PCRE_PARTIAL, for backwards compatibility, and
326 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
327 and may be more useful for multi-segment matching.
328
329 11. Partial matching with pcre_exec() is now more intuitive. A partial match
330 used to be given if ever the end of the subject was reached; now it is
331 given only if matching could not proceed because another character was
332 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
333 string "Z", which now yields "no match" instead of "partial match". In the
334 case of pcre_dfa_exec(), "no match" is given if every matching path for the
335 final character ended with (*FAIL).
336
337 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
338 if the pattern had a "must contain" character that was already found in the
339 earlier partial match, unless partial matching was again requested. For
340 example, with the pattern /dog.(body)?/, the "must contain" character is
341 "g". If the first part-match was for the string "dog", restarting with
342 "sbody" failed. This bug has been fixed.
343
344 13. The string returned by pcre_dfa_exec() after a partial match has been
345 changed so that it starts at the first inspected character rather than the
346 first character of the match. This makes a difference only if the pattern
347 starts with a lookbehind assertion or \b or \B (\K is not supported by
348 pcre_dfa_exec()). It's an incompatible change, but it makes the two
349 matching functions compatible, and I think it's the right thing to do.
350
351 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
352 so that the demonstration program is easily available in environments where
353 PCRE has not been installed from source.
354
355 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
356 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
357 library.
358
359 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
360 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
361 is not the first non-POSIX option to be added. Clearly some people find
362 these options useful.
363
364 17. If a caller to the POSIX matching function regexec() passes a non-zero
365 value for nmatch with a NULL value for pmatch, the value of
366 nmatch is forced to zero.
367
368 18. RunGrepTest did not have a test for the availability of the -u option of
369 the diff command, as RunTest does. It now checks in the same way as
370 RunTest, and also checks for the -b option.
371
372 19. If an odd number of negated classes containing just a single character
373 interposed, within parentheses, between a forward reference to a named
374 subpattern and the definition of the subpattern, compilation crashed with
375 an internal error, complaining that it could not find the referenced
376 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
377 [The bug was that it was starting one character too far in when skipping
378 over the character class, thus treating the ] as data rather than
379 terminating the class. This meant it could skip too much.]
380
381 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
382 /g option in pcretest when the pattern contains \K, which makes it possible
383 to have an empty string match not at the start, even when the pattern is
384 anchored. Updated pcretest and pcredemo to use this option.
385
386 21. If the maximum number of capturing subpatterns in a recursion was greater
387 than the maximum at the outer level, the higher number was returned, but
388 with unset values at the outer level. The correct (outer level) value is
389 now given.
390
391 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
392 PCRE did not set those parentheses (unlike Perl). I have now found a way to
393 make it do so. The string so far is captured, making this feature
394 compatible with Perl.
395
396 23. The tests have been re-organized, adding tests 11 and 12, to make it
397 possible to check the Perl 5.10 features against Perl 5.10.
398
399 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
400 pattern matches a fixed length string. PCRE did not allow this; now it
401 does. Neither allows recursion.
402
403 25. I finally figured out how to implement a request to provide the minimum
404 length of subject string that was needed in order to match a given pattern.
405 (It was back references and recursion that I had previously got hung up
406 on.) This code has now been added to pcre_study(); it finds a lower bound
407 to the length of subject needed. It is not necessarily the greatest lower
408 bound, but using it to avoid searching strings that are too short does give
409 some useful speed-ups. The value is available to calling programs via
410 pcre_fullinfo().
411
412 26. While implementing 25, I discovered to my embarrassment that pcretest had
413 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
414 study optimizations had never been tested with that matching function.
415 Oops. What is worse, even when it was passed study data, there was a bug in
416 pcre_dfa_exec() that meant it never actually used it. Double oops. There
417 were also very few tests of studied patterns with pcre_dfa_exec().
418
419 27. If (?| is used to create subpatterns with duplicate numbers, they are now
420 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
421 on the other side of the coin, they are no longer allowed to have different
422 names, because these cannot be distinguished in PCRE, and this has caused
423 confusion. (This is a difference from Perl.)
424
425 28. When duplicate subpattern names are present (necessarily with different
426 numbers, as required by 27 above), and a test is made by name in a
427 conditional pattern, either for a subpattern having been matched, or for
428 recursion in such a pattern, all the associated numbered subpatterns are
429 tested, and the overall condition is true if the condition is true for any
430 one of them. This is the way Perl works, and is also more like the way
431 testing by number works.
432
433
434 Version 7.9 11-Apr-09
435 ---------------------
436
437 1. When building with support for bzlib/zlib (pcregrep) and/or readline
438 (pcretest), all targets were linked against these libraries. This included
439 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
440 libraries. This caused unwanted dependencies to be created. This problem
441 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
442 pcretest is linked with readline.
443
444 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
445 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
446 moved outside it again, because FALSE and TRUE are already defined in AIX,
447 but BOOL is not.
448
449 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
450 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
451
452 4. The pcregrep documentation said spaces were inserted as well as colons (or
453 hyphens) following file names and line numbers when outputting matching
454 lines. This is not true; no spaces are inserted. I have also clarified the
455 wording for the --colour (or --color) option.
456
457 5. In pcregrep, when --colour was used with -o, the list of matching strings
458 was not coloured; this is different to GNU grep, so I have changed it to be
459 the same.
460
461 6. When --colo(u)r was used in pcregrep, only the first matching substring in
462 each matching line was coloured. Now it goes on to look for further matches
463 of any of the test patterns, which is the same behaviour as GNU grep.
464
465 7. A pattern that could match an empty string could cause pcregrep to loop; it
466 doesn't make sense to accept an empty string match in pcregrep, so I have
467 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
468 seems to be how GNU grep behaves.
469
470 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
471 start or after a newline", because the conditional assertion was not being
472 correctly handled. The rule now is that both the assertion and what follows
473 in the first alternative must satisfy the test.
474
475 9. If auto-callout was enabled in a pattern with a conditional group whose
476 condition was an assertion, PCRE could crash during matching, both with
477 pcre_exec() and pcre_dfa_exec().
478
479 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
480 used for matching.
481
482 11. Unicode property support in character classes was not working for
483 characters (bytes) greater than 127 when not in UTF-8 mode.
484
485 12. Added the -M command line option to pcretest.
486
487 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
488
489 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
490
491 16. Added comments and documentation about mis-use of no_arg in the C++
492 wrapper.
493
494 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
495 from Martin Jerabek that uses macro names for all relevant character and
496 string constants.
497
498 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
499 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
500 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
501 these, but not everybody uses configure.
502
503 19. A conditional group that had only one branch was not being correctly
504 recognized as an item that could match an empty string. This meant that an
505 enclosing group might also not be so recognized, causing infinite looping
506 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
507 with the subject "ab", where knowledge that the repeated group can match
508 nothing is needed in order to break the loop.
509
510 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
511 exec(), but without supplying a callout function, matching went wrong.
512
513 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
514 leak if the size of the offset vector was greater than 30. When the vector
515 is smaller, the saved offsets during recursion go onto a local stack
516 vector, but for larger vectors malloc() is used. It was failing to free
517 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
518 error, in fact).
519
520 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
521 heapframe that is used only when UTF-8 support is enabled. This caused no
522 problem, but was untidy.
523
524 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
525 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
526 included within another project.
527
528 24. Steven Van Ingelgem's patches to add more options to the CMake support,
529 slightly modified by me:
530
531 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
532 not building pcregrep.
533
534 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
535 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
536
537 25. Forward references, both numeric and by name, in patterns that made use of
538 duplicate group numbers, could behave incorrectly or give incorrect errors,
539 because when scanning forward to find the reference group, PCRE was not
540 taking into account the duplicate group numbers. A pattern such as
541 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
542
543 26. Changed a few more instances of "const unsigned char *" to USPTR, making
544 the feature of a custom pointer more persuasive (as requested by a user).
545
546 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
547 pcretest.c, inside #ifndefs, because it seems they are sometimes already
548 pre-defined.
549
550 28. Added support for (*UTF8) at the start of a pattern.
551
552 29. Arrange for flags added by the "release type" setting in CMake to be shown
553 in the configuration summary.
554
555
556 Version 7.8 05-Sep-08
557 ---------------------
558
559 1. Replaced UCP searching code with optimized version as implemented for Ad
560 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
561 stage table and inline lookup instead of a function, giving speed ups of 2
562 to 5 times on some simple patterns that I tested. Permission was given to
563 distribute the MultiStage2.py script that generates the tables (it's not in
564 the tarball, but is in the Subversion repository).
565
566 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
567 scripts.
568
569 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
570 a group with a zero qualifier. The result of the study could be incorrect,
571 or the function might crash, depending on the pattern.
572
573 4. Caseless matching was not working for non-ASCII characters in back
574 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
575 It now works when Unicode Property Support is available.
576
577 5. In pcretest, an escape such as \x{de} in the data was always generating
578 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
579 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
580 truncation.
581
582 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
583
584 7. Added two (int) casts to pcregrep when printing the difference of two
585 pointers, in case they are 64-bit values.
586
587 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
588 test 2 if it fails.
589
590 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
591 and a #define of that name to empty if it is not externally set. This is to
592 allow users of MSVC to set it if necessary.
593
594 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
595 the convenience functions in the pcre_get.c source file.
596
597 11. An option change at the start of a pattern that had top-level alternatives
598 could cause overwriting and/or a crash. This command provoked a crash in
599 some environments:
600
601 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
602
603 This potential security problem was recorded as CVE-2008-2371.
604
605 12. For a pattern where the match had to start at the beginning or immediately
606 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
607 pcre_dfa_exec() could read past the end of the passed subject if there was
608 no match. To help with detecting such bugs (e.g. with valgrind), I modified
609 pcretest so that it places the subject at the end of its malloc-ed buffer.
610
611 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
612 exec() might read past the end of the data buffer in UTF-8 mode.
613
614 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
615 the data contained the byte 0x85 as part of a UTF-8 character within its
616 first line. This applied both to normal and DFA matching.
617
618 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
619 /^[^d]*?$/8 failed to match "abc".
620
621 16. Added a missing copyright notice to pcrecpp_internal.h.
622
623 17. Make it more clear in the documentation that values returned from
624 pcre_exec() in ovector are byte offsets, not character counts.
625
626 18. Tidied a few places to stop certain compilers from issuing warnings.
627
628 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
629 supplied by Stefan Weber. I made a further small update for 7.8 because
630 there is a change of source arrangements: the pcre_searchfuncs.c module is
631 replaced by pcre_ucd.c.
632
633
634 Version 7.7 07-May-08
635 ---------------------
636
637 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
638 a string to a long long, pretend we don't even have a long long." This is
639 done by checking for the strtoq, strtoll, and _strtoi64 functions.
640
641 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
642 pre-7.6 versions, which defined a global no_arg variable instead of putting
643 it in the RE class. (See also #8 below.)
644
645 3. Remove a line of dead code, identified by coverity and reported by Nuno
646 Lopes.
647
648 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
649
650 (1) The include/exclude patterns were being applied to the whole pathnames
651 of files, instead of just to the final components.
652
653 (2) If there was more than one level of directory, the subdirectories were
654 skipped unless they satisfied the include/exclude conditions. This is
655 inconsistent with GNU grep (and could even be seen as contrary to the
656 pcregrep specification - which I improved to make it absolutely clear).
657 The action now is always to scan all levels of directory, and just
658 apply the include/exclude patterns to regular files.
659
660 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
661 --exclude_dir in the tests to avoid scanning .svn directories.
662
663 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
664 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
665 doesn't support NULs in patterns.
666
667 7. Added some missing "const"s to declarations of static tables in
668 pcre_compile.c and pcre_dfa_exec.c.
669
670 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
671 caused by fix #2 above. (Subsequently also a second patch to fix the
672 first patch. And a third patch - this was a messy problem.)
673
674 9. Applied Craig's patch to remove the use of push_back().
675
676 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
677 matching function regexec().
678
679 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
680 which, however, unlike Perl's \g{...}, are subroutine calls, not back
681 references. PCRE supports relative numbers with this syntax (I don't think
682 Oniguruma does).
683
684 12. Previously, a group with a zero repeat such as (...){0} was completely
685 omitted from the compiled regex. However, this means that if the group
686 was called as a subroutine from elsewhere in the pattern, things went wrong
687 (an internal error was given). Such groups are now left in the compiled
688 pattern, with a new opcode that causes them to be skipped at execution
689 time.
690
691 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
692 to the way PCRE behaves:
693
694 (a) A lone ] character is dis-allowed (Perl treats it as data).
695
696 (b) A back reference to an unmatched subpattern matches an empty string
697 (Perl fails the current match path).
698
699 (c) A data ] in a character class must be notated as \] because if the
700 first data character in a class is ], it defines an empty class. (In
701 Perl it is not possible to have an empty class.) The empty class []
702 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
703 The negative empty class [^] matches any one character, independently
704 of the DOTALL setting.
705
706 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
707 non-existent subpattern following a character class starting with ']' and
708 containing () gave an internal compiling error instead of "reference to
709 non-existent subpattern". Fortunately, when the pattern did exist, the
710 compiled code was correct. (When scanning forwards to check for the
711 existencd of the subpattern, it was treating the data ']' as terminating
712 the class, so got the count wrong. When actually compiling, the reference
713 was subsequently set up correctly.)
714
715 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
716 it was being rejected as not supported by pcre_dfa_exec(), even though
717 other assertions are supported. I have made pcre_dfa_exec() support
718 (*FAIL).
719
720 16. The implementation of 13c above involved the invention of a new opcode,
721 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
722 cannot be changed at match time, I realized I could make a small
723 improvement to matching performance by compiling OP_ALLANY instead of
724 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
725 on the OP_ANY path.
726
727 17. Compiling pcretest on Windows with readline support failed without the
728 following two fixes: (1) Make the unistd.h include conditional on
729 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
730
731 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
732 ncurses library to be included for pcretest when ReadLine support is
733 requested, but also to allow for it to be overridden. This patch came from
734 Daniel Bergström.
735
736 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
737 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
738 any errors with the current Unicode tables. Thanks to Peter Kankowski for
739 spotting this.
740
741
742 Version 7.6 28-Jan-08
743 ---------------------
744
745 1. A character class containing a very large number of characters with
746 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
747 overflow.
748
749 2. Patch to cut out the "long long" test in pcrecpp_unittest when
750 HAVE_LONG_LONG is not defined.
751
752 3. Applied Christian Ehrlicher's patch to update the CMake build files to
753 bring them up to date and include new features. This patch includes:
754
755 - Fixed PH's badly added libz and libbz2 support.
756 - Fixed a problem with static linking.
757 - Added pcredemo. [But later removed - see 7 below.]
758 - Fixed dftables problem and added an option.
759 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
760 HAVE_LONG_LONG.
761 - Added readline support for pcretest.
762 - Added an listing of the option settings after cmake has run.
763
764 4. A user submitted a patch to Makefile that makes it easy to create
765 "pcre.dll" under mingw when using Configure/Make. I added stuff to
766 Makefile.am that cause it to include this special target, without
767 affecting anything else. Note that the same mingw target plus all
768 the other distribution libraries and programs are now supported
769 when configuring with CMake (see 6 below) instead of with
770 Configure/Make.
771
772 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
773 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
774 exported in the Windows port". It has not yet been confirmed that the patch
775 solves the problem, but it does no harm.
776
777 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
778 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
779 with CMake, and also correct the comment about stack recursion.
780
781 7. Remove the automatic building of pcredemo from the ./configure system and
782 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
783 of a program that users should build themselves after PCRE is installed, so
784 building it automatically is not really right. What is more, it gave
785 trouble in some build environments.
786
787 8. Further tidies to CMakeLists.txt from Sheri and Christian.
788
789
790 Version 7.5 10-Jan-08
791 ---------------------
792
793 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
794 values in parens when parsing an RE using the C++ wrapper."
795
796 2. Negative specials like \S did not work in character classes in UTF-8 mode.
797 Characters greater than 255 were excluded from the class instead of being
798 included.
799
800 3. The same bug as (2) above applied to negated POSIX classes such as
801 [:^space:].
802
803 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
804 defined or documented. It seems to have been a typo for PCRE_STATIC, so
805 I have changed it.
806
807 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
808 first named subpattern) and a construct such as (?&a) would reference the
809 first named subpattern whose name started with "a" (in other words, the
810 length check was missing). Both these problems are fixed. "Subpattern name
811 expected" is now given for (?&) (a zero-length name), and this patch also
812 makes it give the same error for \k'' (previously it complained that that
813 was a reference to a non-existent subpattern).
814
815 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
816 this is right because (?- can be followed by option settings as well as by
817 digits. I have, however, made the messages clearer.
818
819 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
820 than the number used in the conditional) now cause a compile-time error.
821 This is actually not compatible with Perl, which accepts such patterns, but
822 treats the conditional as always being FALSE (as PCRE used to), but it
823 seems to me that giving a diagnostic is better.
824
825 8. Change "alphameric" to the more common word "alphanumeric" in comments
826 and messages.
827
828 9. Fix two occurrences of "backslash" in comments that should have been
829 "backspace".
830
831 10. Remove two redundant lines of code that can never be obeyed (their function
832 was moved elsewhere).
833
834 11. The program that makes PCRE's Unicode character property table had a bug
835 which caused it to generate incorrect table entries for sequences of
836 characters that have the same character type, but are in different scripts.
837 It amalgamated them into a single range, with the script of the first of
838 them. In other words, some characters were in the wrong script. There were
839 thirteen such cases, affecting characters in the following ranges:
840
841 U+002b0 - U+002c1
842 U+0060c - U+0060d
843 U+0061e - U+00612
844 U+0064b - U+0065e
845 U+0074d - U+0076d
846 U+01800 - U+01805
847 U+01d00 - U+01d77
848 U+01d9b - U+01dbf
849 U+0200b - U+0200f
850 U+030fc - U+030fe
851 U+03260 - U+0327f
852 U+0fb46 - U+0fbb1
853 U+10450 - U+1049d
854
855 12. The -o option (show only the matching part of a line) for pcregrep was not
856 compatible with GNU grep in that, if there was more than one match in a
857 line, it showed only the first of them. It now behaves in the same way as
858 GNU grep.
859
860 13. If the -o and -v options were combined for pcregrep, it printed a blank
861 line for every non-matching line. GNU grep prints nothing, and pcregrep now
862 does the same. The return code can be used to tell if there were any
863 non-matching lines.
864
865 14. Added --file-offsets and --line-offsets to pcregrep.
866
867 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
868 infinitely looping recursion. The bug was that positive lookaheads were not
869 being skipped when checking for a possible empty match (negative lookaheads
870 and both kinds of lookbehind were skipped).
871
872 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
873 inclusion of <windows.h> to before rather than after the definition of
874 INVALID_FILE_ATTRIBUTES (patch from David Byron).
875
876 17. Specifying a possessive quantifier with a specific limit for a Unicode
877 character property caused pcre_compile() to compile bad code, which led at
878 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
879 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
880 caused the error; without that there was no problem.
881
882 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
883
884 19. Added --enable-pcretest-libreadline.
885
886 20. In pcrecpp.cc, the variable 'count' was incremented twice in
887 RE::GlobalReplace(). As a result, the number of replacements returned was
888 double what it should be. I removed one of the increments, but Craig sent a
889 later patch that removed the other one (the right fix) and added unit tests
890 that check the return values (which was not done before).
891
892 21. Several CMake things:
893
894 (1) Arranged that, when cmake is used on Unix, the libraries end up with
895 the names libpcre and libpcreposix, not just pcre and pcreposix.
896
897 (2) The above change means that pcretest and pcregrep are now correctly
898 linked with the newly-built libraries, not previously installed ones.
899
900 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
901
902 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
903 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
904 UTF-8 newline character). The key issue is that the pattern starts .*;
905 this means that the match must be either at the beginning, or after a
906 newline. The bug was in the code for advancing after a failed match and
907 checking that the new position followed a newline. It was not taking
908 account of UTF-8 characters correctly.
909
910 23. PCRE was behaving differently from Perl in the way it recognized POSIX
911 character classes. PCRE was not treating the sequence [:...:] as a
912 character class unless the ... were all letters. Perl, however, seems to
913 allow any characters between [: and :], though of course it rejects as
914 unknown any "names" that contain non-letters, because all the known class
915 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
916 for example, whereas PCRE did not - it did not recognize a POSIX character
917 class. This seemed a bit dangerous, so the code has been changed to be
918 closer to Perl. The behaviour is not identical to Perl, because PCRE will
919 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
920 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
921 Perl does, and where it didn't before.
922
923 24. Rewrite so as to remove the single use of %n from pcregrep because in some
924 Windows environments %n is disabled by default.
925
926
927 Version 7.4 21-Sep-07
928 ---------------------
929
930 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
931 means that a class such as [\s] counted as "explicit reference to CR or
932 LF". That isn't really right - the whole point of the change was to try to
933 help when there was an actual mention of one of the two characters. So now
934 the change happens only if \r or \n (or a literal CR or LF) character is
935 encountered.
936
937 2. The 32-bit options word was also used for 6 internal flags, but the numbers
938 of both had grown to the point where there were only 3 bits left.
939 Fortunately, there was spare space in the data structure, and so I have
940 moved the internal flags into a new 16-bit field to free up more option
941 bits.
942
943 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
944 but did not set the internal JCHANGED flag - either of these is enough to
945 control the way the "get" function works - but the PCRE_INFO_JCHANGED
946 facility is supposed to tell if (?J) was ever used, so now (?J) at the
947 start sets both bits.
948
949 4. Added options (at build time, compile time, exec time) to change \R from
950 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
951
952 5. doc/pcresyntax.html was missing from the distribution.
953
954 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
955 compatibility, even though it is no longer used.
956
957 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
958 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
959 windows.h file is present (where different names are used). [This was
960 reversed later after testing - see 16 below.]
961
962 8. Changed all #include <config.h> to #include "config.h". There were also
963 some further <pcre.h> cases that I changed to "pcre.h".
964
965 9. When pcregrep was used with the --colour option, it missed the line ending
966 sequence off the lines that it output.
967
968 10. It was pointed out to me that arrays of string pointers cause lots of
969 relocations when a shared library is dynamically loaded. A technique of
970 using a single long string with a table of offsets can drastically reduce
971 these. I have refactored PCRE in four places to do this. The result is
972 dramatic:
973
974 Originally: 290
975 After changing UCP table: 187
976 After changing error message table: 43
977 After changing table of "verbs" 36
978 After changing table of Posix names 22
979
980 Thanks to the folks working on Gregex for glib for this insight.
981
982 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
983 unicode-properties was also set.
984
985 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
986
987 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
988 checked only for CRLF.
989
990 14. Added casts to pcretest.c to avoid compiler warnings.
991
992 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
993
994 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
995 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
996 entirely. This removes changes made in 7 above.
997
998 17. The CMake files have been updated, and there is now more information about
999 building with CMake in the NON-UNIX-USE document.
1000
1001
1002 Version 7.3 28-Aug-07
1003 ---------------------
1004
1005 1. In the rejigging of the build system that eventually resulted in 7.1, the
1006 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1007 brackets there is not right, since it causes compilers to look for an
1008 installed pcre.h, not the version that is in the source that is being
1009 compiled (which of course may be different). I have changed it back to:
1010
1011 #include "pcre.h"
1012
1013 I have a vague recollection that the change was concerned with compiling in
1014 different directories, but in the new build system, that is taken care of
1015 by the VPATH setting the Makefile.
1016
1017 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1018 when the subject happened to end in the byte 0x85 (e.g. if the last
1019 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1020 characters but of course it shouldn't be taken as a newline when it is part
1021 of another character. The bug was that, for an unlimited repeat of . in
1022 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1023 characters when looking for a newline.
1024
1025 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1026
1027 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1028 in debug output.
1029
1030 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1031 long printing in the pcrecpp unittest when running under MinGW.
1032
1033 6. ESC_K was left out of the EBCDIC table.
1034
1035 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1036 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1037 limit also applies to "virtual nesting" when a pattern is recursive, and in
1038 this case 1000 isn't so big. I have been able to remove this limit at the
1039 expense of backing off one optimization in certain circumstances. Normally,
1040 when pcre_exec() would call its internal match() function recursively and
1041 immediately return the result unconditionally, it uses a "tail recursion"
1042 feature to save stack. However, when a subpattern that can match an empty
1043 string has an unlimited repetition quantifier, it no longer makes this
1044 optimization. That gives it a stack frame in which to save the data for
1045 checking that an empty string has been matched. Previously this was taken
1046 from the 1000-entry workspace that had been reserved. So now there is no
1047 explicit limit, but more stack is used.
1048
1049 8. Applied Daniel's patches to solve problems with the import/export magic
1050 syntax that is required for Windows, and which was going wrong for the
1051 pcreposix and pcrecpp parts of the library. These were overlooked when this
1052 problem was solved for the main library.
1053
1054 9. There were some crude static tests to avoid integer overflow when computing
1055 the size of patterns that contain repeated groups with explicit upper
1056 limits. As the maximum quantifier is 65535, the maximum group length was
1057 set at 30,000 so that the product of these two numbers did not overflow a
1058 32-bit integer. However, it turns out that people want to use groups that
1059 are longer than 30,000 bytes (though not repeat them that many times).
1060 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1061 made it possible to implement the integer overflow checks in a much more
1062 dynamic way, which I have now done. The artificial limitation on group
1063 length has been removed - we now have only the limit on the total length of
1064 the compiled pattern, which depends on the LINK_SIZE setting.
1065
1066 10. Fixed a bug in the documentation for get/copy named substring when
1067 duplicate names are permitted. If none of the named substrings are set, the
1068 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1069 empty string.
1070
1071 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1072 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1073 because the ] is interpreted as the first data character and the
1074 terminating ] is not found. PCRE has been made compatible with Perl in this
1075 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1076 cause memory overwriting.
1077
1078 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1079 string has been matched (to stop an infinite loop). It was not recognizing
1080 a conditional subpattern that could match an empty string if that
1081 subpattern was within another subpattern. For example, it looped when
1082 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1083 condition was not nested. This bug has been fixed.
1084
1085 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1086 past the start of the subject in the presence of bytes with the top bit
1087 set, for example "\x8aBCD".
1088
1089 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1090 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1091
1092 14. Optimized (?!) to (*FAIL).
1093
1094 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1095 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1096 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1097 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1098 does: it's just the validity check that is more restrictive.
1099
1100 16. Inserted checks for integer overflows during escape sequence (backslash)
1101 processing, and also fixed erroneous offset values for syntax errors during
1102 backslash processing.
1103
1104 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1105 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1106
1107 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1108 caused an overrun.
1109
1110 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1111 something other than just ASCII characters) inside a group that had an
1112 unlimited repeat caused a loop at compile time (while checking to see
1113 whether the group could match an empty string).
1114
1115 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1116 [\P{Any}] did so. (Error in the code for printing property names.)
1117
1118 21. An orphan \E inside a character class could cause a crash.
1119
1120 22. A repeated capturing bracket such as (A)? could cause a wild memory
1121 reference during compilation.
1122
1123 23. There are several functions in pcre_compile() that scan along a compiled
1124 expression for various reasons (e.g. to see if it's fixed length for look
1125 behind). There were bugs in these functions when a repeated \p or \P was
1126 present in the pattern. These operators have additional parameters compared
1127 with \d, etc, and these were not being taken into account when moving along
1128 the compiled data. Specifically:
1129
1130 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1131 length.
1132
1133 (b) An item such as \pL+ within a repeated group could cause crashes or
1134 loops.
1135
1136 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1137 "reference to non-existent subpattern" error.
1138
1139 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1140
1141 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1142 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1143
1144 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1145
1146 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1147 character were causing crashes (broken optimization).
1148
1149 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1150 \p or \P) caused a compile-time loop.
1151
1152 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1153 break. For example, the unstudied pattern [\r\n]A does not match the string
1154 "\r\nA" because change 7.0/46 below moves the current point on by two
1155 characters after failing to match at the start. However, the pattern \nA
1156 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1157 the same is true. There doesn't seem any very clean way out of this, but
1158 what I have chosen to do makes the common cases work: PCRE now takes note
1159 of whether there can be an explicit match for \r or \n anywhere in the
1160 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1161 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1162 pattern has explicit CR or LF references.
1163
1164 29. Added (*CR) etc for changing newline setting at start of pattern.
1165
1166
1167 Version 7.2 19-Jun-07
1168 ---------------------
1169
1170 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1171 which is apparently normally available under Windows.
1172
1173 2. Re-jig the pcregrep tests with different newline settings in an attempt
1174 to make them independent of the local environment's newline setting.
1175
1176 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1177
1178 4. Some of the "internals" tests were previously cut out when the link size
1179 was not 2, because the output contained actual offsets. The recent new
1180 "Z" feature of pcretest means that these can be cut out, making the tests
1181 usable with all link sizes.
1182
1183 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1184 stack recursion. This gives a massive performance boost under BSD, but just
1185 a small improvement under Linux. However, it saves one field in the frame
1186 in all cases.
1187
1188 6. Added more features from the forthcoming Perl 5.10:
1189
1190 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1191 recursion call. It refers to the nth most recently opened parentheses.
1192
1193 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1194 to be opened parentheses.
1195
1196 (c) Conditions that refer to capturing parentheses can be specified
1197 relatively, for example, (?(-2)... or (?(+3)...
1198
1199 (d) \K resets the start of the current match so that everything before
1200 is not part of it.
1201
1202 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1203
1204 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1205 reference syntax.
1206
1207 (g) (?| introduces a group in which the numbering of parentheses in each
1208 alternative starts with the same number.
1209
1210 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1211
1212 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1213 PCRE_INFO_JCHANGED.
1214
1215 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1216 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1217 for detecting groups that can match an empty string.
1218
1219 9. A pattern with a very large number of alternatives (more than several
1220 hundred) was running out of internal workspace during the pre-compile
1221 phase, where pcre_compile() figures out how much memory will be needed. A
1222 bit of new cunning has reduced the workspace needed for groups with
1223 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1224 workspace instead of running out of the 4096 that are available.
1225
1226 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1227
1228 11. Applied patch from Google to remove an optimization that didn't quite work.
1229 The report of the bug said:
1230
1231 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1232 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1233 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1234
1235 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1236 it matched the wrong number of bytes.
1237
1238
1239 Version 7.1 24-Apr-07
1240 ---------------------
1241
1242 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1243 that is more "standard", making use of automake and other Autotools. There
1244 is some re-arrangement of the files and adjustment of comments consequent
1245 on this.
1246
1247 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1248 for recursive directory scanning broke on some systems because the files
1249 are not scanned in any specific order and on different systems the order
1250 was different. A call to "sort" has been inserted into RunGrepTest for the
1251 approprate test as a short-term fix. In the longer term there may be an
1252 alternative.
1253
1254 3. I had an email from Eric Raymond about problems translating some of PCRE's
1255 man pages to HTML (despite the fact that I distribute HTML pages, some
1256 people do their own conversions for various reasons). The problems
1257 concerned the use of low-level troff macros .br and .in. I have therefore
1258 removed all such uses from the man pages (some were redundant, some could
1259 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1260 HTML has been updated to handle .nf/.fi and to complain if it encounters
1261 .br or .in.
1262
1263 4. Updated comments in configure.ac that get placed in config.h.in and also
1264 arranged for config.h to be included in the distribution, with the name
1265 config.h.generic, for the benefit of those who have to compile without
1266 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1267
1268 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1269 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1270 makevp.bat for the current PCRE, using the additional files
1271 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1272
1273 6. A Windows user reported a minor discrepancy with test 2, which turned out
1274 to be caused by a trailing space on an input line that had got lost in his
1275 copy. The trailing space was an accident, so I've just removed it.
1276
1277 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1278 that is needed.
1279
1280 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1281 as "const" (a) because they are and (b) because it helps the PHP
1282 maintainers who have recently made a script to detect big data structures
1283 in the php code that should be moved to the .rodata section. I remembered
1284 to update Builducptable as well, so it won't revert if ucptable.h is ever
1285 re-created.
1286
1287 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1288 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1289 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1290 support is not required. This saves 1.5-2K of code, which is important in
1291 some applications.
1292
1293 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1294 so as not to refer to the tables, even though these functions will never be
1295 called when UTF-8 support is disabled. Otherwise there are problems with a
1296 shared library.
1297
1298 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1299
1300 (a) It was defining its arguments as char * instead of void *.
1301
1302 (b) It was assuming that all moves were upwards in memory; this was true
1303 a long time ago when I wrote it, but is no longer the case.
1304
1305 The emulated memove() is provided for those environments that have neither
1306 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1307 is clearly not the case, as these two bugs were recently reported.
1308
1309 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1310 and Detrail to create the HTML documentation, the .txt form of the man
1311 pages, and it removes trailing spaces from listed files. It also creates
1312 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1313 case, it wraps all the #defines with #ifndefs. This script should be run
1314 before "make dist".
1315
1316 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1317 with Unicode property support.
1318
1319 (a) For a maximizing quantifier, if the two different cases of the
1320 character were of different lengths in their UTF-8 codings (there are
1321 some cases like this - I found 11), and the matching function had to
1322 back up over a mixture of the two cases, it incorrectly assumed they
1323 were both the same length.
1324
1325 (b) When PCRE was configured to use the heap rather than the stack for
1326 recursion during matching, it was not correctly preserving the data for
1327 the other case of a UTF-8 character when checking ahead for a match
1328 while processing a minimizing repeat. If the check also involved
1329 matching a wide character, but failed, corruption could cause an
1330 erroneous result when trying to check for a repeat of the original
1331 character.
1332
1333 13. Some tidying changes to the testing mechanism:
1334
1335 (a) The RunTest script now detects the internal link size and whether there
1336 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1337 values substituted by "configure". (The RunGrepTest script already did
1338 this for UTF-8.) The configure.ac script no longer substitutes the
1339 relevant variables.
1340
1341 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1342 with length and offset values. This means that the output is different
1343 for different internal link sizes. Test 2 is skipped for link sizes
1344 other than 2 because of this, bypassing the problem. Unfortunately,
1345 there was also a test in test 3 (the locale tests) that used /B and
1346 failed for link sizes other than 2. Rather than cut the whole test out,
1347 I have added a new /Z option to pcretest that replaces the length and
1348 offset values with spaces. This is now used to make test 3 independent
1349 of link size. (Test 2 will be tidied up later.)
1350
1351 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1352 segmentation fault instead of returning the appropriate error message.
1353
1354 15. In multiline mode when the newline sequence was set to "any", the pattern
1355 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1356 This doesn't seem right; it now treats the CRLF combination as the line
1357 ending, and so does not match in that case. It's only a pattern such as ^$
1358 that would hit this one: something like ^ABC$ would have failed after \r
1359 and then tried again after \r\n.
1360
1361 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1362 in an attempt to make files that differ only in their line terminators
1363 compare equal. This works on Linux.
1364
1365 17. Under certain error circumstances pcregrep might try to free random memory
1366 as it exited. This is now fixed, thanks to valgrind.
1367
1368 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1369 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1370 was because its rules for how to advance for /g after matching an empty
1371 string at the end of a line did not allow for this case. They now check for
1372 it specially.
1373
1374 20. pcretest is supposed to handle patterns and data of any length, by
1375 extending its buffers when necessary. It was getting this wrong when the
1376 buffer for a data line had to be extended.
1377
1378 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1379 CRLF as a newline sequence.
1380
1381 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1382 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1383 I have nevertheless tidied it up.
1384
1385 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1386
1387 24. Added a man page for pcre-config.
1388
1389
1390 Version 7.0 19-Dec-06
1391 ---------------------
1392
1393 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1394 moving to gcc 4.1.1.
1395
1396 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1397 sys/time.h, which is documented as needed for this function. It doesn't
1398 seem to matter on Linux, but it showed up on some releases of OS X.
1399
1400 3. It seems that there are systems where bytes whose values are greater than
1401 127 match isprint() in the "C" locale. The "C" locale should be the
1402 default when a C program starts up. In most systems, only ASCII printing
1403 characters match isprint(). This difference caused the output from pcretest
1404 to vary, making some of the tests fail. I have changed pcretest so that:
1405
1406 (a) When it is outputting text in the compiled version of a pattern, bytes
1407 other than 32-126 are always shown as hex escapes.
1408
1409 (b) When it is outputting text that is a matched part of a subject string,
1410 it does the same, unless a different locale has been set for the match
1411 (using the /L modifier). In this case, it uses isprint() to decide.
1412
1413 4. Fixed a major bug that caused incorrect computation of the amount of memory
1414 required for a compiled pattern when options that changed within the
1415 pattern affected the logic of the preliminary scan that determines the
1416 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1417 that the computed length was too small. The symptoms of this bug were
1418 either the PCRE error "internal error: code overflow" from pcre_compile(),
1419 or a glibc crash with a message such as "pcretest: free(): invalid next
1420 size (fast)". Examples of patterns that provoked this bug (shown in
1421 pcretest format) are:
1422
1423 /(?-x: )/x
1424 /(?x)(?-x: \s*#\s*)/
1425 /((?i)[\x{c0}])/8
1426 /(?i:[\x{c0}])/8
1427
1428 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1429 is now done differently.
1430
1431 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1432 wrapper classes; (b) implement a new function in the C++ scanner that is
1433 more efficient than the old way of doing things because it avoids levels of
1434 recursion in the regex matching; (c) add a paragraph to the documentation
1435 for the FullMatch() function.
1436
1437 6. The escape sequence \n was being treated as whatever was defined as
1438 "newline". Not only was this contrary to the documentation, which states
1439 that \n is character 10 (hex 0A), but it also went horribly wrong when
1440 "newline" was defined as CRLF. This has been fixed.
1441
1442 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1443 was being set to -1 for the "end of line" case (supposedly a value that no
1444 character can have). Though this value is never used (the check for end of
1445 line is "zero bytes in current character"), it caused compiler complaints.
1446 I've changed it to 0xffffffff.
1447
1448 8. In pcre_version.c, the version string was being built by a sequence of
1449 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1450 string (as it is for production releases) called a macro with an empty
1451 argument. The C standard says the result of this is undefined. The gcc
1452 compiler treats it as an empty string (which was what was wanted) but it is
1453 reported that Visual C gives an error. The source has been hacked around to
1454 avoid this problem.
1455
1456 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1457 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1458 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1459 of them did).
1460
1461 10. Originally, pcretest opened its input and output without "b"; then I was
1462 told that "b" was needed in some environments, so it was added for release
1463 5.0 to both the input and output. (It makes no difference on Unix-like
1464 systems.) Later I was told that it is wrong for the input on Windows. I've
1465 now abstracted the modes into two macros, to make it easier to fiddle with
1466 them, and removed "b" from the input mode under Windows.
1467
1468 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1469
1470 12. Added -help and --help to pcretest as an official way of being reminded
1471 of the options.
1472
1473 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1474 and pcrecpp.cc because they annoy compilers at high warning levels.
1475
1476 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1477
1478 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1479 scripts are not C programs :-) and which was not noticed because it works
1480 on Linux.
1481
1482 16. pcretest is supposed to handle any length of pattern and data line (as one
1483 line or as a continued sequence of lines) by extending its input buffer if
1484 necessary. This feature was broken for very long pattern lines, leading to
1485 a string of junk being passed to pcre_compile() if the pattern was longer
1486 than about 50K.
1487
1488 17. I have done a major re-factoring of the way pcre_compile() computes the
1489 amount of memory needed for a compiled pattern. Previously, there was code
1490 that made a preliminary scan of the pattern in order to do this. That was
1491 OK when PCRE was new, but as the facilities have expanded, it has become
1492 harder and harder to keep it in step with the real compile phase, and there
1493 have been a number of bugs (see for example, 4 above). I have now found a
1494 cunning way of running the real compile function in a "fake" mode that
1495 enables it to compute how much memory it would need, while actually only
1496 ever using a few hundred bytes of working memory and without too many
1497 tests of the mode. This should make future maintenance and development
1498 easier. A side effect of this work is that the limit of 200 on the nesting
1499 depth of parentheses has been removed (though this was never a serious
1500 limitation, I suspect). However, there is a downside: pcre_compile() now
1501 runs more slowly than before (30% or more, depending on the pattern). I
1502 hope this isn't a big issue. There is no effect on runtime performance.
1503
1504 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1505 newline (only possible for the last line of a file) and it was a
1506 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1507
1508 19. Added additional timing features to pcretest. (1) The -tm option now times
1509 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1510 separate command line item, by a number that specifies the number of
1511 repeats to use when timing. The default is 50000; this gives better
1512 precision, but takes uncomfortably long for very large patterns.
1513
1514 20. Extended pcre_study() to be more clever in cases where a branch of a
1515 subpattern has no definite first character. For example, (a*|b*)[cd] would
1516 previously give no result from pcre_study(). Now it recognizes that the
1517 first character must be a, b, c, or d.
1518
1519 21. There was an incorrect error "recursive call could loop indefinitely" if
1520 a subpattern (or the entire pattern) that was being tested for matching an
1521 empty string contained only one non-empty item after a nested subpattern.
1522 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1523 incorrectly, because the \d was being skipped in the check.
1524
1525 22. The pcretest program now has a new pattern option /B and a command line
1526 option -b, which is equivalent to adding /B to every pattern. This causes
1527 it to show the compiled bytecode, without the additional information that
1528 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1529 is the same as /B/I).
1530
1531 23. A new optimization is now able automatically to treat some sequences such
1532 as a*b as a*+b. More specifically, if something simple (such as a character
1533 or a simple class like \d) has an unlimited quantifier, and is followed by
1534 something that cannot possibly match the quantified thing, the quantifier
1535 is automatically "possessified".
1536
1537 24. A recursive reference to a subpattern whose number was greater than 39
1538 went wrong under certain circumstances in UTF-8 mode. This bug could also
1539 have affected the operation of pcre_study().
1540
1541 25. Realized that a little bit of performance could be had by replacing
1542 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1543
1544 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1545
1546 27. Possessive quantifiers such as a++ were previously implemented by turning
1547 them into atomic groups such as ($>a+). Now they have their own opcodes,
1548 which improves performance. This includes the automatically created ones
1549 from 23 above.
1550
1551 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1552 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1553 the first matched character to be a colon. This applied both to named and
1554 numbered groups.
1555
1556 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1557
1558 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1559 building PCRE on Windows easier, so I have included it in the distribution.
1560
1561 31. There is now a check in pcretest against a ridiculously large number being
1562 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1563 loop, the loop is abandoned.
1564
1565 32. Forward references to subpatterns in conditions such as (?(2)...) where
1566 subpattern 2 is defined later cause pcre_compile() to search forwards in
1567 the pattern for the relevant set of parentheses. This search went wrong
1568 when there were unescaped parentheses in a character class, parentheses
1569 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1570
1571 33. "Subroutine" calls and backreferences were previously restricted to
1572 referencing subpatterns earlier in the regex. This restriction has now
1573 been removed.
1574
1575 34. Added a number of extra features that are going to be in Perl 5.10. On the
1576 whole, these are just syntactic alternatives for features that PCRE had
1577 previously implemented using the Python syntax or my own invention. The
1578 other formats are all retained for compatibility.
1579
1580 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1581 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1582 also .NET compatible.
1583
1584 (b) A recursion or subroutine call to a named group can now be defined as
1585 (?&name) as well as (?P>name).
1586
1587 (c) A backreference to a named group can now be defined as \k<name> or
1588 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1589 5.10, are also .NET compatible.
1590
1591 (d) A conditional reference to a named group can now use the syntax
1592 (?(<name>) or (?('name') as well as (?(name).
1593
1594 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1595 groups (named and numbered) that are never evaluated inline, but can be
1596 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1597 is always false. There may be only one alternative in such a group.
1598
1599 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1600 as the simple (?(R). The condition is true only if the most recent
1601 recursion is that of the given number or name. It does not search out
1602 through the entire recursion stack.
1603
1604 (g) The escape \gN or \g{N} has been added, where N is a positive or
1605 negative number, specifying an absolute or relative reference.
1606
1607 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1608 some "unreachable code" warnings.
1609
1610 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1611 things, this adds five new scripts.
1612
1613 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1614 There were also incompatibilities regarding the handling of \Q..\E inside
1615 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1616 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1617
1618 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1619 matches an empty string, and forcibly breaks the loop. There were bugs in
1620 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1621 against aaaa the result was just "a" rather than "aaaa", for example. Two
1622 separate and independent bugs (that affected different cases) have been
1623 fixed.
1624
1625 39. Refactored the code to abolish the use of different opcodes for small
1626 capturing bracket numbers. This is a tidy that I avoided doing when I
1627 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1628 The new approach is not only tidier, it makes it possible to reduce the
1629 memory needed to fix the previous bug (38).
1630
1631 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1632 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1633 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1634 mode.
1635
1636 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1637 report.
1638
1639 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1640 copy construction and assignment in the C++ wrapper.
1641
1642 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1643 couple of bugs that could have given wrong results in the "--newline=crlf"
1644 case.
1645
1646 44. Added a number of casts and did some reorganization of signed/unsigned int
1647 variables following suggestions from Dair Grant. Also renamed the variable
1648 "this" as "item" because it is a C++ keyword.
1649
1650 45. Arranged for dftables to add
1651
1652 #include "pcre_internal.h"
1653
1654 to pcre_chartables.c because without it, gcc 4.x may remove the array
1655 definition from the final binary if PCRE is built into a static library and
1656 dead code stripping is activated.
1657
1658 46. For an unanchored pattern, if a match attempt fails at the start of a
1659 newline sequence, and the newline setting is CRLF or ANY, and the next two
1660 characters are CRLF, advance by two characters instead of one.
1661
1662
1663 Version 6.7 04-Jul-06
1664 ---------------------
1665
1666 1. In order to handle tests when input lines are enormously long, pcretest has
1667 been re-factored so that it automatically extends its buffers when
1668 necessary. The code is crude, but this _is_ just a test program. The
1669 default size has been increased from 32K to 50K.
1670
1671 2. The code in pcre_study() was using the value of the re argument before
1672 testing it for NULL. (Of course, in any sensible call of the function, it
1673 won't be NULL.)
1674
1675 3. The memmove() emulation function in pcre_internal.h, which is used on
1676 systems that lack both memmove() and bcopy() - that is, hardly ever -
1677 was missing a "static" storage class specifier.
1678
1679 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1680 containing an extended class (one that cannot be represented by a bitmap
1681 because it contains high-valued characters or Unicode property items, e.g.
1682 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1683 pattern, but PCRE should not loop if you do not (it no longer does).
1684 [Detail: two cases were found: (a) a repeated subpattern containing an
1685 extended class; (b) a recursive reference to a subpattern that followed a
1686 previous extended class. It wasn't skipping over the extended class
1687 correctly when UTF-8 mode was not set.]
1688
1689 5. A negated single-character class was not being recognized as fixed-length
1690 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1691 compile error "lookbehind assertion is not fixed length".
1692
1693 6. The RunPerlTest auxiliary script was showing an unexpected difference
1694 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1695 write a Perl script that can interpret lines of an input file either as
1696 byte characters or as UTF-8, which is what "perltest" was being required to
1697 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1698 can't do is switch easily at run time between having the "use utf8;" pragma
1699 or not. In the end, I fudged it by using the RunPerlTest script to insert
1700 "use utf8;" explicitly for the UTF-8 tests.
1701
1702 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1703 the end of the subject string, contrary to the documentation and to what
1704 Perl does. This was true of both matching functions. Now it matches only at
1705 the start of the subject and immediately after *internal* newlines.
1706
1707 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1708 a pointer to an int instead of a pointer to an unsigned long int. This
1709 caused problems on 64-bit systems.
1710
1711 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1712 instance of the 'standard' template library not being so standard".
1713
1714 10. There was no check on the number of named subpatterns nor the maximum
1715 length of a subpattern name. The product of these values is used to compute
1716 the size of the memory block for a compiled pattern. By supplying a very
1717 long subpattern name and a large number of named subpatterns, the size
1718 computation could be caused to overflow. This is now prevented by limiting
1719 the length of names to 32 characters, and the number of named subpatterns
1720 to 10,000.
1721
1722 11. Subpatterns that are repeated with specific counts have to be replicated in
1723 the compiled pattern. The size of memory for this was computed from the
1724 length of the subpattern and the repeat count. The latter is limited to
1725 65535, but there was no limit on the former, meaning that integer overflow
1726 could in principle occur. The compiled length of a repeated subpattern is
1727 now limited to 30,000 bytes in order to prevent this.
1728
1729 12. Added the optional facility to have named substrings with the same name.
1730
1731 13. Added the ability to use a named substring as a condition, using the
1732 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1733 are numbers (not recommended). Forward references are permitted.
1734
1735 14. Added forward references in named backreferences (if you see what I mean).
1736
1737 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1738 pattern could run off the end of the subject. For example, the pattern
1739 "(?s)(.{1,5})"8 did this with the subject "ab".
1740
1741 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1742 PCRE_CASELESS was set when matching characters that were quantified with ?
1743 or *.
1744
1745 17. A character class other than a single negated character that had a minimum
1746 but no maximum quantifier - for example [ab]{6,} - was not handled
1747 correctly by pce_dfa_exec(). It would match only one character.
1748
1749 18. A valid (though odd) pattern that looked like a POSIX character
1750 class but used an invalid character after [ (for example [[,abc,]]) caused
1751 pcre_compile() to give the error "Failed: internal error: code overflow" or
1752 in some cases to crash with a glibc free() error. This could even happen if
1753 the pattern terminated after [[ but there just happened to be a sequence of
1754 letters, a binary zero, and a closing ] in the memory that followed.
1755
1756 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1757 over the years. Originally (before any Unicode support), just the bottom 8
1758 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1759 output from "man perlunicode" includes this:
1760
1761 The regular expression compiler produces polymorphic opcodes. That
1762 is, the pattern adapts to the data and automatically switches to
1763 the Unicode character scheme when presented with Unicode data--or
1764 instead uses a traditional byte scheme when presented with byte
1765 data.
1766
1767 Sadly, a wide octal escape does not cause a switch, and in a string with
1768 no other multibyte characters, these octal escapes are treated as before.
1769 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1770 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1771 Unicode string.
1772
1773 I have not perpetrated such confusion in PCRE. Up till now, it took just
1774 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1775 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1776 translate to the appropriate multibyte character.
1777
1778 29. Applied some refactoring to reduce the number of warnings from Microsoft
1779 and Borland compilers. This has included removing the fudge introduced
1780 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1781 a warning about an unused variable.
1782
1783 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1784 characters since release 4.0, because Perl (from release 5.004) does not.
1785 [Or at least, is documented not to: some releases seem to be in conflict
1786 with the documentation.] However, when a pattern was studied with
1787 pcre_study() and all its branches started with \s, PCRE still included VT
1788 as a possible starting character. Of course, this did no harm; it just
1789 caused an unnecessary match attempt.
1790
1791 22. Removed a now-redundant internal flag bit that recorded the fact that case
1792 dependency changed within the pattern. This was once needed for "required
1793 byte" processing, but is no longer used. This recovers a now-scarce options
1794 bit. Also moved the least significant internal flag bit to the most-
1795 significant bit of the word, which was not previously used (hangover from
1796 the days when it was an int rather than a uint) to free up another bit for
1797 the future.
1798
1799 23. Added support for CRLF line endings as well as CR and LF. As well as the
1800 default being selectable at build time, it can now be changed at runtime
1801 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1802 specify that it is scanning data with non-default line endings.
1803
1804 24. Changed the definition of CXXLINK to make it agree with the definition of
1805 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1806
1807 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1808 recursions. This makes a big different to stack usage for some patterns.
1809
1810 26. If a subpattern containing a named recursion or subroutine reference such
1811 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1812 the space required for the compiled pattern went wrong and gave too small a
1813 value. Depending on the environment, this could lead to "Failed: internal
1814 error: code overflow at offset 49" or "glibc detected double free or
1815 corruption" errors.
1816
1817 27. Applied patches from Google (a) to support the new newline modes and (b) to
1818 advance over multibyte UTF-8 characters in GlobalReplace.
1819
1820 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1821 difference for some implementation of PCRE in some Windows version.
1822
1823 29. Added some extra testing facilities to pcretest:
1824
1825 \q<number> in a data line sets the "match limit" value
1826 \Q<number> in a data line sets the "match recursion limt" value
1827 -S <number> sets the stack size, where <number> is in megabytes
1828
1829 The -S option isn't available for Windows.
1830
1831
1832 Version 6.6 06-Feb-06
1833 ---------------------
1834
1835 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1836 in pcreposix.h. I have copied the definition from pcre.h.
1837
1838 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1839 because pcre.h is no longer a built file.
1840
1841 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1842 not normally included in the compiled code.
1843
1844
1845 Version 6.5 01-Feb-06
1846 ---------------------
1847
1848 1. When using the partial match feature with pcre_dfa_exec(), it was not
1849 anchoring the second and subsequent partial matches at the new starting
1850 point. This could lead to incorrect results. For example, with the pattern
1851 /1234/, partially matching against "123" and then "a4" gave a match.
1852
1853 2. Changes to pcregrep:
1854
1855 (a) All non-match returns from pcre_exec() were being treated as failures
1856 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1857 error message is output. Some extra information is given for the
1858 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1859 probably the only errors that are likely to be caused by users (by
1860 specifying a regex that has nested indefinite repeats, for instance).
1861 If there are more than 20 of these errors, pcregrep is abandoned.
1862
1863 (b) A binary zero was treated as data while matching, but terminated the
1864 output line if it was written out. This has been fixed: binary zeroes
1865 are now no different to any other data bytes.
1866
1867 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1868 used to set a locale for matching. The --locale=xxxx long option has
1869 been added (no short equivalent) to specify a locale explicitly on the
1870 pcregrep command, overriding the environment variables.
1871
1872 (d) When -B was used with -n, some line numbers in the output were one less
1873 than they should have been.
1874
1875 (e) Added the -o (--only-matching) option.
1876
1877 (f) If -A or -C was used with -c (count only), some lines of context were
1878 accidentally printed for the final match.
1879
1880 (g) Added the -H (--with-filename) option.
1881
1882 (h) The combination of options -rh failed to suppress file names for files
1883 that were found from directory arguments.
1884
1885 (i) Added the -D (--devices) and -d (--directories) options.
1886
1887 (j) Added the -F (--fixed-strings) option.
1888
1889 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1890
1891 (l) Added the --colo(u)r option.
1892
1893 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1894 is not present by default.
1895
1896 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1897 items such as (?R) or (?1), when the recursion could match a number of
1898 alternatives. If it matched one of the alternatives, but subsequently,
1899 outside the recursion, there was a failure, the code tried to back up into
1900 the recursion. However, because of the way PCRE is implemented, this is not
1901 possible, and the result was an incorrect result from the match.
1902
1903 In order to prevent this happening, the specification of recursion has
1904 been changed so that all such subpatterns are automatically treated as
1905 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1906
1907 4. I had overlooked the fact that, in some locales, there are characters for
1908 which isalpha() is true but neither isupper() nor islower() are true. In
1909 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1910 and ordfeminine) are like this. This affected the treatment of \w and \W
1911 when they appeared in character classes, but not when they appeared outside
1912 a character class. The bit map for "word" characters is now created
1913 separately from the results of isalnum() instead of just taking it from the
1914 upper, lower, and digit maps. (Plus the underscore character, of course.)
1915
1916 5. The above bug also affected the handling of POSIX character classes such as
1917 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1918 permanent tables. Instead, the bit maps for such a class were previously
1919 created as the appropriate unions of the upper, lower, and digit bitmaps.
1920 Now they are created by subtraction from the [[:word:]] class, which has
1921 its own bitmap.
1922
1923 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1924 It is created by subtracting the vertical space characters (\x09, \x0a,
1925 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1926 subtraction was done in the overall bitmap for a character class, meaning
1927 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1928 be recognized. This bug has been fixed.
1929
1930 7. Patches from the folks at Google:
1931
1932 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1933 real life, but is still worth protecting against".
1934
1935 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1936 regular expressions".
1937
1938 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1939 have it.
1940
1941 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1942 "configure" and the latter not, in order to fix a problem somebody had
1943 with compiling the Arg class on HP-UX.
1944
1945 (e) Improve the error-handling of the C++ wrapper a little bit.
1946
1947 (f) New tests for checking recursion limiting.
1948
1949 8. The pcre_memmove() function, which is used only if the environment does not
1950 have a standard memmove() function (and is therefore rarely compiled),
1951 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1952 returning a result (though PCRE never actually uses the result).
1953
1954 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1955 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1956 returned instead of calling malloc() with an overflowing number that would
1957 most likely cause subsequent chaos.
1958
1959 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1960
1961 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1962 with this option is matched, the nmatch and pmatch options of regexec() are
1963 ignored.
1964
1965 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1966 provided in case anyone wants to the the POSIX interface with UTF-8
1967 strings.
1968
1969 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1970 C++ linking (needed for some HP-UX environments).
1971
1972 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1973 (unused parameter) and in the pcre_printint() function (omitted "default"
1974 switch label when the default is to do nothing).
1975
1976 15. Added some code to make it possible, when PCRE is compiled as a C++
1977 library, to replace subject pointers for pcre_exec() with a smart pointer
1978 class, thus making it possible to process discontinuous strings.
1979
1980 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1981 much the same function. They were added by different people who were trying
1982 to make PCRE easy to compile on non-Unix systems. It has been suggested
1983 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1984 for compiling on Windows systems. I have therefore replaced it with
1985 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1986 defaults to "extern" for C or "extern C" for C++, which works fine on
1987 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1988 SCOPE with something explicit in config.h. In addition:
1989
1990 (a) pcreposix.h still had just "extern" instead of either of these macros;
1991 I have replaced it with PCRE_DATA_SCOPE.
1992
1993 (b) Functions such as _pcre_xclass(), which are internal to the library,
1994 but external in the C sense, all had PCRE_EXPORT in their definitions.
1995 This is apparently wrong for the Windows case, so I have removed it.
1996 (It makes no difference on Unix-like systems.)
1997
1998 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1999 of recursive calls to match(). This is different to MATCH_LIMIT because
2000 that limits the total number of calls to match(), not all of which increase
2001 the depth of recursion. Limiting the recursion depth limits the amount of
2002 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2003 when PCRE is compiled, and changed at run time. A patch from Google adds
2004 this functionality to the C++ interface.
2005
2006 18. Changes to the handling of Unicode character properties:
2007
2008 (a) Updated the table to Unicode 4.1.0.
2009
2010 (b) Recognize characters that are not in the table as "Cn" (undefined).
2011
2012 (c) I revised the way the table is implemented to a much improved format
2013 which includes recognition of ranges. It now supports the ranges that
2014 are defined in UnicodeData.txt, and it also amalgamates other
2015 characters into ranges. This has reduced the number of entries in the
2016 table from around 16,000 to around 3,000, thus reducing its size
2017 considerably. I realized I did not need to use a tree structure after
2018 all - a binary chop search is just as efficient. Having reduced the
2019 number of entries, I extended their size from 6 bytes to 8 bytes to
2020 allow for more data.
2021
2022 (d) Added support for Unicode script names via properties such as \p{Han}.
2023
2024 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2025 matching that character.
2026
2027 20. When matching a repeated Unicode property with a minimum greater than zero,
2028 (for example \pL{2,}), PCRE could look past the end of the subject if it
2029 reached it while seeking the minimum number of characters. This could
2030 happen only if some of the characters were more than one byte long, because
2031 there is a check for at least the minimum number of bytes.
2032
2033 21. Refactored the implementation of \p and \P so as to be more general, to
2034 allow for more different types of property in future. This has changed the
2035 compiled form incompatibly. Anybody with saved compiled patterns that use
2036 \p or \P will have to recompile them.
2037
2038 22. Added "Any" and "L&" to the supported property types.
2039
2040 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2041 but give a compile time error if the value is greater than 0xff.
2042
2043 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2044 accidentally not being installed or uninstalled.
2045
2046 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2047 made were to insert the current release number. This seemed silly, because
2048 it made things harder for people building PCRE on systems that don't run
2049 "configure". I have turned pcre.h into a distributed file, no longer built
2050 by "configure", with the version identification directly included. There is
2051 no longer a pcre.h.in file.
2052
2053 However, this change necessitated a change to the pcre-config script as
2054 well. It is built from pcre-config.in, and one of the substitutions was the
2055 release number. I have updated configure.ac so that ./configure now finds
2056 the release number by grepping pcre.h.
2057
2058 26. Added the ability to run the tests under valgrind.
2059
2060
2061 Version 6.4 05-Sep-05
2062 ---------------------
2063
2064 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2065 "--" to be printed when multiple files were scanned, even when none of the
2066 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2067 consider it to be a bug, and have restored the previous behaviour.
2068
2069 2. A couple of code tidies to get rid of compiler warnings.
2070
2071 3. The pcretest program used to cheat by referring to symbols in the library
2072 whose names begin with _pcre_. These are internal symbols that are not
2073 really supposed to be visible externally, and in some environments it is
2074 possible to suppress them. The cheating is now confined to including
2075 certain files from the library's source, which is a bit cleaner.
2076
2077 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2078 file's purpose clearer.
2079
2080 5. Reorganized pcre_ucp_findchar().
2081
2082
2083 Version 6.3 15-Aug-05
2084 ---------------------
2085
2086 1. The file libpcre.pc.in did not have general read permission in the tarball.
2087
2088 2. There were some problems when building without C++ support:
2089
2090 (a) If C++ support was not built, "make install" and "make test" still
2091 tried to test it.
2092
2093 (b) There were problems when the value of CXX was explicitly set. Some
2094 changes have been made to try to fix these, and ...
2095
2096 (c) --disable-cpp can now be used to explicitly disable C++ support.
2097
2098 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2099 backslash in a target when C++ was disabled. This confuses some
2100 versions of "make", apparently. Using an intermediate variable solves
2101 this. (Same for CPP_LOBJ.)
2102
2103 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2104 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2105 necessary on certain architectures.
2106
2107 4. Added a setting of -export-symbols-regex to the link command to remove
2108 those symbols that are exported in the C sense, but actually are local
2109 within the library, and not documented. Their names all begin with
2110 "_pcre_". This is not a perfect job, because (a) we have to except some
2111 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2112 available (and never for static libraries). I have made a note to try to
2113 find a way round (a) in the future.
2114
2115
2116 Version 6.2 01-Aug-05
2117 ---------------------
2118
2119 1. There was no test for integer overflow of quantifier values. A construction
2120 such as {1111111111111111} would give undefined results. What is worse, if
2121 a minimum quantifier for a parenthesized subpattern overflowed and became
2122 negative, the calculation of the memory size went wrong. This could have
2123 led to memory overwriting.
2124
2125 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2126
2127 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2128 operating environments where this matters.
2129
2130 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2131 PCRE options from within the C++ wrapper.
2132
2133 5. Named capturing subpatterns were not being correctly counted when a pattern
2134 was compiled. This caused two problems: (a) If there were more than 100
2135 such subpatterns, the calculation of the memory needed for the whole
2136 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2137 back references of the form \12, where the number was greater than 9, were
2138 not recognized as back references, even though there were sufficient
2139 previous subpatterns.
2140
2141 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2142 versions of gcc, e.g. 2.95.4.
2143
2144
2145 Version 6.1 21-Jun-05
2146 ---------------------
2147
2148 1. There was one reference to the variable "posix" in pcretest.c that was not
2149 surrounded by "#if !defined NOPOSIX".
2150
2151 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2152 the cross-check on the old pcre_info() function, for the benefit of the
2153 cut-down version of PCRE that is currently imported into Exim.
2154
2155 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2156 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2157 patterns that start (?i) but I don't think that matters. The use of (?i) is
2158 just an example; this all applies to the other options as well.
2159
2160 4. Since libtool seems to echo the compile commands it is issuing, the output
2161 from "make" can be reduced a bit by putting "@" in front of each libtool
2162 compile command.
2163
2164 5. Patch from the folks at Google for configure.in to be a bit more thorough
2165 in checking for a suitable C++ installation before trying to compile the
2166 C++ stuff. This should fix a reported problem when a compiler was present,
2167 but no suitable headers.
2168
2169 6. The man pages all had just "PCRE" as their title. I have changed them to
2170 be the relevant file name. I have also arranged that these names are
2171 retained in the file doc/pcre.txt, which is a concatenation in text format
2172 of all the man pages except the little individual ones for each function.
2173
2174 7. The NON-UNIX-USE file had not been updated for the different set of source
2175 files that come with release 6. I also added a few comments about the C++
2176 wrapper.
2177
2178
2179 Version 6.0 07-Jun-05
2180 ---------------------
2181
2182 1. Some minor internal re-organization to help with my DFA experiments.
2183
2184 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2185 didn't matter for the library itself when fully configured, but did matter
2186 when compiling without UCP support, or within Exim, where the ucp files are
2187 not imported.
2188
2189 3. Refactoring of the library code to split up the various functions into
2190 different source modules. The addition of the new DFA matching code (see
2191 below) to a single monolithic source would have made it really too
2192 unwieldy, quite apart from causing all the code to be include in a
2193 statically linked application, when only some functions are used. This is
2194 relevant even without the DFA addition now that patterns can be compiled in
2195 one application and matched in another.
2196
2197 The downside of splitting up is that there have to be some external
2198 functions and data tables that are used internally in different modules of
2199 the library but which are not part of the API. These have all had their
2200 names changed to start with "_pcre_" so that they are unlikely to clash
2201 with other external names.
2202
2203 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2204 a different (DFA) algorithm. Although it is slower than the original
2205 function, it does have some advantages for certain types of matching
2206 problem.
2207
2208 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2209 including restarting after a partial match.
2210
2211 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2212 defined when compiling for Windows was sent to me. I have put it into the
2213 code, though I have no means of testing or verifying it.
2214
2215 7. Added the pcre_refcount() auxiliary function.
2216
2217 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2218 match before or at the first newline in the subject string. In pcretest,
2219 the /f option on a pattern can be used to set this.
2220
2221 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2222 would behave wrongly. This has been present in PCRE since release 4.0.
2223
2224 10. A number of changes to the pcregrep command:
2225
2226 (a) Refactored how -x works; insert ^(...)$ instead of setting
2227 PCRE_ANCHORED and checking the length, in preparation for adding
2228 something similar for -w.
2229
2230 (b) Added the -w (match as a word) option.
2231
2232 (c) Refactored the way lines are read and buffered so as to have more
2233 than one at a time available.
2234
2235 (d) Implemented a pcregrep test script.
2236
2237 (e) Added the -M (multiline match) option. This allows patterns to match
2238 over several lines of the subject. The buffering ensures that at least
2239 8K, or the rest of the document (whichever is the shorter) is available
2240 for matching (and similarly the previous 8K for lookbehind assertions).
2241
2242 (f) Changed the --help output so that it now says
2243
2244 -w, --word-regex(p)
2245
2246 instead of two lines, one with "regex" and the other with "regexp"
2247 because that confused at least one person since the short forms are the
2248 same. (This required a bit of code, as the output is generated
2249 automatically from a table. It wasn't just a text change.)
2250
2251 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2252 option but starts with a hyphen. Could be a pattern or a path name
2253 starting with a hyphen, for instance.
2254
2255 (h) "-" can be given as a file name to represent stdin.
2256
2257 (i) When file names are being printed, "(standard input)" is used for
2258 the standard input, for compatibility with GNU grep. Previously
2259 "<stdin>" was used.
2260
2261 (j) The option --label=xxx can be used to supply a name to be used for
2262 stdin when file names are being printed. There is no short form.
2263
2264 (k) Re-factored the options decoding logic because we are going to add
2265 two more options that take data. Such options can now be given in four
2266 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2267
2268 (l) Added the -A, -B, and -C options for requesting that lines of context
2269 around matches be printed.
2270
2271 (m) Added the -L option to print the names of files that do not contain
2272 any matching lines, that is, the complement of -l.
2273
2274 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2275 continue to scan other files.
2276
2277 (o) The -s option was incorrectly implemented. For compatibility with other
2278 greps, it now suppresses the error message for a non-existent or non-
2279 accessible file (but not the return code). There is a new option called
2280 -q that suppresses the output of matching lines, which was what -s was
2281 previously doing.
2282
2283 (p) Added --include and --exclude options to specify files for inclusion
2284 and exclusion when recursing.
2285
2286 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2287 Hopefully, it now does.
2288
2289 12. Missing cast in pcre_study().
2290
2291 13. Added an "uninstall" target to the makefile.
2292
2293 14. Replaced "extern" in the function prototypes in Makefile.in with
2294 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2295 world, but is set differently for Windows.
2296
2297 15. Added a second compiling function called pcre_compile2(). The only
2298 difference is that it has an extra argument, which is a pointer to an
2299 integer error code. When there is a compile-time failure, this is set
2300 non-zero, in addition to the error test pointer being set to point to an
2301 error message. The new argument may be NULL if no error number is required
2302 (but then you may as well call pcre_compile(), which is now just a
2303 wrapper). This facility is provided because some applications need a
2304 numeric error indication, but it has also enabled me to tidy up the way
2305 compile-time errors are handled in the POSIX wrapper.
2306
2307 16. Added VPATH=.libs to the makefile; this should help when building with one
2308 prefix path and installing with another. (Or so I'm told by someone who
2309 knows more about this stuff than I do.)
2310
2311 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2312 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2313 match everything, including newlines. This is not POSIX-compatible, but
2314 somebody wanted the feature. From pcretest it can be activated by using
2315 both the P and the s flags.
2316
2317 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2318
2319 19. libpcre.pc was being incorrectly installed as executable.
2320
2321 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2322 it now also looks for '\r' so that it will work unmodified on Windows.
2323
2324 21. Added Google's contributed C++ wrapper to the distribution.
2325
2326 22. Added some untidy missing memory free() calls in pcretest, to keep
2327 Electric Fence happy when testing.
2328
2329
2330
2331 Version 5.0 13-Sep-04
2332 ---------------------
2333
2334 1. Internal change: literal characters are no longer packed up into items
2335 containing multiple characters in a single byte-string. Each character
2336 is now matched using a separate opcode. However, there may be more than one
2337 byte in the character in UTF-8 mode.
2338
2339 2. The pcre_callout_block structure has two new fields: pattern_position and
2340 next_item_length. These contain the offset in the pattern to the next match
2341 item, and its length, respectively.
2342
2343 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2344 insertion of callouts before each pattern item. Added the /C option to
2345 pcretest to make use of this.
2346
2347 4. On the advice of a Windows user, the lines
2348
2349 #if defined(_WIN32) || defined(WIN32)
2350 _setmode( _fileno( stdout ), 0x8000 );
2351 #endif /* defined(_WIN32) || defined(WIN32) */
2352
2353 have been added to the source of pcretest. This apparently does useful
2354 magic in relation to line terminators.
2355
2356 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2357 for the benefit of those environments where the "b" makes a difference.
2358
2359 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2360 to know about it. I have put a hack into configure.in that adds in code
2361 to set GCC=yes if CC=icc. This seems to end up at a point in the
2362 generated configure script that is early enough to affect the setting of
2363 compiler options, which is what is needed, but I have no means of testing
2364 whether it really works. (The user who reported this had patched the
2365 generated configure script, which of course I cannot do.)
2366
2367 LATER: After change 22 below (new libtool files), the configure script
2368 seems to know about icc (and also ecc). Therefore, I have commented out
2369 this hack in configure.in.
2370
2371 7. Added support for pkg-config (2 patches were sent in).
2372
2373 8. Negated POSIX character classes that used a combination of internal tables
2374 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2375 [[:^ascii]]. Typically, they would match almost any characters. The other
2376 POSIX classes were not broken in this way.
2377
2378 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2379 to find the match, as PCRE was deluded into thinking that the match had to
2380 start at the start point or following a newline. The same bug applied to
2381 patterns with negative forward assertions or any backward assertions
2382 preceding ".*" at the start, unless the pattern required a fixed first
2383 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2384
2385 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2386 starting at the last subject character, bytes beyond the end of the subject
2387 string were read.
2388
2389 11. Renamed the variable "class" as "classbits" to make life easier for C++
2390 users. (Previously there was a macro definition, but it apparently wasn't
2391 enough.)
2392
2393 12. Added the new field "tables" to the extra data so that tables can be passed
2394 in at exec time, or the internal tables can be re-selected. This allows
2395 a compiled regex to be saved and re-used at a later time by a different
2396 program that might have everything at different addresses.
2397
2398 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2399 -R library as well as a -L library.
2400
2401 14. The debugging options of pcretest (-d on the command line or D on a
2402 pattern) showed incorrect output for anything following an extended class
2403 that contained multibyte characters and which was followed by a quantifier.
2404
2405 15. Added optional support for general category Unicode character properties
2406 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2407 support. It adds about 90K to the size of the library. The meanings of the
2408 inbuilt class escapes such as \d and \s have NOT been changed.
2409
2410 16. Updated pcredemo.c to include calls to free() to release the memory for the
2411 compiled pattern.
2412
2413 17. The generated file chartables.c was being created in the source directory
2414 instead of in the building directory. This caused the build to fail if the
2415 source directory was different from the building directory, and was
2416 read-only.
2417
2418 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2419 file. No doubt somebody will tell me if they don't make sense... Also added
2420 Dan Mooney's comments about building on OpenVMS.
2421
2422 19. Added support for partial matching via the PCRE_PARTIAL option for
2423 pcre_exec() and the \P data escape in pcretest.
2424
2425 20. Extended pcretest with 3 new pattern features:
2426
2427 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2428 write the compiled pattern to the file whose name is "rest-of-line".
2429 This is a straight binary dump of the data, with the saved pointer to
2430 the character tables forced to be NULL. The study data, if any, is
2431 written too. After writing, pcretest reads a new pattern.
2432
2433 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2434 compiled pattern from the given file. There must not be any
2435 occurrences of "<" in the file name (pretty unlikely); if there are,
2436 pcretest will instead treat the initial "<" as a pattern delimiter.
2437 After reading in the pattern, pcretest goes on to read data lines as
2438 usual.
2439
2440 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2441 and 16-bit fields in a compiled pattern, to simulate a pattern that
2442 was compiled on a host of opposite endianness.
2443
2444 21. The pcre-exec() function can now cope with patterns that were compiled on
2445 hosts of opposite endianness, with this restriction:
2446
2447 As for any compiled expression that is saved and used later, the tables
2448 pointer field cannot be preserved; the extra_data field in the arguments
2449 to pcre_exec() should be used to pass in a tables address if a value
2450 other than the default internal tables were used at compile time.
2451
2452 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2453 now diagnosed as an error. Previously, most of the time, a negative number
2454 would have been treated as zero, but if in addition "ovector" was passed as
2455 NULL, a crash could occur.
2456
2457 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2458 new versions from the libtool 1.5 distribution (the last one is a copy of
2459 a file called libtool.m4). This seems to have fixed the need to patch
2460 "configure" to support Darwin 1.3 (which I used to do). However, I still
2461 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2462 workstation).
2463
2464 24. Changed the PCRE licence to be the more standard "BSD" licence.
2465
2466
2467 Version 4.5 01-Dec-03
2468 ---------------------
2469
2470 1. There has been some re-arrangement of the code for the match() function so
2471 that it can be compiled in a version that does not call itself recursively.
2472 Instead, it keeps those local variables that need separate instances for
2473 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2474 needs to "recurse". Keeping track of where control must go is done by means
2475 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2476 hide most of the details from the main code, and operates only if
2477 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2478 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2479 operating.
2480
2481 To make it easier for callers to provide specially tailored get/free
2482 functions for this usage, two new functions, pcre_stack_malloc, and
2483 pcre_stack_free, are used. They are always called in strict stacking order,
2484 and the size of block requested is always the same.
2485
2486 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2487 PCRE has been compiled to use the stack or the heap for recursion. The
2488 -C option of pcretest uses this to show which version is compiled.
2489
2490 A new data escape \S, is added to pcretest; it causes the amounts of store
2491 obtained and freed by both kinds of malloc/free at match time to be added
2492 to the output.
2493
2494 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2495 what's available on my current Linux desktop machine.
2496
2497 3. When matching a UTF-8 string, the test for a valid string at the start has
2498 been extended. If start_offset is not zero, PCRE now checks that it points
2499 to a byte that is the start of a UTF-8 character. If not, it returns
2500 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2501 this is necessary because there may be backward assertions in the pattern.
2502 When matching the same subject several times, it may save resources to use
2503 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2504
2505 4. The code for checking the validity of UTF-8 strings has been tightened so
2506 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2507 containing "overlong sequences".
2508
2509 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2510 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2511 should have been "&", but it just so happened that all the cases this let
2512 through by mistake were picked up later in the function.
2513
2514 6. I had used a variable called "isblank" - this is a C99 function, causing
2515 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2516
2517 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2518 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2519 so the version is shown; (c) stop "make test" echoing "./RunTest".
2520
2521 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2522
2523 9. The prototype for memmove() for systems that don't have it was using
2524 size_t, but the inclusion of the header that defines size_t was later. I've
2525 moved the #includes for the C headers earlier to avoid this.
2526
2527 10. Added some adjustments to the code to make it easier to compiler on certain
2528 special systems:
2529
2530 (a) Some "const" qualifiers were missing.
2531 (b) Added the macro EXPORT before all exported functions; by default this
2532 is defined to be empty.
2533 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2534 that it reads its output file name as an argument instead of writing
2535 to the standard output and assuming this can be redirected.
2536
2537 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2538 class containing characters with values greater than 255, PCRE compilation
2539 went into a loop.
2540
2541 12. A recursive reference to a subpattern that was within another subpattern
2542 that had a minimum quantifier of zero caused PCRE to crash. For example,
2543 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2544 recursion. If the recursively-called subpattern itself had a zero repeat,
2545 that was OK.
2546
2547 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2548 buffer into which it was copied (for escape processing) was still set at
2549 1024, so long lines caused crashes.
2550
2551 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2552 "internal error: code overflow...". This applied to any character class
2553 that was followed by a possessive quantifier.
2554
2555 15. Modified the Makefile to add libpcre.la as a prerequisite for
2556 libpcreposix.la because I was told this is needed for a parallel build to
2557 work.
2558
2559 16. If a pattern that contained .* following optional items at the start was
2560 studied, the wrong optimizing data was generated, leading to matching
2561 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2562 matching string must start with a or b or c. The correct conclusion for
2563 this pattern is that a match can start with any character.
2564
2565
2566 Version 4.4 13-Aug-03
2567 ---------------------
2568
2569 1. In UTF-8 mode, a character class containing characters with values between
2570 127 and 255 was not handled correctly if the compiled pattern was studied.
2571 In fixing this, I have also improved the studying algorithm for such
2572 classes (slightly).
2573
2574 2. Three internal functions had redundant arguments passed to them. Removal
2575 might give a very teeny performance improvement.
2576
2577 3. Documentation bug: the value of the capture_top field in a callout is *one
2578 more than* the number of the hightest numbered captured substring.
2579
2580 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2581 in incorrectly linking with a previously installed version. They now link
2582 explicitly with libpcre.la.
2583
2584 5. configure.in no longer needs to recognize Cygwin specially.
2585
2586 6. A problem in pcre.in for Windows platforms is fixed.
2587
2588 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2589 pcretest, it used to include the size of the study block as part of its
2590 output. Unfortunately, the structure contains a field that has a different
2591 size on different hardware architectures. This meant that the tests that
2592 showed this size failed. As the block is currently always of a fixed size,
2593 this information isn't actually particularly useful in pcretest output, so
2594 I have just removed it.
2595
2596 8. Three pre-processor statements accidentally did not start in column 1.
2597 Sadly, there are *still* compilers around that complain, even though
2598 standard C has not required this for well over a decade. Sigh.
2599
2600 9. In pcretest, the code for checking callouts passed small integers in the
2601 callout_data field, which is a void * field. However, some picky compilers
2602 complained about the casts involved for this on 64-bit systems. Now
2603 pcretest passes the address of the small integer instead, which should get
2604 rid of the warnings.
2605
2606 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2607 both compile and run time, and gives an error if an invalid UTF-8 sequence
2608 is found. There is a option for disabling this check in cases where the
2609 string is known to be correct and/or the maximum performance is wanted.
2610
2611 11. In response to a bug report, I changed one line in Makefile.in from
2612
2613 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2614 to
2615 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2616
2617 to look similar to other lines, but I have no way of telling whether this
2618 is the right thing to do, as I do not use Windows. No doubt I'll get told
2619 if it's wrong...
2620
2621
2622 Version 4.3 21-May-03
2623 ---------------------
2624
2625 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2626 Makefile.
2627
2628 2. Some refactoring to improve the quality of the code:
2629
2630 (i) The utf8_table... variables are now declared "const".
2631
2632 (ii) The code for \cx, which used the "case flipping" table to upper case
2633 lower case letters, now just substracts 32. This is ASCII-specific,
2634 but the whole concept of \cx is ASCII-specific, so it seems
2635 reasonable.
2636
2637 (iii) PCRE was using its character types table to recognize decimal and
2638 hexadecimal digits in the pattern. This is silly, because it handles
2639 only 0-9, a-f, and A-F, but the character types table is locale-
2640 specific, which means strange things might happen. A private
2641 table is now used for this - though it costs 256 bytes, a table is
2642 much faster than multiple explicit tests. Of course, the standard
2643 character types table is still used for matching digits in subject
2644 strings against \d.
2645
2646 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2647 ending in _t are). So I've renamed it as ESC_tee.
2648
2649 3. The first argument for regexec() in the POSIX wrapper should have been
2650 defined as "const".
2651
2652 4. Changed pcretest to use malloc() for its buffers so that they can be
2653 Electric Fenced for debugging.
2654
2655 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2656 to read one or more bytes before the start of the subject string. Often this
2657 had no effect on PCRE's behaviour, but in some circumstances it could
2658 provoke a segmentation fault.
2659
2660 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2661 to try to read one or more bytes before the start of the subject string.
2662
2663 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2664 UTF-8 support could misbehave in various ways if the subject string
2665 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2666 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2667 back over UTF-8 characters.)
2668
2669
2670 Version 4.2 14-Apr-03
2671 ---------------------
2672
2673 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2674
2675 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2676 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2677 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2678 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2679 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2680 and BUILD_EXEEXT
2681 Note: automatic setting of the BUILD variables is not yet working
2682 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2683 compile-time but not at link-time
2684 [LINK]: use for linking executables only
2685 make different versions for Windows and non-Windows
2686 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2687 libraries
2688 [LINK_FOR_BUILD]: new variable
2689 [OBJEXT]: use throughout
2690 [EXEEXT]: use throughout
2691 <winshared>: new target
2692 <wininstall>: new target
2693 <dftables.o>: use native compiler
2694 <dftables>: use native linker
2695 <install>: handle Windows platform correctly
2696 <clean>: ditto
2697 <check>: ditto
2698 copy DLL to top builddir before testing
2699
2700 As part of these changes, -no-undefined was removed again. This was reported
2701 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2702 in any case.
2703
2704 3. Some tidies to get rid of compiler warnings:
2705
2706 . In the match_data structure, match_limit was an unsigned long int, whereas
2707 match_call_count was an int. I've made them both unsigned long ints.
2708
2709 . In pcretest the fact that a const uschar * doesn't automatically cast to
2710 a void * provoked a warning.
2711
2712 . Turning on some more compiler warnings threw up some "shadow" variables
2713 and a few more missing casts.
2714
2715 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2716 option, a class that contained a single character with a value between 128
2717 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2718
2719 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2720 option, a class that contained several characters, but with at least one
2721 whose value was between 128 and 255 caused PCRE to crash.
2722
2723
2724 Version 4.1 12-Mar-03
2725 ---------------------
2726
2727 1. Compiling with gcc -pedantic found a couple of places where casts were
2728 needed, and a string in dftables.c that was longer than standard compilers are
2729 required to support.
2730
2731 2. Compiling with Sun's compiler found a few more places where the code could
2732 be tidied up in order to avoid warnings.
2733
2734 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2735 first of these names is deprecated in the latest Autoconf in favour of the name
2736 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2737 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2738 analogy I have changed it to CFLAGS_FOR_BUILD.
2739
2740 4. Added -no-undefined to the linking command in the Makefile, because this is
2741 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2742 linking step for the pcreposix library.
2743
2744 5. PCRE was failing to diagnose the case of two named groups with the same
2745 name.
2746
2747 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2748 literal character that is needed in the subject for a match, and scans along to
2749 ensure that it is present before embarking on the full matching process. This
2750 saves time in cases of nested unlimited repeats that are never going to match.
2751 Problem: the scan can take a lot of time if the subject is very long (e.g.
2752 megabytes), thus penalizing straightforward matches. It is now done only if the
2753 amount of subject to be scanned is less than 1000 bytes.
2754
2755 7. A lesser problem with the same optimization is that it was recording the
2756 first character of an anchored pattern as "needed", thus provoking a search
2757 right along the subject, even when the first match of the pattern was going to
2758 fail. The "needed" character is now not set for anchored patterns, unless it
2759 follows something in the pattern that is of non-fixed length. Thus, it still
2760 fulfils its original purpose of finding quick non-matches in cases of nested
2761 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2762
2763
2764 Version 4.0 17-Feb-03
2765 ---------------------
2766
2767 1. If a comment in an extended regex that started immediately after a meta-item
2768 extended to the end of string, PCRE compiled incorrect data. This could lead to
2769 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2770
2771 2. Moved to autoconf 2.53 and libtool 1.4.2.
2772
2773 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2774 the special perltest8 script is no longer needed - all the tests can be run
2775 from a single perltest script.
2776
2777 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2778 by \s. It has now been removed in PCRE. This means it isn't recognized as
2779 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2780 class [:space:] *does* include VT, thereby creating a mess.
2781
2782 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2783 space and tab.
2784
2785 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2786 its new features into the main test script, reducing the number of scripts.
2787
2788 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2789 were backward compatible, and made the (?i) apply to the whole pattern, as if
2790 /i were given. Now it behaves more logically, and applies the option setting
2791 only to what follows. PCRE has been changed to follow suit. However, if it
2792 finds options settings right at the start of the pattern, it extracts them into
2793 the global options, as before. Thus, they show up in the info data.
2794
2795 8. Added support for the \Q...\E escape sequence. Characters in between are
2796 treated as literals. This is slightly different from Perl in that $ and @ are
2797 also handled as literals inside the quotes. In Perl, they will cause variable
2798 interpolation. Note the following examples:
2799
2800 Pattern PCRE matches Perl matches
2801
2802 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2803 \Qabc\$xyz\E abc\$xyz abc\$xyz
2804 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2805
2806 For compatibility with Perl, \Q...\E sequences are recognized inside character
2807 classes as well as outside them.
2808
2809 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2810 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2811 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2812 signed/unsigned warnings.
2813
2814 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2815 option for pcretest, so I've replaced it by a simple function that does just
2816 that job.
2817
2818 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2819 "pcregrep -".
2820
2821 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2822 Java package. This provides some syntactic sugar for simple cases of what my
2823 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2824 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2825 item, you can use this simplified notation. Note that only makes sense with
2826 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2827 greediness, whatever the setting of the PCRE_UNGREEDY option.
2828
2829 13. A change of greediness default within a pattern was not taking effect at
2830 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2831 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2832 was abstracted outside.
2833
2834 14. PCRE now supports the \G assertion. It is true when the current matching
2835 position is at the start point of the match. This differs from \A when the
2836 starting offset is non-zero. Used with the /g option of pcretest (or similar
2837 code), it works in the same way as it does for Perl's /g option. If all
2838 alternatives of a regex begin with \G, the expression is anchored to the start
2839 match position, and the "anchored" flag is set in the compiled expression.
2840
2841 15. Some bugs concerning the handling of certain option changes within patterns
2842 have been fixed. These applied to options other than (?ims). For example,
2843 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2844 been the other way round. Some of this was related to change 7 above.
2845
2846 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2847 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2848 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2849 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2850
2851 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2852 mode. Unlike ".", it always matches newline, whatever the setting of
2853 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2854 assertions. Perl allows it, but it doesn't (in general) work because it can't
2855 calculate the length of the lookbehind. At least, that's the case for Perl
2856 5.8.0 - I've been told they are going to document that it doesn't work in
2857 future.
2858
2859 18. Added an error diagnosis for escapes that PCRE does not support: these are
2860 \L, \l, \N, \P, \p, \U, \u, and \X.
2861
2862 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2863 reading past the end of the pattern in cases such as /[abcd/.
2864
2865 20. PCRE was getting more memory than necessary for patterns with classes that
2866 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2867
2868 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2869 compiling PCRE for use with Virtual Pascal.
2870
2871 22. Small fix to the Makefile to make it work properly if the build is done
2872 outside the source tree.
2873
2874 23. Added a new extension: a condition to go with recursion. If a conditional
2875 subpattern starts with (?(R) the "true" branch is used if recursion has
2876 happened, whereas the "false" branch is used only at the top level.
2877
2878 24. When there was a very long string of literal characters (over 255 bytes
2879 without UTF support, over 250 bytes with UTF support), the computation of how
2880 much memory was required could be incorrect, leading to segfaults or other
2881 strange effects.
2882
2883 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2884 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2885 there was a subsequent back reference to those brackets. This meant that, for
2886 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2887 possible to check for precisely this case. All we can do is abandon the
2888 optimization if .* occurs inside capturing brackets when there are any back
2889 references whatsoever. (See below for a better fix that came later.)
2890
2891 26. The handling of the optimization for finding the first character of a
2892 non-anchored pattern, and for finding a character that is required later in the
2893 match were failing in some cases. This didn't break the matching; it just
2894 failed to optimize when it could. The way this is done has been re-implemented.
2895
2896 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2897
2898 28. Added a new feature that provides some of the functionality that Perl
2899 provides with (?{...}). The facility is termed a "callout". The way it is done
2900 in PCRE is for the caller to provide an optional function, by setting
2901 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2902 global variable. By default it is unset, which disables all calling out. To get
2903 the function called, the regex must include (?C) at appropriate points. This
2904 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2905 This provides a means of identifying different callout points. When PCRE
2906 reaches such a point in the regex, if pcre_callout has been set, the external
2907 function is called. It is provided with data in a structure called
2908 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2909 matching continues; if it returns a non-zero value, the match at the current
2910 point fails. However, backtracking will occur if possible. [This was changed
2911 later and other features added - see item 49 below.]
2912
2913 29. pcretest is upgraded to test the callout functionality. It provides a
2914 callout function that displays information. By default, it shows the start of
2915 the match and the current position in the text. There are some new data escapes
2916 to vary what happens:
2917
2918 \C+ in addition, show current contents of captured substrings
2919 \C- do not supply a callout function
2920 \C!n return 1 when callout number n is reached
2921 \C!n!m return 1 when callout number n is reached for the mth time
2922
2923 30. If pcregrep was called with the -l option and just a single file name, it
2924 output "<stdin>" if a match was found, instead of the file name.
2925
2926 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2927 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2928 pcre_exec(). This saves a malloc/free per call. The default value of
2929 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2930 when configuring.
2931
2932 32. The default maximum size of a compiled pattern is 64K. There have been a
2933 few cases of people hitting this limit. The code now uses macros to handle the
2934 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2935 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2936 configuring. Tests 2 and 5 work only with 2-byte links because they output
2937 debugging information about compiled patterns.
2938
2939 33. Internal code re-arrangements:
2940
2941 (a) Moved the debugging function for printing out a compiled regex into
2942 its own source file (printint.c) and used #include to pull it into
2943 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2944 separate copies.
2945
2946 (b) Defined the list of op-code names for debugging as a macro in
2947 internal.h so that it is next to the definition of the opcodes.
2948
2949 (c) Defined a table of op-code lengths for simpler skipping along compiled
2950 code. This is again a macro in internal.h so that it is next to the
2951 definition of the opcodes.
2952
2953 34. Added support for recursive calls to individual subpatterns, along the
2954 lines of Robin Houston's patch (but implemented somewhat differently).
2955
2956 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2957 allow it to read and process whole directories in Win32. This code was
2958 contributed by Lionel Fourquaux; it has not been tested by me.
2959
2960 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2961 used to name a group. Names consist of alphanumerics and underscores, and must
2962 be unique. Back references use the syntax (?P=name) and recursive calls use
2963 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2964 numbers. The function pcre_fullinfo() can be used after compilation to extract
2965 a name/number map. There are three relevant calls:
2966
2967 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2968 PCRE_INFO_NAMECOUNT yields the number of entries
2969 PCRE_INFO_NAMETABLE yields a pointer to the map.
2970
2971 The map is a vector of fixed-size entries. The size of each entry depends on
2972 the length of the longest name used. The first two bytes of each entry are the
2973 group number, most significant byte first. There follows the corresponding
2974 name, zero terminated. The names are in alphabetical order.
2975
2976 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2977 case instead of 255. Making it the same both with and without UTF-8 support
2978 means that the same test output works with both.
2979
2980 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2981 calling malloc() with a zero argument.
2982
2983 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2984 optimization. I've improved things by keeping a bitmap of backreferences with
2985 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2986 fact referenced, the optimization can be applied. It is unlikely that a
2987 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2988 the match to follow \n) will appear inside brackets with a number greater than
2989 31, but if it does, any back reference > 31 suppresses the optimization.
2990
2991 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2992 of disabling numbered capturing parentheses. Any opening parenthesis that is
2993 not followed by ? behaves as if it were followed by ?: but named parentheses
2994 can still be used for capturing (and they will acquire numbers in the usual
2995 way).
2996
2997 41. Redesigned the return codes from the match() function into yes/no/error so
2998 that errors can be passed back from deep inside the nested calls. A malloc
2999 failure while inside a recursive subpattern call now causes the
3000 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3001
3002 42. It is now possible to set a limit on the number of times the match()
3003 function is called in a call to pcre_exec(). This facility makes it possible to
3004 limit the amount of recursion and backtracking, though not in a directly
3005 obvious way, because the match() function is used in a number of different
3006 circumstances. The count starts from zero for each position in the subject
3007 string (for non-anchored patterns). The default limit is, for compatibility, a
3008 large number, namely 10 000 000. You can change this in two ways:
3009
3010 (a) When configuring PCRE before making, you can use --with-match-limit=n
3011 to set a default value for the compiled library.
3012
3013 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3014 a different value is set. See 45 below.
3015
3016 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3017
3018 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3019 of things that can be changed at compile time. The first argument specifies
3020 what is wanted and the second points to where the information is to be placed.
3021 The current list of available information is:
3022
3023 PCRE_CONFIG_UTF8
3024
3025 The output is an integer that is set to one if UTF-8 support is available;
3026 otherwise it is set to zero.
3027
3028 PCRE_CONFIG_NEWLINE
3029
3030 The output is an integer that it set to the value of the code that is used for
3031 newline. It is either LF (10) or CR (13).
3032
3033 PCRE_CONFIG_LINK_SIZE
3034
3035 The output is an integer that contains the number of bytes used for internal
3036 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3037
3038 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3039
3040 The output is an integer that contains the threshold above which the POSIX
3041 interface uses malloc() for output vectors. See item 31 above.
3042
3043 PCRE_CONFIG_MATCH_LIMIT
3044
3045 The output is an unsigned integer that contains the default limit of the number
3046 of match() calls in a pcre_exec() execution. See 42 above.
3047
3048 44. pcretest has been upgraded by the addition of the -C option. This causes it
3049 to extract all the available output from the new pcre_config() function, and to
3050 output it. The program then exits immediately.
3051
3052 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3053 order to support additional features. One way would have been to define
3054 pcre_exec2() (for example) with extra arguments, but this would not have been
3055 extensible, and would also have required all calls to the original function to
3056 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3057 is used for passing in "extra" data from pcre_study().
3058
3059 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3060 contains the following fields:
3061
3062 flags a bitmap indicating which of the following fields are set
3063 study_data opaque data from pcre_study()
3064 match_limit a way of specifying a limit on match() calls for a specific
3065 call to pcre_exec()
3066 callout_data data for callouts (see 49 below)
3067
3068 The flag bits are also defined in pcre.h, and are
3069
3070 PCRE_EXTRA_STUDY_DATA
3071 PCRE_EXTRA_MATCH_LIMIT
3072 PCRE_EXTRA_CALLOUT_DATA
3073
3074 The pcre_study() function now returns one of these new pcre_extra blocks, with
3075 the actual study data pointed to by the study_data field, and the
3076 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3077 before. That is, this change is entirely upwards-compatible and requires no
3078 change to existing code.
3079
3080 If you want to pass in additional data to pcre_exec(), you can either place it
3081 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3082 block.
3083
3084 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3085 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3086 times with different match limits, until it finds the minimum value needed for
3087 pcre_exec() to complete. The value is then output. This can be instructive; for
3088 most simple matches the number is quite small, but for pathological cases it
3089 gets very large very quickly.
3090
3091 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3092 returns the size of the data block pointed to by the study_data field in a
3093 pcre_extra block, that is, the value that was passed as the argument to
3094 pcre_malloc() when PCRE was getting memory in which to place the information
3095 created by pcre_study(). The fourth argument should point to a size_t variable.
3096 pcretest has been extended so that this information is shown after a successful
3097 pcre_study() call when information about the compiled regex is being displayed.
3098
3099 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3100 because what follows is always an absolute path. (Later: it turns out that this
3101 is more than cosmetic for MinGW, because it doesn't like empty path
3102 components.)
3103
3104 49. Some changes have been made to the callout feature (see 28 above):
3105
3106 (i) A callout function now has three choices for what it returns:
3107
3108 0 => success, carry on matching
3109 > 0 => failure at this point, but backtrack if possible
3110 < 0 => serious error, return this value from pcre_exec()
3111
3112 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3113 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3114 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3115 use by callout functions. It will never be used by PCRE itself.
3116
3117 (ii) The pcre_extra structure (see 45 above) has a void * field called
3118 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3119 pcre_callout_block structure has a field of the same name. The contents of
3120 the field passed in the pcre_extra structure are passed to the callout
3121 function in the corresponding field in the callout block. This makes it
3122 easier to use the same callout-containing regex from multiple threads. For
3123 testing, the pcretest program has a new data escape
3124
3125 \C*n pass the number n (may be negative) as callout_data
3126
3127 If the callout function in pcretest receives a non-zero value as
3128 callout_data, it returns that value.
3129
3130 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3131 there were some redundant $(CFLAGS) in commands that are now specified as
3132 $(LINK), which already includes $(CFLAGS).
3133
3134 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3135 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3136 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3137 one-byte characters throughout. Note that case-insensitive matching applies
3138 only to characters whose values are less than 256. PCRE doesn't support the
3139 notion of cases for higher-valued characters.
3140
3141 (i) A character class whose characters are all within 0-255 is handled as
3142 a bit map, and the map is inverted for negative classes. Previously, a
3143 character > 255 always failed to match such a class; however it should
3144 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3145
3146 (ii) A negated character class with a single character < 255 is coded as
3147 "not this character" (OP_NOT). This wasn't working properly when the test
3148 character was multibyte, either singly or repeated.
3149
3150 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3151 mode, for example: \x{100}{2,3}.
3152
3153 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3154 singly or repeated) now correctly test multibyte characters. However,
3155 PCRE doesn't recognize any characters with values greater than 255 as
3156 digits, spaces, or word characters. Such characters always match \D, \S,
3157 and \W, and never match \d, \s, or \w.
3158
3159 (v) Classes may now contain characters and character ranges with values
3160 greater than 255. For example: [ab\x{100}-\x{400}].
3161
3162 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3163 PCRE in UTF-8 mode.
3164
3165 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3166 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3167 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3168 value.)
3169
3170 53. The single man page has become too large. I have therefore split it up into
3171 a number of separate man pages. These also give rise to individual HTML pages;
3172 these are now put in a separate directory, and there is an index.html page that
3173 lists them all. Some hyperlinking between the pages has been installed.
3174
3175 54. Added convenience functions for handling named capturing parentheses.
3176
3177 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3178 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3179 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3180 are faulted.
3181
3182 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3183 calling configure. These values are used when compiling the dftables.c program
3184 which is run to generate the source of the default character tables. They
3185 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3186 you will need to set these values.
3187
3188 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3189
3190
3191 Version 3.9 02-Jan-02
3192 ---------------------
3193
3194 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3195
3196 2. If --disable-static was given, the building process failed when trying to
3197 build pcretest and pcregrep. (For some reason it was using libtool to compile
3198 them, which is not right, as they aren't part of the library.)
3199
3200
3201 Version 3.8 18-Dec-01
3202 ---------------------
3203
3204 1. The experimental UTF-8 code was completely screwed up. It was packing the
3205 bytes in the wrong order. How dumb can you get?
3206
3207
3208 Version 3.7 29-Oct-01
3209 ---------------------
3210
3211 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3212 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3213 this didn't happen under Solaris 8, where I normally test things.
3214
3215 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3216 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3217 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3218 this fix an hour or so after the initial 3.7 release.)
3219
3220
3221 Version 3.6 23-Oct-01
3222 ---------------------
3223
3224 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3225 offsets passed as NULL with zero offset count.
3226
3227 2. The config.guess and config.sub files had not been updated when I moved to
3228 the latest autoconf.
3229
3230
3231 Version 3.5 15-Aug-01
3232 ---------------------
3233
3234 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3235 had been forgotten.
3236
3237 2. By using declared but undefined structures, we can avoid using "void"
3238 definitions in pcre.h while keeping the internal definitions of the structures
3239 private.
3240
3241 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3242 user point of view, this means that both static and shared libraries are built
3243 by default, but this can be individually controlled. More of the work of
3244 handling this static/shared cases is now inside libtool instead of PCRE's make
3245 file.
3246
3247 4. The pcretest utility is now installed along with pcregrep because it is
3248 useful for users (to test regexs) and by doing this, it automatically gets
3249 relinked by libtool. The documentation has been turned into a man page, so
3250 there are now .1, .txt, and .html versions in /doc.
3251
3252 5. Upgrades to pcregrep:
3253 (i) Added long-form option names like gnu grep.
3254 (ii) Added --help to list all options with an explanatory phrase.
3255 (iii) Added -r, --recursive to recurse into sub-directories.
3256 (iv) Added -f, --file to read patterns from a file.
3257
3258 6. pcre_exec() was referring to its "code" argument before testing that
3259 argument for NULL (and giving an error if it was NULL).
3260
3261 7. Upgraded Makefile.in to allow for compiling in a different directory from
3262 the source directory.
3263
3264 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3265 options bits, the pointer it was passed was to an int instead of to an unsigned
3266 long int. This mattered only on 64-bit systems.
3267
3268 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3269 generated) instead of pcre.in, which it its source. Also made the same change
3270 in several of the .c files.
3271
3272 10. A new release of gcc defines printf() as a macro, which broke pcretest
3273 because it had an ifdef in the middle of a string argument for printf(). Fixed
3274 by using separate calls to printf().
3275
3276 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3277 script, to force use of CR or LF instead of \n in the source. On non-Unix
3278 systems, the value can be set in config.h.
3279
3280 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3281 absolute limit. Changed the text of the error message to make this clear, and
3282 likewise updated the man page.
3283
3284 13. The limit of 99 on the number of capturing subpatterns has been removed.
3285 The new limit is 65535, which I hope will not be a "real" limit.
3286
3287
3288 Version 3.4 22-Aug-00
3289 ---------------------
3290
3291 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3292
3293 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3294
3295
3296 Version 3.3 01-Aug-00
3297 ---------------------
3298
3299 1. If an octal character was given, but the value was greater than \377, it
3300 was not getting masked to the least significant bits, as documented. This could
3301 lead to crashes in some systems.
3302
3303 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3304 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3305
3306 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3307 These just pass their arguments on to (pcre_free)(), but they are provided
3308 because some uses of PCRE bind it to non-C systems that can call its functions,
3309 but cannot call free() or pcre_free() directly.
3310
3311 4. Add "make test" as a synonym for "make check". Corrected some comments in
3312 the Makefile.
3313
3314 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3315 Makefile.
3316
3317 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3318 command called pgrep for grepping around the active processes.
3319
3320 7. Added the beginnings of support for UTF-8 character strings.
3321
3322 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3323 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3324 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3325 out for the ar command.)
3326
3327
3328 Version 3.2 12-May-00
3329 ---------------------
3330
3331 This is purely a bug fixing release.
3332
3333 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3334 of ZA. This was just one example of several cases that could provoke this bug,
3335 which was introduced by change 9 of version 2.00. The code for breaking
3336 infinite loops after an iteration that matches an empty string was't working
3337 correctly.
3338
3339 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3340 when matched against abbab (for example). After matching an empty string, it
3341 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3342 caused it to match further down the string than it should.
3343
3344 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3345 was there because it doesn't seem to be needed, and it causes trouble on some
3346 systems, as it is not a Standard C header. It has been removed.
3347
3348 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3349 were reported on the Macintosh. The changes were from
3350
3351 while ((c = *(++ptr)) != 0 && c != '\n');
3352 to
3353 while ((c = *(++ptr)) != 0 && c != '\n') ;
3354
3355 Totally extraordinary, but if that's what it takes...
3356
3357 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3358 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3359 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3360 assumes the way PCRE uses memmove() (always moving upwards).
3361
3362 6. PCRE is being used in one environment where strchr() is not available. There
3363 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3364 faster code anyway.
3365
3366
3367 Version 3.1 09-Feb-00
3368 ---------------------
3369
3370 The only change in this release is the fixing of some bugs in Makefile.in for
3371 the "install" target:
3372
3373 (1) It was failing to install pcreposix.h.
3374
3375 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3376
3377
3378 Version 3.0 01-Feb-00
3379 ---------------------
3380
3381 1. Add support for the /+ modifier to perltest (to output $` like it does in
3382 pcretest).
3383
3384 2. Add support for the /g modifier to perltest.
3385
3386 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3387 matches null strings.
3388
3389 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3390 pattern. Perl treats empty patterns specially - it reuses the most recent
3391 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3392 effect.
3393
3394 5. The POSIX interface was broken in that it was just handing over the POSIX
3395 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3396 required a bigger vector, with some working space on the end. This means that
3397 the POSIX wrapper now has to get and free some memory, and copy the results.
3398
3399 6. Added some simple autoconf support, placing the test data and the
3400 documentation in separate directories, re-organizing some of the
3401 information files, and making it build pcre-config (a GNU standard). Also added
3402 libtool support for building PCRE as a shared library, which is now the
3403 default.
3404
3405 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3406 09 are not valid octal constants. Single digits will be used for minor values
3407 less than 10.
3408
3409 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3410 existing programs that set these in the POSIX interface can use PCRE without
3411 modification.
3412
3413 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3414 return all that pcre_info() returns, plus additional data. The pcre_info()
3415 function is retained for compatibility, but is considered to be obsolete.
3416
3417 10. Added experimental recursion feature (?R) to handle one common case that
3418 Perl 5.6 will be able to do with (?p{...}).
3419
3420 11. Added support for POSIX character classes like [:alpha:], which Perl is
3421 adopting.
3422
3423
3424 Version 2.08 31-Aug-99
3425 ----------------------
3426
3427 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3428 trying to match at the startoffset position, but instead was moving forward to
3429 the next newline as if a previous match had failed.
3430
3431 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3432 and could get into a loop if a null string was matched other than at the start
3433 of the subject.
3434
3435 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3436 be distinguished at compile time, and for completeness also added PCRE_DATE.
3437
3438 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3439 in GnuWin32 environments.
3440
3441
3442 Version 2.07 29-Jul-99
3443 ----------------------
3444
3445 1. The documentation is now supplied in plain text form and HTML as well as in
3446 the form of man page sources.
3447
3448 2. C++ compilers don't like assigning (void *) values to other pointer types.
3449 In particular this affects malloc(). Although there is no problem in Standard
3450 C, I've put in casts to keep C++ compilers happy.
3451
3452 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3453 should be (const char *).
3454
3455 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3456 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3457 However, I haven't made this a standard facility. The documentation doesn't
3458 mention it, and the Makefile doesn't support it.
3459
3460 5. The Makefile now contains an "install" target, with editable destinations at
3461 the top of the file. The pcretest program is not installed.
3462
3463 6. pgrep -V now gives the PCRE version number and date.
3464
3465 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3466 causing the entire string to be ignored, instead of just the last character.
3467
3468 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3469 non-matching string, it can take a very, very long time, even for strings of
3470 quite modest length, because of the nested recursion. PCRE now does better in
3471 some of these cases. It does this by remembering the last required literal
3472 character in the pattern, and pre-searching the subject to ensure it is present
3473 before running the real match. In other words, it applies a heuristic to detect
3474 some types of certain failure quickly, and in the above example, if presented
3475 with a string that has no trailing " it gives "no match" very quickly.
3476
3477 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3478 other alternatives are tried instead.
3479
3480
3481 Version 2.06 09-Jun-99
3482 ----------------------
3483
3484 1. Change pcretest's output for amount of store used to show just the code
3485 space, because the remainder (the data block) varies in size between 32-bit and
3486 64-bit systems.
3487
3488 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3489 start matching at. This allows lookbehinds to work when searching for multiple
3490 occurrences in a string.
3491
3492 3. Added additional options to pcretest for testing multiple occurrences:
3493
3494 /+ outputs the rest of the string that follows a match
3495 /g loops for multiple occurrences, using the new startoffset argument
3496 /G loops for multiple occurrences by passing an incremented pointer
3497
3498 4. PCRE wasn't doing the "first character" optimization for patterns starting
3499 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3500 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3501 the letter 'x'. On long subject strings, this gives a significant speed-up.
3502
3503
3504 Version 2.05 21-Apr-99
3505 ----------------------
3506
3507 1. Changed the type of magic_number from int to long int so that it works
3508 properly on 16-bit systems.
3509
3510 2. Fixed a bug which caused patterns starting with .* not to work correctly
3511 when the subject string contained newline characters. PCRE was assuming
3512 anchoring for such patterns in all cases, which is not correct because .* will
3513 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3514 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3515 must be retried after every newline in the subject.
3516
3517
3518 Version 2.04 18-Feb-99
3519 ----------------------
3520
3521 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3522 computation of the store needed to hold the pattern was incorrect (too large).
3523 If such patterns were nested a few deep, this could multiply and become a real
3524 problem.
3525
3526 2. Added /M option to pcretest to show the memory requirement of a specific
3527 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3528
3529 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3530 compiled in such a way that the backtracking after subsequent failure was
3531 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3532 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3533
3534
3535 Version 2.03 02-Feb-99
3536 ----------------------
3537
3538 1. Fixed typo and small mistake in man page.
3539
3540 2. Added 4th condition (GPL supersedes if conflict) and created separate
3541 LICENCE file containing the conditions.
3542
3543 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3544 Perl, that is the internal \ allows the delimiter to be included in the
3545 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3546 the final delimiter, add \ to the end of the pattern (to test the error).
3547
3548 4. Added the convenience functions for extracting substrings after a successful
3549 match. Updated pcretest to make it able to test these functions.
3550
3551
3552 Version 2.02 14-Jan-99
3553 ----------------------
3554
3555 1. Initialized the working variables associated with each extraction so that
3556 their saving and restoring doesn't refer to uninitialized store.
3557
3558 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3559 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3560 fix the problem.
3561
3562 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3563 calls, and wasn't printing the correct value for compiling calls. Increased the
3564 default value of LOOPREPEAT, and the number of significant figures in the
3565 times.
3566
3567 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3568
3569 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3570 a building problem on Windows NT with a FAT file system.
3571
3572
3573 Version 2.01 21-Oct-98
3574 ----------------------
3575
3576 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3577 to character tables built by pcre_maketables() in the current locale. If NULL
3578 is passed, the default tables are used.
3579
3580
3581 Version 2.00 24-Sep-98
3582 ----------------------
3583
3584 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3585 it any more.
3586
3587 2. Allow quantification of (?>) groups, and make it work correctly.
3588
3589 3. The first character computation wasn't working for (?>) groups.
3590
3591 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3592 end of the subject) and add 5.005's \z, which really does match only at the
3593 very end of the subject.
3594
3595 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3596
3597 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3598 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3599 localized options. All options to pcre_study() were also removed.
3600
3601 7. Add other new features from 5.005:
3602
3603 $(?<= positive lookbehind
3604 $(?<! negative lookbehind
3605 (?imsx-imsx) added the unsetting capability
3606 such a setting is global if at outer level; local otherwise
3607 (?imsx-imsx:) non-capturing groups with option setting
3608 (?(cond)re|re) conditional pattern matching
3609
3610 A backreference to itself in a repeated group matches the previous
3611 captured string.
3612
3613 8. General tidying up of studying (both automatic and via "study")
3614 consequential on the addition of new assertions.
3615
3616 9. As in 5.005, unlimited repeated groups that could match an empty substring
3617 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3618 runtime if any iteration does actually match an empty substring.
3619
3620 10. Include the RunTest script in the distribution.
3621
3622 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3623 discrepancies, some of which were old and were also with respect to 5.004. They
3624 have now been fixed.
3625
3626
3627 Version 1.09 28-Apr-98
3628 ----------------------
3629
3630 1. A negated single character class followed by a quantifier with a minimum
3631 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3632 program crashes, or just wrong answers. This did not apply to negated classes
3633 containing more than one character, or to minima other than one.
3634
3635
3636 Version 1.08 27-Mar-98
3637 ----------------------
3638
3639 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3640
3641 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3642 latter must appear before anything that relies on it in the pattern.
3643
3644
3645 Version 1.07 16-Feb-98
3646 ----------------------
3647
3648 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3649 repeat of a potentially empty string).
3650
3651
3652 Version 1.06 23-Jan-98
3653 ----------------------
3654
3655 1. Added Markus Oberhumer's little patches for C++.
3656
3657 2. Literal strings longer than 255 characters were broken.
3658
3659
3660 Version 1.05 23-Dec-97
3661 ----------------------
3662
3663 1. Negated character classes containing more than one character were failing if
3664 PCRE_CASELESS was set at run time.
3665
3666
3667 Version 1.04 19-Dec-97
3668 ----------------------
3669
3670 1. Corrected the man page, where some "const" qualifiers had been omitted.
3671
3672 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3673 input syntax.
3674
3675 3. Fixed memory leak which occurred when a regex with back references was
3676 matched with an offsets vector that wasn't big enough. The temporary memory
3677 that is used in this case wasn't being freed if the match failed.
3678
3679 4. Tidied pcretest to ensure it frees memory that it gets.
3680
3681 5. Temporary memory was being obtained in the case where the passed offsets
3682 vector was exactly big enough.
3683
3684 6. Corrected definition of offsetof() from change 5 below.
3685
3686 7. I had screwed up change 6 below and broken the rules for the use of
3687 setjmp(). Now fixed.
3688
3689
3690 Version 1.03 18-Dec-97
3691 ----------------------
3692
3693 1. A erroneous regex with a missing opening parenthesis was correctly
3694 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3695 on some systems.
3696
3697 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3698 it was reported that one broken compiler failed on the former because "code" is
3699 also an independent variable.
3700
3701 3. The erroneous regex a[]b caused an array overrun reference.
3702
3703 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3704 fail on data ending with that character. (It was going on too far, and checking
3705 the next character, typically a binary zero.) This was specific to the
3706 optimized code for single-character negative classes.
3707
3708 5. Added a contributed patch from the TIN world which does the following:
3709
3710 + Add an undef for memmove, in case the the system defines a macro for it.
3711
3712 + Add a definition of offsetof(), in case there isn't one. (I don't know
3713 the reason behind this - offsetof() is part of the ANSI standard - but
3714 it does no harm).
3715
3716 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3717 most of the places where whitespace preceded '#'. I have given up and
3718 allowed the remaining 2 cases to be at the margin.
3719
3720 + Rename some variables in pcre to eliminate shadowing. This seems very
3721 pedantic, but does no harm, of course.
3722
3723 6. Moved the call to setjmp() into its own function, to get rid of warnings
3724 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3725
3726 7. Constructs such as \d{8,} were compiling into the equivalent of
3727 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3728 outcome, but in this particular case used more store than had been allocated,
3729 which caused the bug to be discovered because it threw up an internal error.
3730
3731 8. The debugging code in both pcre and pcretest for outputting the compiled
3732 form of a regex was going wrong in the case of back references followed by
3733 curly-bracketed repeats.
3734
3735
3736 Version 1.02 12-Dec-97
3737 ----------------------
3738
3739 1. Typos in pcre.3 and comments in the source fixed.
3740
3741 2. Applied a contributed patch to get rid of places where it used to remove
3742 'const' from variables, and fixed some signed/unsigned and uninitialized
3743 variable warnings.
3744
3745 3. Added the "runtest" target to Makefile.
3746
3747 4. Set default compiler flag to -O2 rather than just -O.
3748
3749
3750 Version 1.01 19-Nov-97
3751 ----------------------
3752
3753 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3754 like /([ab]*)*/, that is, for classes with more than one character in them.
3755
3756 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3757 as /((?>a*))*/ (a PCRE_EXTRA facility).
3758
3759
3760 Version 1.00 18-Nov-97
3761 ----------------------
3762
3763 1. Added compile-time macros to support systems such as SunOS4 which don't have
3764 memmove() or strerror() but have other things that can be used instead.
3765
3766 2. Arranged that "make clean" removes the executables.
3767
3768
3769 Version 0.99 27-Oct-97
3770 ----------------------
3771
3772 1. Fixed bug in code for optimizing classes with only one character. It was
3773 initializing a 32-byte map regardless, which could cause it to run off the end
3774 of the memory it had got.
3775
3776 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3777
3778
3779 Version 0.98 22-Oct-97
3780 ----------------------
3781
3782 1. Fixed bug in code for handling temporary memory usage when there are more
3783 back references than supplied space in the ovector. This could cause segfaults.
3784
3785
3786 Version 0.97 21-Oct-97
3787 ----------------------
3788
3789 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3790
3791 2. Optimized negated single characters not to use a bit map.
3792
3793 3. Brought error texts together as macro definitions; clarified some of them;
3794 fixed one that was wrong - it said "range out of order" when it meant "invalid
3795 escape sequence".
3796
3797 4. Changed some char * arguments to const char *.
3798
3799 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3800
3801 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3802 pcretest.
3803
3804
3805 Version 0.96 16-Oct-97
3806 ----------------------
3807
3808 1. Added a simple "pgrep" utility to the distribution.
3809
3810 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3811 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3812 where "ddd" means "one or more decimal digits".
3813
3814 3. Fixed serious bug. If a pattern had a back reference, but the call to
3815 pcre_exec() didn't supply a large enough ovector to record the related
3816 identifying subpattern, the match always failed. PCRE now remembers the number
3817 of the largest back reference, and gets some temporary memory in which to save
3818 the offsets during matching if necessary, in order to ensure that
3819 backreferences always work.
3820
3821 4. Increased the compatibility with Perl in a number of ways:
3822
3823 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3824 to request this handling. The option can be set at compile or exec time.
3825
3826 (b) $ matches before a terminating newline by default; an option
3827 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3828 mode). The option can be set at compile or exec time.
3829
3830 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3831 the same as Perl's. If the decimal number it represents is less than 10
3832 or there aren't that many previous left capturing parentheses, an octal
3833 escape is read. Inside a character class, it's always an octal escape,
3834 even if it is a single digit.
3835
3836 (d) An escaped but undefined alphabetic character is taken as a literal,
3837 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3838 escapes.
3839
3840 (e) {0} is now permitted. (The previous item is removed from the compiled
3841 pattern).
3842
3843 5. Changed all the names of code files so that the basic parts are no longer
3844 than 10 characters, and abolished the teeny "globals.c" file.
3845
3846 6. Changed the handling of character classes; they are now done with a 32-byte
3847 bit map always.
3848
3849 7. Added the -d and /D options to pcretest to make it possible to look at the
3850 internals of compilation without having to recompile pcre.
3851
3852
3853 Version 0.95 23-Sep-97
3854 ----------------------
3855
3856 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3857 \x20 at the start of a run of normal characters. These were being treated as
3858 real characters, instead of the source characters being re-checked.
3859
3860
3861 Version 0.94 18-Sep-97
3862 ----------------------
3863
3864 1. The functions are now thread-safe, with the caveat that the global variables
3865 containing pointers to malloc() and free() or alternative functions are the
3866 same for all threads.
3867
3868 2. Get pcre_study() to generate a bitmap of initial characters for non-
3869 anchored patterns when this is possible, and use it if passed to pcre_exec().
3870
3871
3872 Version 0.93 15-Sep-97
3873 ----------------------
3874
3875 1. /(b)|(:+)/ was computing an incorrect first character.
3876
3877 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3878 but not actually doing anything yet.
3879
3880 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3881 as Perl does (e.g. [-az] or [az-]).
3882
3883 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3884 all possible positions.
3885
3886 5. Split up into different modules to avoid including unneeded functions in a
3887 compiled binary. However, compile and exec are still in one module. The "study"
3888 function is split off.
3889
3890 6. The character tables are now in a separate module whose source is generated
3891 by an auxiliary program - but can then be edited by hand if required. There are
3892 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3893 toupper() in the code.
3894
3895 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3896 make them global. Abolish the function for setting them, as the caller can now
3897 set them directly.
3898
3899
3900 Version 0.92 11-Sep-97
3901 ----------------------
3902
3903 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3904 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3905
3906 2. Caseless matching was not working in character classes if the characters in
3907 the pattern were in upper case.
3908
3909 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3910
3911 4. Make PCRE_ANCHORED public and accept as a compile option.
3912
3913 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3914 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3915 pass them.
3916
3917 6. Give an error if bad option bits passed at compile or run time.
3918
3919 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3920 pcretest to cause it to pass that flag.
3921
3922 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3923 options, and the first character, if set.
3924
3925 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3926
3927
3928 Version 0.91 10-Sep-97
3929 ----------------------
3930
3931 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3932 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3933
3934 2. PCRE was looping on encountering an indefinitely repeated back reference to
3935 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3936 Perl does - treats the match as successful.
3937
3938 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5