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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 532 - (show annotations)
Tue Jun 1 16:21:42 2010 UTC (4 years, 11 months ago) by ph10
File size: 188309 byte(s)
Error occurred while calculating annotation data.
Extend auto-possessify to handle some Unicode properties.
1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.10 03 May-2010
5 ------------------------
6
7 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
8 THEN.
9
10 2. (*ACCEPT) was not working when inside an atomic group.
11
12 3. Inside a character class, \B is treated as a literal by default, but
13 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14 causes the error). The code is unchanged, but I tidied the documentation.
15
16 4. Inside a character class, PCRE always treated \R and \X as literals,
17 whereas Perl faults them if its -w option is set. I have changed PCRE so
18 that it faults them when PCRE_EXTRA is set.
19
20 5. Added support for \N, which always matches any character other than
21 newline. (It is the same as "." when PCRE_DOTALL is not set.)
22
23 6. When compiling pcregrep with newer versions of gcc which may have
24 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25 declared with attribute warn_unused_result" were given. Just casting the
26 result to (void) does not stop the warnings; a more elaborate fudge is
27 needed. I've used a macro to implement this.
28
29 7. Minor change to pcretest.c to avoid a compiler warning.
30
31 8. Added four artifical Unicode properties to help with an option to make
32 \s etc use properties (see next item). The new properties are: Xan
33 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
34
35 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
36 use Unicode properties. (*UCP) at the start of a pattern can be used to set
37 this option. Modified pcretest to add /W to test this facility. Added
38 REG_UCP to make it available via the POSIX interface.
39
40 10. Added --line-buffered to pcregrep.
41
42 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
43 studied, and the match started with a letter with a code point greater than
44 127 whose first byte was different to the first byte of the other case of
45 the letter, the other case of this starting letter was not recognized.
46
47 12. pcreposix.c included pcre.h before including pcre_internal.h. This caused a
48 conflict in the definition of PCRE_EXP_DECL. I have removed the include of
49 pcre.h as pcre_internal.h includes pcre.h itself. (This may be a bit of
50 historical tidying that never got done.)
51
52 13. If a pattern that was studied started with a repeated Unicode property
53 test, for example, \p{Nd}+, there was the theoretical possibility of
54 setting up an incorrect bitmap of starting bytes, but fortunately it could
55 not have actually happened in practice until change 8 above was made (it
56 added property types that matched character-matching opcodes).
57
58 14. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
59 possible starting bytes for non-anchored patterns.
60
61 15. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
62 \R, and also a number of cases that involve Unicode properties, both
63 explicit and implicit when PCRE_UCP is set.
64
65 16. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
66 input, it could crash or give wrong results if characters with values
67 greater than 0xc0 were present in the subject string. (Detail: it assumed
68 UTF-8 input when processing these items.)
69
70 17. Added a lot of (int) casts to avoid compiler warnings in systems where
71 size_t is 64-bit.
72
73 18. Added a check for running out of memory when PCRE is compiled with
74 --disable-stack-for-recursion.
75
76
77
78 Version 8.02 19-Mar-2010
79 ------------------------
80
81 1. The Unicode data tables have been updated to Unicode 5.2.0.
82
83 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
84 configured.
85
86 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
87 original author of that file, following a query about its status.
88
89 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
90 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
91
92 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
93 quantifier applied to a forward-referencing subroutine call, could compile
94 incorrect code or give the error "internal error: previously-checked
95 referenced subpattern not found".
96
97 6. Both MS Visual Studio and Symbian OS have problems with initializing
98 variables to point to external functions. For these systems, therefore,
99 pcre_malloc etc. are now initialized to local functions that call the
100 relevant global functions.
101
102 7. There were two entries missing in the vectors called coptable and poptable
103 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
104 I've fixed the data, and added a kludgy way of testing at compile time that
105 the lengths are correct (equal to the number of opcodes).
106
107 8. Following on from 7, I added a similar kludge to check the length of the
108 eint vector in pcreposix.c.
109
110 9. Error texts for pcre_compile() are held as one long string to avoid too
111 much relocation at load time. To find a text, the string is searched,
112 counting zeros. There was no check for running off the end of the string,
113 which could happen if a new error number was added without updating the
114 string.
115
116 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
117
118 11. \K was not working if it appeared in an atomic group or in a group that
119 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
120 \K is "not well defined" if used in an assertion. PCRE now accepts it if
121 the assertion is positive, but not if it is negative.
122
123 12. Change 11 fortuitously reduced the size of the stack frame used in the
124 "match()" function of pcre_exec.c by one pointer. Forthcoming
125 implementation of support for (*MARK) will need an extra pointer on the
126 stack; I have reserved it now, so that the stack frame size does not
127 decrease.
128
129 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
130 item in branch that calls a recursion is a subroutine call - as in the
131 second branch in the above example - was incorrectly given the compile-
132 time error "recursive call could loop indefinitely" because pcre_compile()
133 was not correctly checking the subroutine for matching a non-empty string.
134
135 14. The checks for overrunning compiling workspace could trigger after an
136 overrun had occurred. This is a "should never occur" error, but it can be
137 triggered by pathological patterns such as hundreds of nested parentheses.
138 The checks now trigger 100 bytes before the end of the workspace.
139
140 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
141
142
143 Version 8.01 19-Jan-2010
144 ------------------------
145
146 1. If a pattern contained a conditional subpattern with only one branch (in
147 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
148 computed the wrong minimum data length (which is of course zero for such
149 subpatterns). This could cause incorrect "no match" results.
150
151 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
152 the pattern is reset in the first branch, pcre_compile() failed with
153 "internal error: code overflow at offset...". This happened only when
154 the reset was to the original external option setting. (An optimization
155 abstracts leading options settings into an external setting, which was the
156 cause of this.)
157
158 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
159 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
160 assertion pattern did not match (meaning that the assertion was true), it
161 was incorrectly treated as false if the SKIP had been reached during the
162 matching. This also applied to assertions used as conditions.
163
164 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
165 assertion subpattern, including such a pattern used as a condition,
166 unpredictable results occurred, instead of the error return
167 PCRE_ERROR_DFA_UITEM.
168
169 5. The C++ GlobalReplace function was not working like Perl for the special
170 situation when an empty string is matched. It now does the fancy magic
171 stuff that is necessary.
172
173 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
174 removed. (These were left over from very, very early versions of PCRE.)
175
176 7. Some cosmetic changes to the code to make life easier when compiling it
177 as part of something else:
178
179 (a) Change DEBUG to PCRE_DEBUG.
180
181 (b) In pcre_compile(), rename the member of the "branch_chain" structure
182 called "current" as "current_branch", to prevent a collision with the
183 Linux macro when compiled as a kernel module.
184
185 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
186 prevent a collision with the Linux macro when compiled as a kernel
187 module.
188
189 8. In pcre_compile() there are some checks for integer overflows that used to
190 cast potentially large values to (double). This has been changed to that
191 when building, a check for int64_t is made, and if it is found, it is used
192 instead, thus avoiding the use of floating point arithmetic. (There is no
193 other use of FP in PCRE.) If int64_t is not found, the fallback is to
194 double.
195
196 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
197 2005 (difference between two addresses compared to an unsigned value).
198
199 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
200 custom one, because of the following reported problem in Windows:
201
202 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
203 under Win32.
204 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
205 therefore missing the function definition.
206 - The compiler thus generates a "C" signature for the test function.
207 - The linker fails to find the "C" function.
208 - PCRE fails to configure if asked to do so against libbz2.
209
210 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
211 messages were output:
212
213 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
214 rerunning libtoolize, to keep the correct libtool macros in-tree.
215 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
216
217 I have done both of these things.
218
219 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
220 most of the time, it *can* run out if it is given a pattern that contains a
221 runaway infinite recursion. I updated the discussion in the pcrestack man
222 page.
223
224 13. Now that we have gone to the x.xx style of version numbers, the minor
225 version may start with zero. Using 08 or 09 is a bad idea because users
226 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
227 interpreted as invalid octal numbers. I've updated the previous comment in
228 configure.ac, and also added a check that gives an error if 08 or 09 are
229 used.
230
231 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
232 causing partial matching to fail when the end of the subject matched \W
233 in a UTF-8 pattern where \W was quantified with a minimum of 3.
234
235 15. There were some discrepancies between the declarations in pcre_internal.h
236 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
237 their definitions. The declarations used "const uschar *" and the
238 definitions used USPTR. Even though USPTR is normally defined as "const
239 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
240 reported that: "This difference in casting confuses some C++ compilers, for
241 example, SunCC recognizes above declarations as different functions and
242 generates broken code for hbpcre." I have changed the declarations to use
243 USPTR.
244
245 16. GNU libtool is named differently on some systems. The autogen.sh script now
246 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
247 (FreeBSD).
248
249 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
250 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
251 comment: "Figure out how to create a longlong from a string: strtoll and
252 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
253 instance, but it only takes 2 args instead of 3!"
254
255 18. A subtle bug concerned with back references has been fixed by a change of
256 specification, with a corresponding code fix. A pattern such as
257 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
258 refers, was giving matches when it shouldn't. For example, xa=xaaa would
259 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
260 same bug. Such groups have to be quantified to be useful, or contained
261 inside another quantified group. (If there's no repetition, the reference
262 can never match.) The problem arises because, having left the group and
263 moved on to the rest of the pattern, a later failure that backtracks into
264 the group uses the captured value from the final iteration of the group
265 rather than the correct earlier one. I have fixed this in PCRE by forcing
266 any group that contains a reference to itself to be an atomic group; that
267 is, there cannot be any backtracking into it once it has completed. This is
268 similar to recursive and subroutine calls.
269
270
271 Version 8.00 19-Oct-09
272 ----------------------
273
274 1. The table for translating pcre_compile() error codes into POSIX error codes
275 was out-of-date, and there was no check on the pcre_compile() error code
276 being within the table. This could lead to an OK return being given in
277 error.
278
279 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
280 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
281 in a Windows environment.
282
283 3. The pcregrep --count option prints the count for each file even when it is
284 zero, as does GNU grep. However, pcregrep was also printing all files when
285 --files-with-matches was added. Now, when both options are given, it prints
286 counts only for those files that have at least one match. (GNU grep just
287 prints the file name in this circumstance, but including the count seems
288 more useful - otherwise, why use --count?) Also ensured that the
289 combination -clh just lists non-zero counts, with no names.
290
291 4. The long form of the pcregrep -F option was incorrectly implemented as
292 --fixed_strings instead of --fixed-strings. This is an incompatible change,
293 but it seems right to fix it, and I didn't think it was worth preserving
294 the old behaviour.
295
296 5. The command line items --regex=pattern and --regexp=pattern were not
297 recognized by pcregrep, which required --regex pattern or --regexp pattern
298 (with a space rather than an '='). The man page documented the '=' forms,
299 which are compatible with GNU grep; these now work.
300
301 6. No libpcreposix.pc file was created for pkg-config; there was just
302 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
303
304 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
305 when UCP support is not needed, by modifying the Python script that
306 generates it from Unicode data files. This should not matter if the module
307 is correctly used as a library, but I received one complaint about 50K of
308 unwanted data. My guess is that the person linked everything into his
309 program rather than using a library. Anyway, it does no harm.
310
311 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
312 was a minimum greater than 1 for a wide character in a possessive
313 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
314 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
315 character. Chaos in the form of incorrect output or a compiling loop could
316 result.
317
318 9. The restrictions on what a pattern can contain when partial matching is
319 requested for pcre_exec() have been removed. All patterns can now be
320 partially matched by this function. In addition, if there are at least two
321 slots in the offset vector, the offset of the earliest inspected character
322 for the match and the offset of the end of the subject are set in them when
323 PCRE_ERROR_PARTIAL is returned.
324
325 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
326 synonymous with PCRE_PARTIAL, for backwards compatibility, and
327 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
328 and may be more useful for multi-segment matching.
329
330 11. Partial matching with pcre_exec() is now more intuitive. A partial match
331 used to be given if ever the end of the subject was reached; now it is
332 given only if matching could not proceed because another character was
333 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
334 string "Z", which now yields "no match" instead of "partial match". In the
335 case of pcre_dfa_exec(), "no match" is given if every matching path for the
336 final character ended with (*FAIL).
337
338 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
339 if the pattern had a "must contain" character that was already found in the
340 earlier partial match, unless partial matching was again requested. For
341 example, with the pattern /dog.(body)?/, the "must contain" character is
342 "g". If the first part-match was for the string "dog", restarting with
343 "sbody" failed. This bug has been fixed.
344
345 13. The string returned by pcre_dfa_exec() after a partial match has been
346 changed so that it starts at the first inspected character rather than the
347 first character of the match. This makes a difference only if the pattern
348 starts with a lookbehind assertion or \b or \B (\K is not supported by
349 pcre_dfa_exec()). It's an incompatible change, but it makes the two
350 matching functions compatible, and I think it's the right thing to do.
351
352 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
353 so that the demonstration program is easily available in environments where
354 PCRE has not been installed from source.
355
356 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
357 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
358 library.
359
360 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
361 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
362 is not the first non-POSIX option to be added. Clearly some people find
363 these options useful.
364
365 17. If a caller to the POSIX matching function regexec() passes a non-zero
366 value for nmatch with a NULL value for pmatch, the value of
367 nmatch is forced to zero.
368
369 18. RunGrepTest did not have a test for the availability of the -u option of
370 the diff command, as RunTest does. It now checks in the same way as
371 RunTest, and also checks for the -b option.
372
373 19. If an odd number of negated classes containing just a single character
374 interposed, within parentheses, between a forward reference to a named
375 subpattern and the definition of the subpattern, compilation crashed with
376 an internal error, complaining that it could not find the referenced
377 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
378 [The bug was that it was starting one character too far in when skipping
379 over the character class, thus treating the ] as data rather than
380 terminating the class. This meant it could skip too much.]
381
382 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
383 /g option in pcretest when the pattern contains \K, which makes it possible
384 to have an empty string match not at the start, even when the pattern is
385 anchored. Updated pcretest and pcredemo to use this option.
386
387 21. If the maximum number of capturing subpatterns in a recursion was greater
388 than the maximum at the outer level, the higher number was returned, but
389 with unset values at the outer level. The correct (outer level) value is
390 now given.
391
392 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
393 PCRE did not set those parentheses (unlike Perl). I have now found a way to
394 make it do so. The string so far is captured, making this feature
395 compatible with Perl.
396
397 23. The tests have been re-organized, adding tests 11 and 12, to make it
398 possible to check the Perl 5.10 features against Perl 5.10.
399
400 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
401 pattern matches a fixed length string. PCRE did not allow this; now it
402 does. Neither allows recursion.
403
404 25. I finally figured out how to implement a request to provide the minimum
405 length of subject string that was needed in order to match a given pattern.
406 (It was back references and recursion that I had previously got hung up
407 on.) This code has now been added to pcre_study(); it finds a lower bound
408 to the length of subject needed. It is not necessarily the greatest lower
409 bound, but using it to avoid searching strings that are too short does give
410 some useful speed-ups. The value is available to calling programs via
411 pcre_fullinfo().
412
413 26. While implementing 25, I discovered to my embarrassment that pcretest had
414 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
415 study optimizations had never been tested with that matching function.
416 Oops. What is worse, even when it was passed study data, there was a bug in
417 pcre_dfa_exec() that meant it never actually used it. Double oops. There
418 were also very few tests of studied patterns with pcre_dfa_exec().
419
420 27. If (?| is used to create subpatterns with duplicate numbers, they are now
421 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
422 on the other side of the coin, they are no longer allowed to have different
423 names, because these cannot be distinguished in PCRE, and this has caused
424 confusion. (This is a difference from Perl.)
425
426 28. When duplicate subpattern names are present (necessarily with different
427 numbers, as required by 27 above), and a test is made by name in a
428 conditional pattern, either for a subpattern having been matched, or for
429 recursion in such a pattern, all the associated numbered subpatterns are
430 tested, and the overall condition is true if the condition is true for any
431 one of them. This is the way Perl works, and is also more like the way
432 testing by number works.
433
434
435 Version 7.9 11-Apr-09
436 ---------------------
437
438 1. When building with support for bzlib/zlib (pcregrep) and/or readline
439 (pcretest), all targets were linked against these libraries. This included
440 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
441 libraries. This caused unwanted dependencies to be created. This problem
442 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
443 pcretest is linked with readline.
444
445 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
446 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
447 moved outside it again, because FALSE and TRUE are already defined in AIX,
448 but BOOL is not.
449
450 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
451 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
452
453 4. The pcregrep documentation said spaces were inserted as well as colons (or
454 hyphens) following file names and line numbers when outputting matching
455 lines. This is not true; no spaces are inserted. I have also clarified the
456 wording for the --colour (or --color) option.
457
458 5. In pcregrep, when --colour was used with -o, the list of matching strings
459 was not coloured; this is different to GNU grep, so I have changed it to be
460 the same.
461
462 6. When --colo(u)r was used in pcregrep, only the first matching substring in
463 each matching line was coloured. Now it goes on to look for further matches
464 of any of the test patterns, which is the same behaviour as GNU grep.
465
466 7. A pattern that could match an empty string could cause pcregrep to loop; it
467 doesn't make sense to accept an empty string match in pcregrep, so I have
468 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
469 seems to be how GNU grep behaves.
470
471 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
472 start or after a newline", because the conditional assertion was not being
473 correctly handled. The rule now is that both the assertion and what follows
474 in the first alternative must satisfy the test.
475
476 9. If auto-callout was enabled in a pattern with a conditional group whose
477 condition was an assertion, PCRE could crash during matching, both with
478 pcre_exec() and pcre_dfa_exec().
479
480 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
481 used for matching.
482
483 11. Unicode property support in character classes was not working for
484 characters (bytes) greater than 127 when not in UTF-8 mode.
485
486 12. Added the -M command line option to pcretest.
487
488 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
489
490 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
491
492 16. Added comments and documentation about mis-use of no_arg in the C++
493 wrapper.
494
495 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
496 from Martin Jerabek that uses macro names for all relevant character and
497 string constants.
498
499 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
500 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
501 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
502 these, but not everybody uses configure.
503
504 19. A conditional group that had only one branch was not being correctly
505 recognized as an item that could match an empty string. This meant that an
506 enclosing group might also not be so recognized, causing infinite looping
507 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
508 with the subject "ab", where knowledge that the repeated group can match
509 nothing is needed in order to break the loop.
510
511 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
512 exec(), but without supplying a callout function, matching went wrong.
513
514 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
515 leak if the size of the offset vector was greater than 30. When the vector
516 is smaller, the saved offsets during recursion go onto a local stack
517 vector, but for larger vectors malloc() is used. It was failing to free
518 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
519 error, in fact).
520
521 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
522 heapframe that is used only when UTF-8 support is enabled. This caused no
523 problem, but was untidy.
524
525 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
526 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
527 included within another project.
528
529 24. Steven Van Ingelgem's patches to add more options to the CMake support,
530 slightly modified by me:
531
532 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
533 not building pcregrep.
534
535 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
536 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
537
538 25. Forward references, both numeric and by name, in patterns that made use of
539 duplicate group numbers, could behave incorrectly or give incorrect errors,
540 because when scanning forward to find the reference group, PCRE was not
541 taking into account the duplicate group numbers. A pattern such as
542 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
543
544 26. Changed a few more instances of "const unsigned char *" to USPTR, making
545 the feature of a custom pointer more persuasive (as requested by a user).
546
547 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
548 pcretest.c, inside #ifndefs, because it seems they are sometimes already
549 pre-defined.
550
551 28. Added support for (*UTF8) at the start of a pattern.
552
553 29. Arrange for flags added by the "release type" setting in CMake to be shown
554 in the configuration summary.
555
556
557 Version 7.8 05-Sep-08
558 ---------------------
559
560 1. Replaced UCP searching code with optimized version as implemented for Ad
561 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
562 stage table and inline lookup instead of a function, giving speed ups of 2
563 to 5 times on some simple patterns that I tested. Permission was given to
564 distribute the MultiStage2.py script that generates the tables (it's not in
565 the tarball, but is in the Subversion repository).
566
567 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
568 scripts.
569
570 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
571 a group with a zero qualifier. The result of the study could be incorrect,
572 or the function might crash, depending on the pattern.
573
574 4. Caseless matching was not working for non-ASCII characters in back
575 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
576 It now works when Unicode Property Support is available.
577
578 5. In pcretest, an escape such as \x{de} in the data was always generating
579 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
580 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
581 truncation.
582
583 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
584
585 7. Added two (int) casts to pcregrep when printing the difference of two
586 pointers, in case they are 64-bit values.
587
588 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
589 test 2 if it fails.
590
591 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
592 and a #define of that name to empty if it is not externally set. This is to
593 allow users of MSVC to set it if necessary.
594
595 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
596 the convenience functions in the pcre_get.c source file.
597
598 11. An option change at the start of a pattern that had top-level alternatives
599 could cause overwriting and/or a crash. This command provoked a crash in
600 some environments:
601
602 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
603
604 This potential security problem was recorded as CVE-2008-2371.
605
606 12. For a pattern where the match had to start at the beginning or immediately
607 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
608 pcre_dfa_exec() could read past the end of the passed subject if there was
609 no match. To help with detecting such bugs (e.g. with valgrind), I modified
610 pcretest so that it places the subject at the end of its malloc-ed buffer.
611
612 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
613 exec() might read past the end of the data buffer in UTF-8 mode.
614
615 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
616 the data contained the byte 0x85 as part of a UTF-8 character within its
617 first line. This applied both to normal and DFA matching.
618
619 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
620 /^[^d]*?$/8 failed to match "abc".
621
622 16. Added a missing copyright notice to pcrecpp_internal.h.
623
624 17. Make it more clear in the documentation that values returned from
625 pcre_exec() in ovector are byte offsets, not character counts.
626
627 18. Tidied a few places to stop certain compilers from issuing warnings.
628
629 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
630 supplied by Stefan Weber. I made a further small update for 7.8 because
631 there is a change of source arrangements: the pcre_searchfuncs.c module is
632 replaced by pcre_ucd.c.
633
634
635 Version 7.7 07-May-08
636 ---------------------
637
638 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
639 a string to a long long, pretend we don't even have a long long." This is
640 done by checking for the strtoq, strtoll, and _strtoi64 functions.
641
642 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
643 pre-7.6 versions, which defined a global no_arg variable instead of putting
644 it in the RE class. (See also #8 below.)
645
646 3. Remove a line of dead code, identified by coverity and reported by Nuno
647 Lopes.
648
649 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
650
651 (1) The include/exclude patterns were being applied to the whole pathnames
652 of files, instead of just to the final components.
653
654 (2) If there was more than one level of directory, the subdirectories were
655 skipped unless they satisfied the include/exclude conditions. This is
656 inconsistent with GNU grep (and could even be seen as contrary to the
657 pcregrep specification - which I improved to make it absolutely clear).
658 The action now is always to scan all levels of directory, and just
659 apply the include/exclude patterns to regular files.
660
661 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
662 --exclude_dir in the tests to avoid scanning .svn directories.
663
664 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
665 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
666 doesn't support NULs in patterns.
667
668 7. Added some missing "const"s to declarations of static tables in
669 pcre_compile.c and pcre_dfa_exec.c.
670
671 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
672 caused by fix #2 above. (Subsequently also a second patch to fix the
673 first patch. And a third patch - this was a messy problem.)
674
675 9. Applied Craig's patch to remove the use of push_back().
676
677 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
678 matching function regexec().
679
680 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
681 which, however, unlike Perl's \g{...}, are subroutine calls, not back
682 references. PCRE supports relative numbers with this syntax (I don't think
683 Oniguruma does).
684
685 12. Previously, a group with a zero repeat such as (...){0} was completely
686 omitted from the compiled regex. However, this means that if the group
687 was called as a subroutine from elsewhere in the pattern, things went wrong
688 (an internal error was given). Such groups are now left in the compiled
689 pattern, with a new opcode that causes them to be skipped at execution
690 time.
691
692 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
693 to the way PCRE behaves:
694
695 (a) A lone ] character is dis-allowed (Perl treats it as data).
696
697 (b) A back reference to an unmatched subpattern matches an empty string
698 (Perl fails the current match path).
699
700 (c) A data ] in a character class must be notated as \] because if the
701 first data character in a class is ], it defines an empty class. (In
702 Perl it is not possible to have an empty class.) The empty class []
703 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
704 The negative empty class [^] matches any one character, independently
705 of the DOTALL setting.
706
707 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
708 non-existent subpattern following a character class starting with ']' and
709 containing () gave an internal compiling error instead of "reference to
710 non-existent subpattern". Fortunately, when the pattern did exist, the
711 compiled code was correct. (When scanning forwards to check for the
712 existencd of the subpattern, it was treating the data ']' as terminating
713 the class, so got the count wrong. When actually compiling, the reference
714 was subsequently set up correctly.)
715
716 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
717 it was being rejected as not supported by pcre_dfa_exec(), even though
718 other assertions are supported. I have made pcre_dfa_exec() support
719 (*FAIL).
720
721 16. The implementation of 13c above involved the invention of a new opcode,
722 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
723 cannot be changed at match time, I realized I could make a small
724 improvement to matching performance by compiling OP_ALLANY instead of
725 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
726 on the OP_ANY path.
727
728 17. Compiling pcretest on Windows with readline support failed without the
729 following two fixes: (1) Make the unistd.h include conditional on
730 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
731
732 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
733 ncurses library to be included for pcretest when ReadLine support is
734 requested, but also to allow for it to be overridden. This patch came from
735 Daniel Bergström.
736
737 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
738 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
739 any errors with the current Unicode tables. Thanks to Peter Kankowski for
740 spotting this.
741
742
743 Version 7.6 28-Jan-08
744 ---------------------
745
746 1. A character class containing a very large number of characters with
747 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
748 overflow.
749
750 2. Patch to cut out the "long long" test in pcrecpp_unittest when
751 HAVE_LONG_LONG is not defined.
752
753 3. Applied Christian Ehrlicher's patch to update the CMake build files to
754 bring them up to date and include new features. This patch includes:
755
756 - Fixed PH's badly added libz and libbz2 support.
757 - Fixed a problem with static linking.
758 - Added pcredemo. [But later removed - see 7 below.]
759 - Fixed dftables problem and added an option.
760 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
761 HAVE_LONG_LONG.
762 - Added readline support for pcretest.
763 - Added an listing of the option settings after cmake has run.
764
765 4. A user submitted a patch to Makefile that makes it easy to create
766 "pcre.dll" under mingw when using Configure/Make. I added stuff to
767 Makefile.am that cause it to include this special target, without
768 affecting anything else. Note that the same mingw target plus all
769 the other distribution libraries and programs are now supported
770 when configuring with CMake (see 6 below) instead of with
771 Configure/Make.
772
773 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
774 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
775 exported in the Windows port". It has not yet been confirmed that the patch
776 solves the problem, but it does no harm.
777
778 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
779 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
780 with CMake, and also correct the comment about stack recursion.
781
782 7. Remove the automatic building of pcredemo from the ./configure system and
783 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
784 of a program that users should build themselves after PCRE is installed, so
785 building it automatically is not really right. What is more, it gave
786 trouble in some build environments.
787
788 8. Further tidies to CMakeLists.txt from Sheri and Christian.
789
790
791 Version 7.5 10-Jan-08
792 ---------------------
793
794 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
795 values in parens when parsing an RE using the C++ wrapper."
796
797 2. Negative specials like \S did not work in character classes in UTF-8 mode.
798 Characters greater than 255 were excluded from the class instead of being
799 included.
800
801 3. The same bug as (2) above applied to negated POSIX classes such as
802 [:^space:].
803
804 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
805 defined or documented. It seems to have been a typo for PCRE_STATIC, so
806 I have changed it.
807
808 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
809 first named subpattern) and a construct such as (?&a) would reference the
810 first named subpattern whose name started with "a" (in other words, the
811 length check was missing). Both these problems are fixed. "Subpattern name
812 expected" is now given for (?&) (a zero-length name), and this patch also
813 makes it give the same error for \k'' (previously it complained that that
814 was a reference to a non-existent subpattern).
815
816 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
817 this is right because (?- can be followed by option settings as well as by
818 digits. I have, however, made the messages clearer.
819
820 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
821 than the number used in the conditional) now cause a compile-time error.
822 This is actually not compatible with Perl, which accepts such patterns, but
823 treats the conditional as always being FALSE (as PCRE used to), but it
824 seems to me that giving a diagnostic is better.
825
826 8. Change "alphameric" to the more common word "alphanumeric" in comments
827 and messages.
828
829 9. Fix two occurrences of "backslash" in comments that should have been
830 "backspace".
831
832 10. Remove two redundant lines of code that can never be obeyed (their function
833 was moved elsewhere).
834
835 11. The program that makes PCRE's Unicode character property table had a bug
836 which caused it to generate incorrect table entries for sequences of
837 characters that have the same character type, but are in different scripts.
838 It amalgamated them into a single range, with the script of the first of
839 them. In other words, some characters were in the wrong script. There were
840 thirteen such cases, affecting characters in the following ranges:
841
842 U+002b0 - U+002c1
843 U+0060c - U+0060d
844 U+0061e - U+00612
845 U+0064b - U+0065e
846 U+0074d - U+0076d
847 U+01800 - U+01805
848 U+01d00 - U+01d77
849 U+01d9b - U+01dbf
850 U+0200b - U+0200f
851 U+030fc - U+030fe
852 U+03260 - U+0327f
853 U+0fb46 - U+0fbb1
854 U+10450 - U+1049d
855
856 12. The -o option (show only the matching part of a line) for pcregrep was not
857 compatible with GNU grep in that, if there was more than one match in a
858 line, it showed only the first of them. It now behaves in the same way as
859 GNU grep.
860
861 13. If the -o and -v options were combined for pcregrep, it printed a blank
862 line for every non-matching line. GNU grep prints nothing, and pcregrep now
863 does the same. The return code can be used to tell if there were any
864 non-matching lines.
865
866 14. Added --file-offsets and --line-offsets to pcregrep.
867
868 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
869 infinitely looping recursion. The bug was that positive lookaheads were not
870 being skipped when checking for a possible empty match (negative lookaheads
871 and both kinds of lookbehind were skipped).
872
873 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
874 inclusion of <windows.h> to before rather than after the definition of
875 INVALID_FILE_ATTRIBUTES (patch from David Byron).
876
877 17. Specifying a possessive quantifier with a specific limit for a Unicode
878 character property caused pcre_compile() to compile bad code, which led at
879 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
880 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
881 caused the error; without that there was no problem.
882
883 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
884
885 19. Added --enable-pcretest-libreadline.
886
887 20. In pcrecpp.cc, the variable 'count' was incremented twice in
888 RE::GlobalReplace(). As a result, the number of replacements returned was
889 double what it should be. I removed one of the increments, but Craig sent a
890 later patch that removed the other one (the right fix) and added unit tests
891 that check the return values (which was not done before).
892
893 21. Several CMake things:
894
895 (1) Arranged that, when cmake is used on Unix, the libraries end up with
896 the names libpcre and libpcreposix, not just pcre and pcreposix.
897
898 (2) The above change means that pcretest and pcregrep are now correctly
899 linked with the newly-built libraries, not previously installed ones.
900
901 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
902
903 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
904 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
905 UTF-8 newline character). The key issue is that the pattern starts .*;
906 this means that the match must be either at the beginning, or after a
907 newline. The bug was in the code for advancing after a failed match and
908 checking that the new position followed a newline. It was not taking
909 account of UTF-8 characters correctly.
910
911 23. PCRE was behaving differently from Perl in the way it recognized POSIX
912 character classes. PCRE was not treating the sequence [:...:] as a
913 character class unless the ... were all letters. Perl, however, seems to
914 allow any characters between [: and :], though of course it rejects as
915 unknown any "names" that contain non-letters, because all the known class
916 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
917 for example, whereas PCRE did not - it did not recognize a POSIX character
918 class. This seemed a bit dangerous, so the code has been changed to be
919 closer to Perl. The behaviour is not identical to Perl, because PCRE will
920 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
921 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
922 Perl does, and where it didn't before.
923
924 24. Rewrite so as to remove the single use of %n from pcregrep because in some
925 Windows environments %n is disabled by default.
926
927
928 Version 7.4 21-Sep-07
929 ---------------------
930
931 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
932 means that a class such as [\s] counted as "explicit reference to CR or
933 LF". That isn't really right - the whole point of the change was to try to
934 help when there was an actual mention of one of the two characters. So now
935 the change happens only if \r or \n (or a literal CR or LF) character is
936 encountered.
937
938 2. The 32-bit options word was also used for 6 internal flags, but the numbers
939 of both had grown to the point where there were only 3 bits left.
940 Fortunately, there was spare space in the data structure, and so I have
941 moved the internal flags into a new 16-bit field to free up more option
942 bits.
943
944 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
945 but did not set the internal JCHANGED flag - either of these is enough to
946 control the way the "get" function works - but the PCRE_INFO_JCHANGED
947 facility is supposed to tell if (?J) was ever used, so now (?J) at the
948 start sets both bits.
949
950 4. Added options (at build time, compile time, exec time) to change \R from
951 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
952
953 5. doc/pcresyntax.html was missing from the distribution.
954
955 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
956 compatibility, even though it is no longer used.
957
958 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
959 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
960 windows.h file is present (where different names are used). [This was
961 reversed later after testing - see 16 below.]
962
963 8. Changed all #include <config.h> to #include "config.h". There were also
964 some further <pcre.h> cases that I changed to "pcre.h".
965
966 9. When pcregrep was used with the --colour option, it missed the line ending
967 sequence off the lines that it output.
968
969 10. It was pointed out to me that arrays of string pointers cause lots of
970 relocations when a shared library is dynamically loaded. A technique of
971 using a single long string with a table of offsets can drastically reduce
972 these. I have refactored PCRE in four places to do this. The result is
973 dramatic:
974
975 Originally: 290
976 After changing UCP table: 187
977 After changing error message table: 43
978 After changing table of "verbs" 36
979 After changing table of Posix names 22
980
981 Thanks to the folks working on Gregex for glib for this insight.
982
983 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
984 unicode-properties was also set.
985
986 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
987
988 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
989 checked only for CRLF.
990
991 14. Added casts to pcretest.c to avoid compiler warnings.
992
993 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
994
995 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
996 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
997 entirely. This removes changes made in 7 above.
998
999 17. The CMake files have been updated, and there is now more information about
1000 building with CMake in the NON-UNIX-USE document.
1001
1002
1003 Version 7.3 28-Aug-07
1004 ---------------------
1005
1006 1. In the rejigging of the build system that eventually resulted in 7.1, the
1007 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1008 brackets there is not right, since it causes compilers to look for an
1009 installed pcre.h, not the version that is in the source that is being
1010 compiled (which of course may be different). I have changed it back to:
1011
1012 #include "pcre.h"
1013
1014 I have a vague recollection that the change was concerned with compiling in
1015 different directories, but in the new build system, that is taken care of
1016 by the VPATH setting the Makefile.
1017
1018 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1019 when the subject happened to end in the byte 0x85 (e.g. if the last
1020 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1021 characters but of course it shouldn't be taken as a newline when it is part
1022 of another character. The bug was that, for an unlimited repeat of . in
1023 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1024 characters when looking for a newline.
1025
1026 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1027
1028 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1029 in debug output.
1030
1031 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1032 long printing in the pcrecpp unittest when running under MinGW.
1033
1034 6. ESC_K was left out of the EBCDIC table.
1035
1036 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1037 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1038 limit also applies to "virtual nesting" when a pattern is recursive, and in
1039 this case 1000 isn't so big. I have been able to remove this limit at the
1040 expense of backing off one optimization in certain circumstances. Normally,
1041 when pcre_exec() would call its internal match() function recursively and
1042 immediately return the result unconditionally, it uses a "tail recursion"
1043 feature to save stack. However, when a subpattern that can match an empty
1044 string has an unlimited repetition quantifier, it no longer makes this
1045 optimization. That gives it a stack frame in which to save the data for
1046 checking that an empty string has been matched. Previously this was taken
1047 from the 1000-entry workspace that had been reserved. So now there is no
1048 explicit limit, but more stack is used.
1049
1050 8. Applied Daniel's patches to solve problems with the import/export magic
1051 syntax that is required for Windows, and which was going wrong for the
1052 pcreposix and pcrecpp parts of the library. These were overlooked when this
1053 problem was solved for the main library.
1054
1055 9. There were some crude static tests to avoid integer overflow when computing
1056 the size of patterns that contain repeated groups with explicit upper
1057 limits. As the maximum quantifier is 65535, the maximum group length was
1058 set at 30,000 so that the product of these two numbers did not overflow a
1059 32-bit integer. However, it turns out that people want to use groups that
1060 are longer than 30,000 bytes (though not repeat them that many times).
1061 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1062 made it possible to implement the integer overflow checks in a much more
1063 dynamic way, which I have now done. The artificial limitation on group
1064 length has been removed - we now have only the limit on the total length of
1065 the compiled pattern, which depends on the LINK_SIZE setting.
1066
1067 10. Fixed a bug in the documentation for get/copy named substring when
1068 duplicate names are permitted. If none of the named substrings are set, the
1069 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1070 empty string.
1071
1072 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1073 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1074 because the ] is interpreted as the first data character and the
1075 terminating ] is not found. PCRE has been made compatible with Perl in this
1076 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1077 cause memory overwriting.
1078
1079 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1080 string has been matched (to stop an infinite loop). It was not recognizing
1081 a conditional subpattern that could match an empty string if that
1082 subpattern was within another subpattern. For example, it looped when
1083 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1084 condition was not nested. This bug has been fixed.
1085
1086 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1087 past the start of the subject in the presence of bytes with the top bit
1088 set, for example "\x8aBCD".
1089
1090 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1091 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1092
1093 14. Optimized (?!) to (*FAIL).
1094
1095 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1096 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1097 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1098 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1099 does: it's just the validity check that is more restrictive.
1100
1101 16. Inserted checks for integer overflows during escape sequence (backslash)
1102 processing, and also fixed erroneous offset values for syntax errors during
1103 backslash processing.
1104
1105 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1106 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1107
1108 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1109 caused an overrun.
1110
1111 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1112 something other than just ASCII characters) inside a group that had an
1113 unlimited repeat caused a loop at compile time (while checking to see
1114 whether the group could match an empty string).
1115
1116 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1117 [\P{Any}] did so. (Error in the code for printing property names.)
1118
1119 21. An orphan \E inside a character class could cause a crash.
1120
1121 22. A repeated capturing bracket such as (A)? could cause a wild memory
1122 reference during compilation.
1123
1124 23. There are several functions in pcre_compile() that scan along a compiled
1125 expression for various reasons (e.g. to see if it's fixed length for look
1126 behind). There were bugs in these functions when a repeated \p or \P was
1127 present in the pattern. These operators have additional parameters compared
1128 with \d, etc, and these were not being taken into account when moving along
1129 the compiled data. Specifically:
1130
1131 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1132 length.
1133
1134 (b) An item such as \pL+ within a repeated group could cause crashes or
1135 loops.
1136
1137 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1138 "reference to non-existent subpattern" error.
1139
1140 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1141
1142 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1143 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1144
1145 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1146
1147 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1148 character were causing crashes (broken optimization).
1149
1150 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1151 \p or \P) caused a compile-time loop.
1152
1153 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1154 break. For example, the unstudied pattern [\r\n]A does not match the string
1155 "\r\nA" because change 7.0/46 below moves the current point on by two
1156 characters after failing to match at the start. However, the pattern \nA
1157 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1158 the same is true. There doesn't seem any very clean way out of this, but
1159 what I have chosen to do makes the common cases work: PCRE now takes note
1160 of whether there can be an explicit match for \r or \n anywhere in the
1161 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1162 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1163 pattern has explicit CR or LF references.
1164
1165 29. Added (*CR) etc for changing newline setting at start of pattern.
1166
1167
1168 Version 7.2 19-Jun-07
1169 ---------------------
1170
1171 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1172 which is apparently normally available under Windows.
1173
1174 2. Re-jig the pcregrep tests with different newline settings in an attempt
1175 to make them independent of the local environment's newline setting.
1176
1177 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1178
1179 4. Some of the "internals" tests were previously cut out when the link size
1180 was not 2, because the output contained actual offsets. The recent new
1181 "Z" feature of pcretest means that these can be cut out, making the tests
1182 usable with all link sizes.
1183
1184 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1185 stack recursion. This gives a massive performance boost under BSD, but just
1186 a small improvement under Linux. However, it saves one field in the frame
1187 in all cases.
1188
1189 6. Added more features from the forthcoming Perl 5.10:
1190
1191 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1192 recursion call. It refers to the nth most recently opened parentheses.
1193
1194 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1195 to be opened parentheses.
1196
1197 (c) Conditions that refer to capturing parentheses can be specified
1198 relatively, for example, (?(-2)... or (?(+3)...
1199
1200 (d) \K resets the start of the current match so that everything before
1201 is not part of it.
1202
1203 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1204
1205 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1206 reference syntax.
1207
1208 (g) (?| introduces a group in which the numbering of parentheses in each
1209 alternative starts with the same number.
1210
1211 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1212
1213 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1214 PCRE_INFO_JCHANGED.
1215
1216 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1217 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1218 for detecting groups that can match an empty string.
1219
1220 9. A pattern with a very large number of alternatives (more than several
1221 hundred) was running out of internal workspace during the pre-compile
1222 phase, where pcre_compile() figures out how much memory will be needed. A
1223 bit of new cunning has reduced the workspace needed for groups with
1224 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1225 workspace instead of running out of the 4096 that are available.
1226
1227 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1228
1229 11. Applied patch from Google to remove an optimization that didn't quite work.
1230 The report of the bug said:
1231
1232 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1233 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1234 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1235
1236 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1237 it matched the wrong number of bytes.
1238
1239
1240 Version 7.1 24-Apr-07
1241 ---------------------
1242
1243 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1244 that is more "standard", making use of automake and other Autotools. There
1245 is some re-arrangement of the files and adjustment of comments consequent
1246 on this.
1247
1248 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1249 for recursive directory scanning broke on some systems because the files
1250 are not scanned in any specific order and on different systems the order
1251 was different. A call to "sort" has been inserted into RunGrepTest for the
1252 approprate test as a short-term fix. In the longer term there may be an
1253 alternative.
1254
1255 3. I had an email from Eric Raymond about problems translating some of PCRE's
1256 man pages to HTML (despite the fact that I distribute HTML pages, some
1257 people do their own conversions for various reasons). The problems
1258 concerned the use of low-level troff macros .br and .in. I have therefore
1259 removed all such uses from the man pages (some were redundant, some could
1260 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1261 HTML has been updated to handle .nf/.fi and to complain if it encounters
1262 .br or .in.
1263
1264 4. Updated comments in configure.ac that get placed in config.h.in and also
1265 arranged for config.h to be included in the distribution, with the name
1266 config.h.generic, for the benefit of those who have to compile without
1267 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1268
1269 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1270 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1271 makevp.bat for the current PCRE, using the additional files
1272 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1273
1274 6. A Windows user reported a minor discrepancy with test 2, which turned out
1275 to be caused by a trailing space on an input line that had got lost in his
1276 copy. The trailing space was an accident, so I've just removed it.
1277
1278 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1279 that is needed.
1280
1281 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1282 as "const" (a) because they are and (b) because it helps the PHP
1283 maintainers who have recently made a script to detect big data structures
1284 in the php code that should be moved to the .rodata section. I remembered
1285 to update Builducptable as well, so it won't revert if ucptable.h is ever
1286 re-created.
1287
1288 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1289 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1290 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1291 support is not required. This saves 1.5-2K of code, which is important in
1292 some applications.
1293
1294 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1295 so as not to refer to the tables, even though these functions will never be
1296 called when UTF-8 support is disabled. Otherwise there are problems with a
1297 shared library.
1298
1299 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1300
1301 (a) It was defining its arguments as char * instead of void *.
1302
1303 (b) It was assuming that all moves were upwards in memory; this was true
1304 a long time ago when I wrote it, but is no longer the case.
1305
1306 The emulated memove() is provided for those environments that have neither
1307 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1308 is clearly not the case, as these two bugs were recently reported.
1309
1310 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1311 and Detrail to create the HTML documentation, the .txt form of the man
1312 pages, and it removes trailing spaces from listed files. It also creates
1313 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1314 case, it wraps all the #defines with #ifndefs. This script should be run
1315 before "make dist".
1316
1317 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1318 with Unicode property support.
1319
1320 (a) For a maximizing quantifier, if the two different cases of the
1321 character were of different lengths in their UTF-8 codings (there are
1322 some cases like this - I found 11), and the matching function had to
1323 back up over a mixture of the two cases, it incorrectly assumed they
1324 were both the same length.
1325
1326 (b) When PCRE was configured to use the heap rather than the stack for
1327 recursion during matching, it was not correctly preserving the data for
1328 the other case of a UTF-8 character when checking ahead for a match
1329 while processing a minimizing repeat. If the check also involved
1330 matching a wide character, but failed, corruption could cause an
1331 erroneous result when trying to check for a repeat of the original
1332 character.
1333
1334 13. Some tidying changes to the testing mechanism:
1335
1336 (a) The RunTest script now detects the internal link size and whether there
1337 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1338 values substituted by "configure". (The RunGrepTest script already did
1339 this for UTF-8.) The configure.ac script no longer substitutes the
1340 relevant variables.
1341
1342 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1343 with length and offset values. This means that the output is different
1344 for different internal link sizes. Test 2 is skipped for link sizes
1345 other than 2 because of this, bypassing the problem. Unfortunately,
1346 there was also a test in test 3 (the locale tests) that used /B and
1347 failed for link sizes other than 2. Rather than cut the whole test out,
1348 I have added a new /Z option to pcretest that replaces the length and
1349 offset values with spaces. This is now used to make test 3 independent
1350 of link size. (Test 2 will be tidied up later.)
1351
1352 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1353 segmentation fault instead of returning the appropriate error message.
1354
1355 15. In multiline mode when the newline sequence was set to "any", the pattern
1356 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1357 This doesn't seem right; it now treats the CRLF combination as the line
1358 ending, and so does not match in that case. It's only a pattern such as ^$
1359 that would hit this one: something like ^ABC$ would have failed after \r
1360 and then tried again after \r\n.
1361
1362 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1363 in an attempt to make files that differ only in their line terminators
1364 compare equal. This works on Linux.
1365
1366 17. Under certain error circumstances pcregrep might try to free random memory
1367 as it exited. This is now fixed, thanks to valgrind.
1368
1369 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1370 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1371 was because its rules for how to advance for /g after matching an empty
1372 string at the end of a line did not allow for this case. They now check for
1373 it specially.
1374
1375 20. pcretest is supposed to handle patterns and data of any length, by
1376 extending its buffers when necessary. It was getting this wrong when the
1377 buffer for a data line had to be extended.
1378
1379 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1380 CRLF as a newline sequence.
1381
1382 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1383 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1384 I have nevertheless tidied it up.
1385
1386 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1387
1388 24. Added a man page for pcre-config.
1389
1390
1391 Version 7.0 19-Dec-06
1392 ---------------------
1393
1394 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1395 moving to gcc 4.1.1.
1396
1397 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1398 sys/time.h, which is documented as needed for this function. It doesn't
1399 seem to matter on Linux, but it showed up on some releases of OS X.
1400
1401 3. It seems that there are systems where bytes whose values are greater than
1402 127 match isprint() in the "C" locale. The "C" locale should be the
1403 default when a C program starts up. In most systems, only ASCII printing
1404 characters match isprint(). This difference caused the output from pcretest
1405 to vary, making some of the tests fail. I have changed pcretest so that:
1406
1407 (a) When it is outputting text in the compiled version of a pattern, bytes
1408 other than 32-126 are always shown as hex escapes.
1409
1410 (b) When it is outputting text that is a matched part of a subject string,
1411 it does the same, unless a different locale has been set for the match
1412 (using the /L modifier). In this case, it uses isprint() to decide.
1413
1414 4. Fixed a major bug that caused incorrect computation of the amount of memory
1415 required for a compiled pattern when options that changed within the
1416 pattern affected the logic of the preliminary scan that determines the
1417 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1418 that the computed length was too small. The symptoms of this bug were
1419 either the PCRE error "internal error: code overflow" from pcre_compile(),
1420 or a glibc crash with a message such as "pcretest: free(): invalid next
1421 size (fast)". Examples of patterns that provoked this bug (shown in
1422 pcretest format) are:
1423
1424 /(?-x: )/x
1425 /(?x)(?-x: \s*#\s*)/
1426 /((?i)[\x{c0}])/8
1427 /(?i:[\x{c0}])/8
1428
1429 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1430 is now done differently.
1431
1432 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1433 wrapper classes; (b) implement a new function in the C++ scanner that is
1434 more efficient than the old way of doing things because it avoids levels of
1435 recursion in the regex matching; (c) add a paragraph to the documentation
1436 for the FullMatch() function.
1437
1438 6. The escape sequence \n was being treated as whatever was defined as
1439 "newline". Not only was this contrary to the documentation, which states
1440 that \n is character 10 (hex 0A), but it also went horribly wrong when
1441 "newline" was defined as CRLF. This has been fixed.
1442
1443 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1444 was being set to -1 for the "end of line" case (supposedly a value that no
1445 character can have). Though this value is never used (the check for end of
1446 line is "zero bytes in current character"), it caused compiler complaints.
1447 I've changed it to 0xffffffff.
1448
1449 8. In pcre_version.c, the version string was being built by a sequence of
1450 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1451 string (as it is for production releases) called a macro with an empty
1452 argument. The C standard says the result of this is undefined. The gcc
1453 compiler treats it as an empty string (which was what was wanted) but it is
1454 reported that Visual C gives an error. The source has been hacked around to
1455 avoid this problem.
1456
1457 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1458 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1459 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1460 of them did).
1461
1462 10. Originally, pcretest opened its input and output without "b"; then I was
1463 told that "b" was needed in some environments, so it was added for release
1464 5.0 to both the input and output. (It makes no difference on Unix-like
1465 systems.) Later I was told that it is wrong for the input on Windows. I've
1466 now abstracted the modes into two macros, to make it easier to fiddle with
1467 them, and removed "b" from the input mode under Windows.
1468
1469 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1470
1471 12. Added -help and --help to pcretest as an official way of being reminded
1472 of the options.
1473
1474 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1475 and pcrecpp.cc because they annoy compilers at high warning levels.
1476
1477 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1478
1479 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1480 scripts are not C programs :-) and which was not noticed because it works
1481 on Linux.
1482
1483 16. pcretest is supposed to handle any length of pattern and data line (as one
1484 line or as a continued sequence of lines) by extending its input buffer if
1485 necessary. This feature was broken for very long pattern lines, leading to
1486 a string of junk being passed to pcre_compile() if the pattern was longer
1487 than about 50K.
1488
1489 17. I have done a major re-factoring of the way pcre_compile() computes the
1490 amount of memory needed for a compiled pattern. Previously, there was code
1491 that made a preliminary scan of the pattern in order to do this. That was
1492 OK when PCRE was new, but as the facilities have expanded, it has become
1493 harder and harder to keep it in step with the real compile phase, and there
1494 have been a number of bugs (see for example, 4 above). I have now found a
1495 cunning way of running the real compile function in a "fake" mode that
1496 enables it to compute how much memory it would need, while actually only
1497 ever using a few hundred bytes of working memory and without too many
1498 tests of the mode. This should make future maintenance and development
1499 easier. A side effect of this work is that the limit of 200 on the nesting
1500 depth of parentheses has been removed (though this was never a serious
1501 limitation, I suspect). However, there is a downside: pcre_compile() now
1502 runs more slowly than before (30% or more, depending on the pattern). I
1503 hope this isn't a big issue. There is no effect on runtime performance.
1504
1505 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1506 newline (only possible for the last line of a file) and it was a
1507 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1508
1509 19. Added additional timing features to pcretest. (1) The -tm option now times
1510 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1511 separate command line item, by a number that specifies the number of
1512 repeats to use when timing. The default is 50000; this gives better
1513 precision, but takes uncomfortably long for very large patterns.
1514
1515 20. Extended pcre_study() to be more clever in cases where a branch of a
1516 subpattern has no definite first character. For example, (a*|b*)[cd] would
1517 previously give no result from pcre_study(). Now it recognizes that the
1518 first character must be a, b, c, or d.
1519
1520 21. There was an incorrect error "recursive call could loop indefinitely" if
1521 a subpattern (or the entire pattern) that was being tested for matching an
1522 empty string contained only one non-empty item after a nested subpattern.
1523 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1524 incorrectly, because the \d was being skipped in the check.
1525
1526 22. The pcretest program now has a new pattern option /B and a command line
1527 option -b, which is equivalent to adding /B to every pattern. This causes
1528 it to show the compiled bytecode, without the additional information that
1529 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1530 is the same as /B/I).
1531
1532 23. A new optimization is now able automatically to treat some sequences such
1533 as a*b as a*+b. More specifically, if something simple (such as a character
1534 or a simple class like \d) has an unlimited quantifier, and is followed by
1535 something that cannot possibly match the quantified thing, the quantifier
1536 is automatically "possessified".
1537
1538 24. A recursive reference to a subpattern whose number was greater than 39
1539 went wrong under certain circumstances in UTF-8 mode. This bug could also
1540 have affected the operation of pcre_study().
1541
1542 25. Realized that a little bit of performance could be had by replacing
1543 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1544
1545 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1546
1547 27. Possessive quantifiers such as a++ were previously implemented by turning
1548 them into atomic groups such as ($>a+). Now they have their own opcodes,
1549 which improves performance. This includes the automatically created ones
1550 from 23 above.
1551
1552 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1553 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1554 the first matched character to be a colon. This applied both to named and
1555 numbered groups.
1556
1557 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1558
1559 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1560 building PCRE on Windows easier, so I have included it in the distribution.
1561
1562 31. There is now a check in pcretest against a ridiculously large number being
1563 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1564 loop, the loop is abandoned.
1565
1566 32. Forward references to subpatterns in conditions such as (?(2)...) where
1567 subpattern 2 is defined later cause pcre_compile() to search forwards in
1568 the pattern for the relevant set of parentheses. This search went wrong
1569 when there were unescaped parentheses in a character class, parentheses
1570 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1571
1572 33. "Subroutine" calls and backreferences were previously restricted to
1573 referencing subpatterns earlier in the regex. This restriction has now
1574 been removed.
1575
1576 34. Added a number of extra features that are going to be in Perl 5.10. On the
1577 whole, these are just syntactic alternatives for features that PCRE had
1578 previously implemented using the Python syntax or my own invention. The
1579 other formats are all retained for compatibility.
1580
1581 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1582 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1583 also .NET compatible.
1584
1585 (b) A recursion or subroutine call to a named group can now be defined as
1586 (?&name) as well as (?P>name).
1587
1588 (c) A backreference to a named group can now be defined as \k<name> or
1589 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1590 5.10, are also .NET compatible.
1591
1592 (d) A conditional reference to a named group can now use the syntax
1593 (?(<name>) or (?('name') as well as (?(name).
1594
1595 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1596 groups (named and numbered) that are never evaluated inline, but can be
1597 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1598 is always false. There may be only one alternative in such a group.
1599
1600 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1601 as the simple (?(R). The condition is true only if the most recent
1602 recursion is that of the given number or name. It does not search out
1603 through the entire recursion stack.
1604
1605 (g) The escape \gN or \g{N} has been added, where N is a positive or
1606 negative number, specifying an absolute or relative reference.
1607
1608 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1609 some "unreachable code" warnings.
1610
1611 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1612 things, this adds five new scripts.
1613
1614 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1615 There were also incompatibilities regarding the handling of \Q..\E inside
1616 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1617 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1618
1619 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1620 matches an empty string, and forcibly breaks the loop. There were bugs in
1621 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1622 against aaaa the result was just "a" rather than "aaaa", for example. Two
1623 separate and independent bugs (that affected different cases) have been
1624 fixed.
1625
1626 39. Refactored the code to abolish the use of different opcodes for small
1627 capturing bracket numbers. This is a tidy that I avoided doing when I
1628 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1629 The new approach is not only tidier, it makes it possible to reduce the
1630 memory needed to fix the previous bug (38).
1631
1632 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1633 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1634 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1635 mode.
1636
1637 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1638 report.
1639
1640 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1641 copy construction and assignment in the C++ wrapper.
1642
1643 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1644 couple of bugs that could have given wrong results in the "--newline=crlf"
1645 case.
1646
1647 44. Added a number of casts and did some reorganization of signed/unsigned int
1648 variables following suggestions from Dair Grant. Also renamed the variable
1649 "this" as "item" because it is a C++ keyword.
1650
1651 45. Arranged for dftables to add
1652
1653 #include "pcre_internal.h"
1654
1655 to pcre_chartables.c because without it, gcc 4.x may remove the array
1656 definition from the final binary if PCRE is built into a static library and
1657 dead code stripping is activated.
1658
1659 46. For an unanchored pattern, if a match attempt fails at the start of a
1660 newline sequence, and the newline setting is CRLF or ANY, and the next two
1661 characters are CRLF, advance by two characters instead of one.
1662
1663
1664 Version 6.7 04-Jul-06
1665 ---------------------
1666
1667 1. In order to handle tests when input lines are enormously long, pcretest has
1668 been re-factored so that it automatically extends its buffers when
1669 necessary. The code is crude, but this _is_ just a test program. The
1670 default size has been increased from 32K to 50K.
1671
1672 2. The code in pcre_study() was using the value of the re argument before
1673 testing it for NULL. (Of course, in any sensible call of the function, it
1674 won't be NULL.)
1675
1676 3. The memmove() emulation function in pcre_internal.h, which is used on
1677 systems that lack both memmove() and bcopy() - that is, hardly ever -
1678 was missing a "static" storage class specifier.
1679
1680 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1681 containing an extended class (one that cannot be represented by a bitmap
1682 because it contains high-valued characters or Unicode property items, e.g.
1683 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1684 pattern, but PCRE should not loop if you do not (it no longer does).
1685 [Detail: two cases were found: (a) a repeated subpattern containing an
1686 extended class; (b) a recursive reference to a subpattern that followed a
1687 previous extended class. It wasn't skipping over the extended class
1688 correctly when UTF-8 mode was not set.]
1689
1690 5. A negated single-character class was not being recognized as fixed-length
1691 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1692 compile error "lookbehind assertion is not fixed length".
1693
1694 6. The RunPerlTest auxiliary script was showing an unexpected difference
1695 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1696 write a Perl script that can interpret lines of an input file either as
1697 byte characters or as UTF-8, which is what "perltest" was being required to
1698 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1699 can't do is switch easily at run time between having the "use utf8;" pragma
1700 or not. In the end, I fudged it by using the RunPerlTest script to insert
1701 "use utf8;" explicitly for the UTF-8 tests.
1702
1703 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1704 the end of the subject string, contrary to the documentation and to what
1705 Perl does. This was true of both matching functions. Now it matches only at
1706 the start of the subject and immediately after *internal* newlines.
1707
1708 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1709 a pointer to an int instead of a pointer to an unsigned long int. This
1710 caused problems on 64-bit systems.
1711
1712 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1713 instance of the 'standard' template library not being so standard".
1714
1715 10. There was no check on the number of named subpatterns nor the maximum
1716 length of a subpattern name. The product of these values is used to compute
1717 the size of the memory block for a compiled pattern. By supplying a very
1718 long subpattern name and a large number of named subpatterns, the size
1719 computation could be caused to overflow. This is now prevented by limiting
1720 the length of names to 32 characters, and the number of named subpatterns
1721 to 10,000.
1722
1723 11. Subpatterns that are repeated with specific counts have to be replicated in
1724 the compiled pattern. The size of memory for this was computed from the
1725 length of the subpattern and the repeat count. The latter is limited to
1726 65535, but there was no limit on the former, meaning that integer overflow
1727 could in principle occur. The compiled length of a repeated subpattern is
1728 now limited to 30,000 bytes in order to prevent this.
1729
1730 12. Added the optional facility to have named substrings with the same name.
1731
1732 13. Added the ability to use a named substring as a condition, using the
1733 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1734 are numbers (not recommended). Forward references are permitted.
1735
1736 14. Added forward references in named backreferences (if you see what I mean).
1737
1738 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1739 pattern could run off the end of the subject. For example, the pattern
1740 "(?s)(.{1,5})"8 did this with the subject "ab".
1741
1742 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1743 PCRE_CASELESS was set when matching characters that were quantified with ?
1744 or *.
1745
1746 17. A character class other than a single negated character that had a minimum
1747 but no maximum quantifier - for example [ab]{6,} - was not handled
1748 correctly by pce_dfa_exec(). It would match only one character.
1749
1750 18. A valid (though odd) pattern that looked like a POSIX character
1751 class but used an invalid character after [ (for example [[,abc,]]) caused
1752 pcre_compile() to give the error "Failed: internal error: code overflow" or
1753 in some cases to crash with a glibc free() error. This could even happen if
1754 the pattern terminated after [[ but there just happened to be a sequence of
1755 letters, a binary zero, and a closing ] in the memory that followed.
1756
1757 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1758 over the years. Originally (before any Unicode support), just the bottom 8
1759 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1760 output from "man perlunicode" includes this:
1761
1762 The regular expression compiler produces polymorphic opcodes. That
1763 is, the pattern adapts to the data and automatically switches to
1764 the Unicode character scheme when presented with Unicode data--or
1765 instead uses a traditional byte scheme when presented with byte
1766 data.
1767
1768 Sadly, a wide octal escape does not cause a switch, and in a string with
1769 no other multibyte characters, these octal escapes are treated as before.
1770 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1771 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1772 Unicode string.
1773
1774 I have not perpetrated such confusion in PCRE. Up till now, it took just
1775 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1776 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1777 translate to the appropriate multibyte character.
1778
1779 29. Applied some refactoring to reduce the number of warnings from Microsoft
1780 and Borland compilers. This has included removing the fudge introduced
1781 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1782 a warning about an unused variable.
1783
1784 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1785 characters since release 4.0, because Perl (from release 5.004) does not.
1786 [Or at least, is documented not to: some releases seem to be in conflict
1787 with the documentation.] However, when a pattern was studied with
1788 pcre_study() and all its branches started with \s, PCRE still included VT
1789 as a possible starting character. Of course, this did no harm; it just
1790 caused an unnecessary match attempt.
1791
1792 22. Removed a now-redundant internal flag bit that recorded the fact that case
1793 dependency changed within the pattern. This was once needed for "required
1794 byte" processing, but is no longer used. This recovers a now-scarce options
1795 bit. Also moved the least significant internal flag bit to the most-
1796 significant bit of the word, which was not previously used (hangover from
1797 the days when it was an int rather than a uint) to free up another bit for
1798 the future.
1799
1800 23. Added support for CRLF line endings as well as CR and LF. As well as the
1801 default being selectable at build time, it can now be changed at runtime
1802 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1803 specify that it is scanning data with non-default line endings.
1804
1805 24. Changed the definition of CXXLINK to make it agree with the definition of
1806 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1807
1808 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1809 recursions. This makes a big different to stack usage for some patterns.
1810
1811 26. If a subpattern containing a named recursion or subroutine reference such
1812 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1813 the space required for the compiled pattern went wrong and gave too small a
1814 value. Depending on the environment, this could lead to "Failed: internal
1815 error: code overflow at offset 49" or "glibc detected double free or
1816 corruption" errors.
1817
1818 27. Applied patches from Google (a) to support the new newline modes and (b) to
1819 advance over multibyte UTF-8 characters in GlobalReplace.
1820
1821 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1822 difference for some implementation of PCRE in some Windows version.
1823
1824 29. Added some extra testing facilities to pcretest:
1825
1826 \q<number> in a data line sets the "match limit" value
1827 \Q<number> in a data line sets the "match recursion limt" value
1828 -S <number> sets the stack size, where <number> is in megabytes
1829
1830 The -S option isn't available for Windows.
1831
1832
1833 Version 6.6 06-Feb-06
1834 ---------------------
1835
1836 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1837 in pcreposix.h. I have copied the definition from pcre.h.
1838
1839 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1840 because pcre.h is no longer a built file.
1841
1842 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1843 not normally included in the compiled code.
1844
1845
1846 Version 6.5 01-Feb-06
1847 ---------------------
1848
1849 1. When using the partial match feature with pcre_dfa_exec(), it was not
1850 anchoring the second and subsequent partial matches at the new starting
1851 point. This could lead to incorrect results. For example, with the pattern
1852 /1234/, partially matching against "123" and then "a4" gave a match.
1853
1854 2. Changes to pcregrep:
1855
1856 (a) All non-match returns from pcre_exec() were being treated as failures
1857 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1858 error message is output. Some extra information is given for the
1859 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1860 probably the only errors that are likely to be caused by users (by
1861 specifying a regex that has nested indefinite repeats, for instance).
1862 If there are more than 20 of these errors, pcregrep is abandoned.
1863
1864 (b) A binary zero was treated as data while matching, but terminated the
1865 output line if it was written out. This has been fixed: binary zeroes
1866 are now no different to any other data bytes.
1867
1868 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1869 used to set a locale for matching. The --locale=xxxx long option has
1870 been added (no short equivalent) to specify a locale explicitly on the
1871 pcregrep command, overriding the environment variables.
1872
1873 (d) When -B was used with -n, some line numbers in the output were one less
1874 than they should have been.
1875
1876 (e) Added the -o (--only-matching) option.
1877
1878 (f) If -A or -C was used with -c (count only), some lines of context were
1879 accidentally printed for the final match.
1880
1881 (g) Added the -H (--with-filename) option.
1882
1883 (h) The combination of options -rh failed to suppress file names for files
1884 that were found from directory arguments.
1885
1886 (i) Added the -D (--devices) and -d (--directories) options.
1887
1888 (j) Added the -F (--fixed-strings) option.
1889
1890 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1891
1892 (l) Added the --colo(u)r option.
1893
1894 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1895 is not present by default.
1896
1897 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1898 items such as (?R) or (?1), when the recursion could match a number of
1899 alternatives. If it matched one of the alternatives, but subsequently,
1900 outside the recursion, there was a failure, the code tried to back up into
1901 the recursion. However, because of the way PCRE is implemented, this is not
1902 possible, and the result was an incorrect result from the match.
1903
1904 In order to prevent this happening, the specification of recursion has
1905 been changed so that all such subpatterns are automatically treated as
1906 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1907
1908 4. I had overlooked the fact that, in some locales, there are characters for
1909 which isalpha() is true but neither isupper() nor islower() are true. In
1910 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1911 and ordfeminine) are like this. This affected the treatment of \w and \W
1912 when they appeared in character classes, but not when they appeared outside
1913 a character class. The bit map for "word" characters is now created
1914 separately from the results of isalnum() instead of just taking it from the
1915 upper, lower, and digit maps. (Plus the underscore character, of course.)
1916
1917 5. The above bug also affected the handling of POSIX character classes such as
1918 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1919 permanent tables. Instead, the bit maps for such a class were previously
1920 created as the appropriate unions of the upper, lower, and digit bitmaps.
1921 Now they are created by subtraction from the [[:word:]] class, which has
1922 its own bitmap.
1923
1924 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1925 It is created by subtracting the vertical space characters (\x09, \x0a,
1926 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1927 subtraction was done in the overall bitmap for a character class, meaning
1928 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1929 be recognized. This bug has been fixed.
1930
1931 7. Patches from the folks at Google:
1932
1933 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1934 real life, but is still worth protecting against".
1935
1936 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1937 regular expressions".
1938
1939 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1940 have it.
1941
1942 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1943 "configure" and the latter not, in order to fix a problem somebody had
1944 with compiling the Arg class on HP-UX.
1945
1946 (e) Improve the error-handling of the C++ wrapper a little bit.
1947
1948 (f) New tests for checking recursion limiting.
1949
1950 8. The pcre_memmove() function, which is used only if the environment does not
1951 have a standard memmove() function (and is therefore rarely compiled),
1952 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1953 returning a result (though PCRE never actually uses the result).
1954
1955 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1956 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1957 returned instead of calling malloc() with an overflowing number that would
1958 most likely cause subsequent chaos.
1959
1960 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1961
1962 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1963 with this option is matched, the nmatch and pmatch options of regexec() are
1964 ignored.
1965
1966 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1967 provided in case anyone wants to the the POSIX interface with UTF-8
1968 strings.
1969
1970 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1971 C++ linking (needed for some HP-UX environments).
1972
1973 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1974 (unused parameter) and in the pcre_printint() function (omitted "default"
1975 switch label when the default is to do nothing).
1976
1977 15. Added some code to make it possible, when PCRE is compiled as a C++
1978 library, to replace subject pointers for pcre_exec() with a smart pointer
1979 class, thus making it possible to process discontinuous strings.
1980
1981 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1982 much the same function. They were added by different people who were trying
1983 to make PCRE easy to compile on non-Unix systems. It has been suggested
1984 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1985 for compiling on Windows systems. I have therefore replaced it with
1986 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1987 defaults to "extern" for C or "extern C" for C++, which works fine on
1988 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1989 SCOPE with something explicit in config.h. In addition:
1990
1991 (a) pcreposix.h still had just "extern" instead of either of these macros;
1992 I have replaced it with PCRE_DATA_SCOPE.
1993
1994 (b) Functions such as _pcre_xclass(), which are internal to the library,
1995 but external in the C sense, all had PCRE_EXPORT in their definitions.
1996 This is apparently wrong for the Windows case, so I have removed it.
1997 (It makes no difference on Unix-like systems.)
1998
1999 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2000 of recursive calls to match(). This is different to MATCH_LIMIT because
2001 that limits the total number of calls to match(), not all of which increase
2002 the depth of recursion. Limiting the recursion depth limits the amount of
2003 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2004 when PCRE is compiled, and changed at run time. A patch from Google adds
2005 this functionality to the C++ interface.
2006
2007 18. Changes to the handling of Unicode character properties:
2008
2009 (a) Updated the table to Unicode 4.1.0.
2010
2011 (b) Recognize characters that are not in the table as "Cn" (undefined).
2012
2013 (c) I revised the way the table is implemented to a much improved format
2014 which includes recognition of ranges. It now supports the ranges that
2015 are defined in UnicodeData.txt, and it also amalgamates other
2016 characters into ranges. This has reduced the number of entries in the
2017 table from around 16,000 to around 3,000, thus reducing its size
2018 considerably. I realized I did not need to use a tree structure after
2019 all - a binary chop search is just as efficient. Having reduced the
2020 number of entries, I extended their size from 6 bytes to 8 bytes to
2021 allow for more data.
2022
2023 (d) Added support for Unicode script names via properties such as \p{Han}.
2024
2025 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2026 matching that character.
2027
2028 20. When matching a repeated Unicode property with a minimum greater than zero,
2029 (for example \pL{2,}), PCRE could look past the end of the subject if it
2030 reached it while seeking the minimum number of characters. This could
2031 happen only if some of the characters were more than one byte long, because
2032 there is a check for at least the minimum number of bytes.
2033
2034 21. Refactored the implementation of \p and \P so as to be more general, to
2035 allow for more different types of property in future. This has changed the
2036 compiled form incompatibly. Anybody with saved compiled patterns that use
2037 \p or \P will have to recompile them.
2038
2039 22. Added "Any" and "L&" to the supported property types.
2040
2041 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2042 but give a compile time error if the value is greater than 0xff.
2043
2044 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2045 accidentally not being installed or uninstalled.
2046
2047 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2048 made were to insert the current release number. This seemed silly, because
2049 it made things harder for people building PCRE on systems that don't run
2050 "configure". I have turned pcre.h into a distributed file, no longer built
2051 by "configure", with the version identification directly included. There is
2052 no longer a pcre.h.in file.
2053
2054 However, this change necessitated a change to the pcre-config script as
2055 well. It is built from pcre-config.in, and one of the substitutions was the
2056 release number. I have updated configure.ac so that ./configure now finds
2057 the release number by grepping pcre.h.
2058
2059 26. Added the ability to run the tests under valgrind.
2060
2061
2062 Version 6.4 05-Sep-05
2063 ---------------------
2064
2065 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2066 "--" to be printed when multiple files were scanned, even when none of the
2067 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2068 consider it to be a bug, and have restored the previous behaviour.
2069
2070 2. A couple of code tidies to get rid of compiler warnings.
2071
2072 3. The pcretest program used to cheat by referring to symbols in the library
2073 whose names begin with _pcre_. These are internal symbols that are not
2074 really supposed to be visible externally, and in some environments it is
2075 possible to suppress them. The cheating is now confined to including
2076 certain files from the library's source, which is a bit cleaner.
2077
2078 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2079 file's purpose clearer.
2080
2081 5. Reorganized pcre_ucp_findchar().
2082
2083
2084 Version 6.3 15-Aug-05
2085 ---------------------
2086
2087 1. The file libpcre.pc.in did not have general read permission in the tarball.
2088
2089 2. There were some problems when building without C++ support:
2090
2091 (a) If C++ support was not built, "make install" and "make test" still
2092 tried to test it.
2093
2094 (b) There were problems when the value of CXX was explicitly set. Some
2095 changes have been made to try to fix these, and ...
2096
2097 (c) --disable-cpp can now be used to explicitly disable C++ support.
2098
2099 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2100 backslash in a target when C++ was disabled. This confuses some
2101 versions of "make", apparently. Using an intermediate variable solves
2102 this. (Same for CPP_LOBJ.)
2103
2104 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2105 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2106 necessary on certain architectures.
2107
2108 4. Added a setting of -export-symbols-regex to the link command to remove
2109 those symbols that are exported in the C sense, but actually are local
2110 within the library, and not documented. Their names all begin with
2111 "_pcre_". This is not a perfect job, because (a) we have to except some
2112 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2113 available (and never for static libraries). I have made a note to try to
2114 find a way round (a) in the future.
2115
2116
2117 Version 6.2 01-Aug-05
2118 ---------------------
2119
2120 1. There was no test for integer overflow of quantifier values. A construction
2121 such as {1111111111111111} would give undefined results. What is worse, if
2122 a minimum quantifier for a parenthesized subpattern overflowed and became
2123 negative, the calculation of the memory size went wrong. This could have
2124 led to memory overwriting.
2125
2126 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2127
2128 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2129 operating environments where this matters.
2130
2131 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2132 PCRE options from within the C++ wrapper.
2133
2134 5. Named capturing subpatterns were not being correctly counted when a pattern
2135 was compiled. This caused two problems: (a) If there were more than 100
2136 such subpatterns, the calculation of the memory needed for the whole
2137 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2138 back references of the form \12, where the number was greater than 9, were
2139 not recognized as back references, even though there were sufficient
2140 previous subpatterns.
2141
2142 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2143 versions of gcc, e.g. 2.95.4.
2144
2145
2146 Version 6.1 21-Jun-05
2147 ---------------------
2148
2149 1. There was one reference to the variable "posix" in pcretest.c that was not
2150 surrounded by "#if !defined NOPOSIX".
2151
2152 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2153 the cross-check on the old pcre_info() function, for the benefit of the
2154 cut-down version of PCRE that is currently imported into Exim.
2155
2156 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2157 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2158 patterns that start (?i) but I don't think that matters. The use of (?i) is
2159 just an example; this all applies to the other options as well.
2160
2161 4. Since libtool seems to echo the compile commands it is issuing, the output
2162 from "make" can be reduced a bit by putting "@" in front of each libtool
2163 compile command.
2164
2165 5. Patch from the folks at Google for configure.in to be a bit more thorough
2166 in checking for a suitable C++ installation before trying to compile the
2167 C++ stuff. This should fix a reported problem when a compiler was present,
2168 but no suitable headers.
2169
2170 6. The man pages all had just "PCRE" as their title. I have changed them to
2171 be the relevant file name. I have also arranged that these names are
2172 retained in the file doc/pcre.txt, which is a concatenation in text format
2173 of all the man pages except the little individual ones for each function.
2174
2175 7. The NON-UNIX-USE file had not been updated for the different set of source
2176 files that come with release 6. I also added a few comments about the C++
2177 wrapper.
2178
2179
2180 Version 6.0 07-Jun-05
2181 ---------------------
2182
2183 1. Some minor internal re-organization to help with my DFA experiments.
2184
2185 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2186 didn't matter for the library itself when fully configured, but did matter
2187 when compiling without UCP support, or within Exim, where the ucp files are
2188 not imported.
2189
2190 3. Refactoring of the library code to split up the various functions into
2191 different source modules. The addition of the new DFA matching code (see
2192 below) to a single monolithic source would have made it really too
2193 unwieldy, quite apart from causing all the code to be include in a
2194 statically linked application, when only some functions are used. This is
2195 relevant even without the DFA addition now that patterns can be compiled in
2196 one application and matched in another.
2197
2198 The downside of splitting up is that there have to be some external
2199 functions and data tables that are used internally in different modules of
2200 the library but which are not part of the API. These have all had their
2201 names changed to start with "_pcre_" so that they are unlikely to clash
2202 with other external names.
2203
2204 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2205 a different (DFA) algorithm. Although it is slower than the original
2206 function, it does have some advantages for certain types of matching
2207 problem.
2208
2209 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2210 including restarting after a partial match.
2211
2212 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2213 defined when compiling for Windows was sent to me. I have put it into the
2214 code, though I have no means of testing or verifying it.
2215
2216 7. Added the pcre_refcount() auxiliary function.
2217
2218 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2219 match before or at the first newline in the subject string. In pcretest,
2220 the /f option on a pattern can be used to set this.
2221
2222 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2223 would behave wrongly. This has been present in PCRE since release 4.0.
2224
2225 10. A number of changes to the pcregrep command:
2226
2227 (a) Refactored how -x works; insert ^(...)$ instead of setting
2228 PCRE_ANCHORED and checking the length, in preparation for adding
2229 something similar for -w.
2230
2231 (b) Added the -w (match as a word) option.
2232
2233 (c) Refactored the way lines are read and buffered so as to have more
2234 than one at a time available.
2235
2236 (d) Implemented a pcregrep test script.
2237
2238 (e) Added the -M (multiline match) option. This allows patterns to match
2239 over several lines of the subject. The buffering ensures that at least
2240 8K, or the rest of the document (whichever is the shorter) is available
2241 for matching (and similarly the previous 8K for lookbehind assertions).
2242
2243 (f) Changed the --help output so that it now says
2244
2245 -w, --word-regex(p)
2246
2247 instead of two lines, one with "regex" and the other with "regexp"
2248 because that confused at least one person since the short forms are the
2249 same. (This required a bit of code, as the output is generated
2250 automatically from a table. It wasn't just a text change.)
2251
2252 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2253 option but starts with a hyphen. Could be a pattern or a path name
2254 starting with a hyphen, for instance.
2255
2256 (h) "-" can be given as a file name to represent stdin.
2257
2258 (i) When file names are being printed, "(standard input)" is used for
2259 the standard input, for compatibility with GNU grep. Previously
2260 "<stdin>" was used.
2261
2262 (j) The option --label=xxx can be used to supply a name to be used for
2263 stdin when file names are being printed. There is no short form.
2264
2265 (k) Re-factored the options decoding logic because we are going to add
2266 two more options that take data. Such options can now be given in four
2267 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2268
2269 (l) Added the -A, -B, and -C options for requesting that lines of context
2270 around matches be printed.
2271
2272 (m) Added the -L option to print the names of files that do not contain
2273 any matching lines, that is, the complement of -l.
2274
2275 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2276 continue to scan other files.
2277
2278 (o) The -s option was incorrectly implemented. For compatibility with other
2279 greps, it now suppresses the error message for a non-existent or non-
2280 accessible file (but not the return code). There is a new option called
2281 -q that suppresses the output of matching lines, which was what -s was
2282 previously doing.
2283
2284 (p) Added --include and --exclude options to specify files for inclusion
2285 and exclusion when recursing.
2286
2287 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2288 Hopefully, it now does.
2289
2290 12. Missing cast in pcre_study().
2291
2292 13. Added an "uninstall" target to the makefile.
2293
2294 14. Replaced "extern" in the function prototypes in Makefile.in with
2295 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2296 world, but is set differently for Windows.
2297
2298 15. Added a second compiling function called pcre_compile2(). The only
2299 difference is that it has an extra argument, which is a pointer to an
2300 integer error code. When there is a compile-time failure, this is set
2301 non-zero, in addition to the error test pointer being set to point to an
2302 error message. The new argument may be NULL if no error number is required
2303 (but then you may as well call pcre_compile(), which is now just a
2304 wrapper). This facility is provided because some applications need a
2305 numeric error indication, but it has also enabled me to tidy up the way
2306 compile-time errors are handled in the POSIX wrapper.
2307
2308 16. Added VPATH=.libs to the makefile; this should help when building with one
2309 prefix path and installing with another. (Or so I'm told by someone who
2310 knows more about this stuff than I do.)
2311
2312 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2313 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2314 match everything, including newlines. This is not POSIX-compatible, but
2315 somebody wanted the feature. From pcretest it can be activated by using
2316 both the P and the s flags.
2317
2318 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2319
2320 19. libpcre.pc was being incorrectly installed as executable.
2321
2322 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2323 it now also looks for '\r' so that it will work unmodified on Windows.
2324
2325 21. Added Google's contributed C++ wrapper to the distribution.
2326
2327 22. Added some untidy missing memory free() calls in pcretest, to keep
2328 Electric Fence happy when testing.
2329
2330
2331
2332 Version 5.0 13-Sep-04
2333 ---------------------
2334
2335 1. Internal change: literal characters are no longer packed up into items
2336 containing multiple characters in a single byte-string. Each character
2337 is now matched using a separate opcode. However, there may be more than one
2338 byte in the character in UTF-8 mode.
2339
2340 2. The pcre_callout_block structure has two new fields: pattern_position and
2341 next_item_length. These contain the offset in the pattern to the next match
2342 item, and its length, respectively.
2343
2344 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2345 insertion of callouts before each pattern item. Added the /C option to
2346 pcretest to make use of this.
2347
2348 4. On the advice of a Windows user, the lines
2349
2350 #if defined(_WIN32) || defined(WIN32)
2351 _setmode( _fileno( stdout ), 0x8000 );
2352 #endif /* defined(_WIN32) || defined(WIN32) */
2353
2354 have been added to the source of pcretest. This apparently does useful
2355 magic in relation to line terminators.
2356
2357 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2358 for the benefit of those environments where the "b" makes a difference.
2359
2360 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2361 to know about it. I have put a hack into configure.in that adds in code
2362 to set GCC=yes if CC=icc. This seems to end up at a point in the
2363 generated configure script that is early enough to affect the setting of
2364 compiler options, which is what is needed, but I have no means of testing
2365 whether it really works. (The user who reported this had patched the
2366 generated configure script, which of course I cannot do.)
2367
2368 LATER: After change 22 below (new libtool files), the configure script
2369 seems to know about icc (and also ecc). Therefore, I have commented out
2370 this hack in configure.in.
2371
2372 7. Added support for pkg-config (2 patches were sent in).
2373
2374 8. Negated POSIX character classes that used a combination of internal tables
2375 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2376 [[:^ascii]]. Typically, they would match almost any characters. The other
2377 POSIX classes were not broken in this way.
2378
2379 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2380 to find the match, as PCRE was deluded into thinking that the match had to
2381 start at the start point or following a newline. The same bug applied to
2382 patterns with negative forward assertions or any backward assertions
2383 preceding ".*" at the start, unless the pattern required a fixed first
2384 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2385
2386 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2387 starting at the last subject character, bytes beyond the end of the subject
2388 string were read.
2389
2390 11. Renamed the variable "class" as "classbits" to make life easier for C++
2391 users. (Previously there was a macro definition, but it apparently wasn't
2392 enough.)
2393
2394 12. Added the new field "tables" to the extra data so that tables can be passed
2395 in at exec time, or the internal tables can be re-selected. This allows
2396 a compiled regex to be saved and re-used at a later time by a different
2397 program that might have everything at different addresses.
2398
2399 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2400 -R library as well as a -L library.
2401
2402 14. The debugging options of pcretest (-d on the command line or D on a
2403 pattern) showed incorrect output for anything following an extended class
2404 that contained multibyte characters and which was followed by a quantifier.
2405
2406 15. Added optional support for general category Unicode character properties
2407 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2408 support. It adds about 90K to the size of the library. The meanings of the
2409 inbuilt class escapes such as \d and \s have NOT been changed.
2410
2411 16. Updated pcredemo.c to include calls to free() to release the memory for the
2412 compiled pattern.
2413
2414 17. The generated file chartables.c was being created in the source directory
2415 instead of in the building directory. This caused the build to fail if the
2416 source directory was different from the building directory, and was
2417 read-only.
2418
2419 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2420 file. No doubt somebody will tell me if they don't make sense... Also added
2421 Dan Mooney's comments about building on OpenVMS.
2422
2423 19. Added support for partial matching via the PCRE_PARTIAL option for
2424 pcre_exec() and the \P data escape in pcretest.
2425
2426 20. Extended pcretest with 3 new pattern features:
2427
2428 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2429 write the compiled pattern to the file whose name is "rest-of-line".
2430 This is a straight binary dump of the data, with the saved pointer to
2431 the character tables forced to be NULL. The study data, if any, is
2432 written too. After writing, pcretest reads a new pattern.
2433
2434 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2435 compiled pattern from the given file. There must not be any
2436 occurrences of "<" in the file name (pretty unlikely); if there are,
2437 pcretest will instead treat the initial "<" as a pattern delimiter.
2438 After reading in the pattern, pcretest goes on to read data lines as
2439 usual.
2440
2441 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2442 and 16-bit fields in a compiled pattern, to simulate a pattern that
2443 was compiled on a host of opposite endianness.
2444
2445 21. The pcre-exec() function can now cope with patterns that were compiled on
2446 hosts of opposite endianness, with this restriction:
2447
2448 As for any compiled expression that is saved and used later, the tables
2449 pointer field cannot be preserved; the extra_data field in the arguments
2450 to pcre_exec() should be used to pass in a tables address if a value
2451 other than the default internal tables were used at compile time.
2452
2453 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2454 now diagnosed as an error. Previously, most of the time, a negative number
2455 would have been treated as zero, but if in addition "ovector" was passed as
2456 NULL, a crash could occur.
2457
2458 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2459 new versions from the libtool 1.5 distribution (the last one is a copy of
2460 a file called libtool.m4). This seems to have fixed the need to patch
2461 "configure" to support Darwin 1.3 (which I used to do). However, I still
2462 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2463 workstation).
2464
2465 24. Changed the PCRE licence to be the more standard "BSD" licence.
2466
2467
2468 Version 4.5 01-Dec-03
2469 ---------------------
2470
2471 1. There has been some re-arrangement of the code for the match() function so
2472 that it can be compiled in a version that does not call itself recursively.
2473 Instead, it keeps those local variables that need separate instances for
2474 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2475 needs to "recurse". Keeping track of where control must go is done by means
2476 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2477 hide most of the details from the main code, and operates only if
2478 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2479 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2480 operating.
2481
2482 To make it easier for callers to provide specially tailored get/free
2483 functions for this usage, two new functions, pcre_stack_malloc, and
2484 pcre_stack_free, are used. They are always called in strict stacking order,
2485 and the size of block requested is always the same.
2486
2487 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2488 PCRE has been compiled to use the stack or the heap for recursion. The
2489 -C option of pcretest uses this to show which version is compiled.
2490
2491 A new data escape \S, is added to pcretest; it causes the amounts of store
2492 obtained and freed by both kinds of malloc/free at match time to be added
2493 to the output.
2494
2495 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2496 what's available on my current Linux desktop machine.
2497
2498 3. When matching a UTF-8 string, the test for a valid string at the start has
2499 been extended. If start_offset is not zero, PCRE now checks that it points
2500 to a byte that is the start of a UTF-8 character. If not, it returns
2501 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2502 this is necessary because there may be backward assertions in the pattern.
2503 When matching the same subject several times, it may save resources to use
2504 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2505
2506 4. The code for checking the validity of UTF-8 strings has been tightened so
2507 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2508 containing "overlong sequences".
2509
2510 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2511 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2512 should have been "&", but it just so happened that all the cases this let
2513 through by mistake were picked up later in the function.
2514
2515 6. I had used a variable called "isblank" - this is a C99 function, causing
2516 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2517
2518 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2519 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2520 so the version is shown; (c) stop "make test" echoing "./RunTest".
2521
2522 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2523
2524 9. The prototype for memmove() for systems that don't have it was using
2525 size_t, but the inclusion of the header that defines size_t was later. I've
2526 moved the #includes for the C headers earlier to avoid this.
2527
2528 10. Added some adjustments to the code to make it easier to compiler on certain
2529 special systems:
2530
2531 (a) Some "const" qualifiers were missing.
2532 (b) Added the macro EXPORT before all exported functions; by default this
2533 is defined to be empty.
2534 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2535 that it reads its output file name as an argument instead of writing
2536 to the standard output and assuming this can be redirected.
2537
2538 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2539 class containing characters with values greater than 255, PCRE compilation
2540 went into a loop.
2541
2542 12. A recursive reference to a subpattern that was within another subpattern
2543 that had a minimum quantifier of zero caused PCRE to crash. For example,
2544 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2545 recursion. If the recursively-called subpattern itself had a zero repeat,
2546 that was OK.
2547
2548 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2549 buffer into which it was copied (for escape processing) was still set at
2550 1024, so long lines caused crashes.
2551
2552 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2553 "internal error: code overflow...". This applied to any character class
2554 that was followed by a possessive quantifier.
2555
2556 15. Modified the Makefile to add libpcre.la as a prerequisite for
2557 libpcreposix.la because I was told this is needed for a parallel build to
2558 work.
2559
2560 16. If a pattern that contained .* following optional items at the start was
2561 studied, the wrong optimizing data was generated, leading to matching
2562 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2563 matching string must start with a or b or c. The correct conclusion for
2564 this pattern is that a match can start with any character.
2565
2566
2567 Version 4.4 13-Aug-03
2568 ---------------------
2569
2570 1. In UTF-8 mode, a character class containing characters with values between
2571 127 and 255 was not handled correctly if the compiled pattern was studied.
2572 In fixing this, I have also improved the studying algorithm for such
2573 classes (slightly).
2574
2575 2. Three internal functions had redundant arguments passed to them. Removal
2576 might give a very teeny performance improvement.
2577
2578 3. Documentation bug: the value of the capture_top field in a callout is *one
2579 more than* the number of the hightest numbered captured substring.
2580
2581 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2582 in incorrectly linking with a previously installed version. They now link
2583 explicitly with libpcre.la.
2584
2585 5. configure.in no longer needs to recognize Cygwin specially.
2586
2587 6. A problem in pcre.in for Windows platforms is fixed.
2588
2589 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2590 pcretest, it used to include the size of the study block as part of its
2591 output. Unfortunately, the structure contains a field that has a different
2592 size on different hardware architectures. This meant that the tests that
2593 showed this size failed. As the block is currently always of a fixed size,
2594 this information isn't actually particularly useful in pcretest output, so
2595 I have just removed it.
2596
2597 8. Three pre-processor statements accidentally did not start in column 1.
2598 Sadly, there are *still* compilers around that complain, even though
2599 standard C has not required this for well over a decade. Sigh.
2600
2601 9. In pcretest, the code for checking callouts passed small integers in the
2602 callout_data field, which is a void * field. However, some picky compilers
2603 complained about the casts involved for this on 64-bit systems. Now
2604 pcretest passes the address of the small integer instead, which should get
2605 rid of the warnings.
2606
2607 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2608 both compile and run time, and gives an error if an invalid UTF-8 sequence
2609 is found. There is a option for disabling this check in cases where the
2610 string is known to be correct and/or the maximum performance is wanted.
2611
2612 11. In response to a bug report, I changed one line in Makefile.in from
2613
2614 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2615 to
2616 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2617
2618 to look similar to other lines, but I have no way of telling whether this
2619 is the right thing to do, as I do not use Windows. No doubt I'll get told
2620 if it's wrong...
2621
2622
2623 Version 4.3 21-May-03
2624 ---------------------
2625
2626 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2627 Makefile.
2628
2629 2. Some refactoring to improve the quality of the code:
2630
2631 (i) The utf8_table... variables are now declared "const".
2632
2633 (ii) The code for \cx, which used the "case flipping" table to upper case
2634 lower case letters, now just substracts 32. This is ASCII-specific,
2635 but the whole concept of \cx is ASCII-specific, so it seems
2636 reasonable.
2637
2638 (iii) PCRE was using its character types table to recognize decimal and
2639 hexadecimal digits in the pattern. This is silly, because it handles
2640 only 0-9, a-f, and A-F, but the character types table is locale-
2641 specific, which means strange things might happen. A private
2642 table is now used for this - though it costs 256 bytes, a table is
2643 much faster than multiple explicit tests. Of course, the standard
2644 character types table is still used for matching digits in subject
2645 strings against \d.
2646
2647 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2648 ending in _t are). So I've renamed it as ESC_tee.
2649
2650 3. The first argument for regexec() in the POSIX wrapper should have been
2651 defined as "const".
2652
2653 4. Changed pcretest to use malloc() for its buffers so that they can be
2654 Electric Fenced for debugging.
2655
2656 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2657 to read one or more bytes before the start of the subject string. Often this
2658 had no effect on PCRE's behaviour, but in some circumstances it could
2659 provoke a segmentation fault.
2660
2661 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2662 to try to read one or more bytes before the start of the subject string.
2663
2664 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2665 UTF-8 support could misbehave in various ways if the subject string
2666 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2667 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2668 back over UTF-8 characters.)
2669
2670
2671 Version 4.2 14-Apr-03
2672 ---------------------
2673
2674 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2675
2676 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2677 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2678 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2679 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2680 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2681 and BUILD_EXEEXT
2682 Note: automatic setting of the BUILD variables is not yet working
2683 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2684 compile-time but not at link-time
2685 [LINK]: use for linking executables only
2686 make different versions for Windows and non-Windows
2687 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2688 libraries
2689 [LINK_FOR_BUILD]: new variable
2690 [OBJEXT]: use throughout
2691 [EXEEXT]: use throughout
2692 <winshared>: new target
2693 <wininstall>: new target
2694 <dftables.o>: use native compiler
2695 <dftables>: use native linker
2696 <install>: handle Windows platform correctly
2697 <clean>: ditto
2698 <check>: ditto
2699 copy DLL to top builddir before testing
2700
2701 As part of these changes, -no-undefined was removed again. This was reported
2702 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2703 in any case.
2704
2705 3. Some tidies to get rid of compiler warnings:
2706
2707 . In the match_data structure, match_limit was an unsigned long int, whereas
2708 match_call_count was an int. I've made them both unsigned long ints.
2709
2710 . In pcretest the fact that a const uschar * doesn't automatically cast to
2711 a void * provoked a warning.
2712
2713 . Turning on some more compiler warnings threw up some "shadow" variables
2714 and a few more missing casts.
2715
2716 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2717 option, a class that contained a single character with a value between 128
2718 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2719
2720 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2721 option, a class that contained several characters, but with at least one
2722 whose value was between 128 and 255 caused PCRE to crash.
2723
2724
2725 Version 4.1 12-Mar-03
2726 ---------------------
2727
2728 1. Compiling with gcc -pedantic found a couple of places where casts were
2729 needed, and a string in dftables.c that was longer than standard compilers are
2730 required to support.
2731
2732 2. Compiling with Sun's compiler found a few more places where the code could
2733 be tidied up in order to avoid warnings.
2734
2735 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2736 first of these names is deprecated in the latest Autoconf in favour of the name
2737 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2738 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2739 analogy I have changed it to CFLAGS_FOR_BUILD.
2740
2741 4. Added -no-undefined to the linking command in the Makefile, because this is
2742 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2743 linking step for the pcreposix library.
2744
2745 5. PCRE was failing to diagnose the case of two named groups with the same
2746 name.
2747
2748 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2749 literal character that is needed in the subject for a match, and scans along to
2750 ensure that it is present before embarking on the full matching process. This
2751 saves time in cases of nested unlimited repeats that are never going to match.
2752 Problem: the scan can take a lot of time if the subject is very long (e.g.
2753 megabytes), thus penalizing straightforward matches. It is now done only if the
2754 amount of subject to be scanned is less than 1000 bytes.
2755
2756 7. A lesser problem with the same optimization is that it was recording the
2757 first character of an anchored pattern as "needed", thus provoking a search
2758 right along the subject, even when the first match of the pattern was going to
2759 fail. The "needed" character is now not set for anchored patterns, unless it
2760 follows something in the pattern that is of non-fixed length. Thus, it still
2761 fulfils its original purpose of finding quick non-matches in cases of nested
2762 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2763
2764
2765 Version 4.0 17-Feb-03
2766 ---------------------
2767
2768 1. If a comment in an extended regex that started immediately after a meta-item
2769 extended to the end of string, PCRE compiled incorrect data. This could lead to
2770 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2771
2772 2. Moved to autoconf 2.53 and libtool 1.4.2.
2773
2774 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2775 the special perltest8 script is no longer needed - all the tests can be run
2776 from a single perltest script.
2777
2778 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2779 by \s. It has now been removed in PCRE. This means it isn't recognized as
2780 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2781 class [:space:] *does* include VT, thereby creating a mess.
2782
2783 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2784 space and tab.
2785
2786 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2787 its new features into the main test script, reducing the number of scripts.
2788
2789 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2790 were backward compatible, and made the (?i) apply to the whole pattern, as if
2791 /i were given. Now it behaves more logically, and applies the option setting
2792 only to what follows. PCRE has been changed to follow suit. However, if it
2793 finds options settings right at the start of the pattern, it extracts them into
2794 the global options, as before. Thus, they show up in the info data.
2795
2796 8. Added support for the \Q...\E escape sequence. Characters in between are
2797 treated as literals. This is slightly different from Perl in that $ and @ are
2798 also handled as literals inside the quotes. In Perl, they will cause variable
2799 interpolation. Note the following examples:
2800
2801 Pattern PCRE matches Perl matches
2802
2803 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2804 \Qabc\$xyz\E abc\$xyz abc\$xyz
2805 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2806
2807 For compatibility with Perl, \Q...\E sequences are recognized inside character
2808 classes as well as outside them.
2809
2810 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2811 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2812 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2813 signed/unsigned warnings.
2814
2815 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2816 option for pcretest, so I've replaced it by a simple function that does just
2817 that job.
2818
2819 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2820 "pcregrep -".
2821
2822 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2823 Java package. This provides some syntactic sugar for simple cases of what my
2824 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2825 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2826 item, you can use this simplified notation. Note that only makes sense with
2827 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2828 greediness, whatever the setting of the PCRE_UNGREEDY option.
2829
2830 13. A change of greediness default within a pattern was not taking effect at
2831 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2832 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2833 was abstracted outside.
2834
2835 14. PCRE now supports the \G assertion. It is true when the current matching
2836 position is at the start point of the match. This differs from \A when the
2837 starting offset is non-zero. Used with the /g option of pcretest (or similar
2838 code), it works in the same way as it does for Perl's /g option. If all
2839 alternatives of a regex begin with \G, the expression is anchored to the start
2840 match position, and the "anchored" flag is set in the compiled expression.
2841
2842 15. Some bugs concerning the handling of certain option changes within patterns
2843 have been fixed. These applied to options other than (?ims). For example,
2844 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2845 been the other way round. Some of this was related to change 7 above.
2846
2847 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2848 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2849 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2850 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2851
2852 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2853 mode. Unlike ".", it always matches newline, whatever the setting of
2854 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2855 assertions. Perl allows it, but it doesn't (in general) work because it can't
2856 calculate the length of the lookbehind. At least, that's the case for Perl
2857 5.8.0 - I've been told they are going to document that it doesn't work in
2858 future.
2859
2860 18. Added an error diagnosis for escapes that PCRE does not support: these are
2861 \L, \l, \N, \P, \p, \U, \u, and \X.
2862
2863 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2864 reading past the end of the pattern in cases such as /[abcd/.
2865
2866 20. PCRE was getting more memory than necessary for patterns with classes that
2867 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2868
2869 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2870 compiling PCRE for use with Virtual Pascal.
2871
2872 22. Small fix to the Makefile to make it work properly if the build is done
2873 outside the source tree.
2874
2875 23. Added a new extension: a condition to go with recursion. If a conditional
2876 subpattern starts with (?(R) the "true" branch is used if recursion has
2877 happened, whereas the "false" branch is used only at the top level.
2878
2879 24. When there was a very long string of literal characters (over 255 bytes
2880 without UTF support, over 250 bytes with UTF support), the computation of how
2881 much memory was required could be incorrect, leading to segfaults or other
2882 strange effects.
2883
2884 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2885 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2886 there was a subsequent back reference to those brackets. This meant that, for
2887 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2888 possible to check for precisely this case. All we can do is abandon the
2889 optimization if .* occurs inside capturing brackets when there are any back
2890 references whatsoever. (See below for a better fix that came later.)
2891
2892 26. The handling of the optimization for finding the first character of a
2893 non-anchored pattern, and for finding a character that is required later in the
2894 match were failing in some cases. This didn't break the matching; it just
2895 failed to optimize when it could. The way this is done has been re-implemented.
2896
2897 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2898
2899 28. Added a new feature that provides some of the functionality that Perl
2900 provides with (?{...}). The facility is termed a "callout". The way it is done
2901 in PCRE is for the caller to provide an optional function, by setting
2902 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2903 global variable. By default it is unset, which disables all calling out. To get
2904 the function called, the regex must include (?C) at appropriate points. This
2905 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2906 This provides a means of identifying different callout points. When PCRE
2907 reaches such a point in the regex, if pcre_callout has been set, the external
2908 function is called. It is provided with data in a structure called
2909 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2910 matching continues; if it returns a non-zero value, the match at the current
2911 point fails. However, backtracking will occur if possible. [This was changed
2912 later and other features added - see item 49 below.]
2913
2914 29. pcretest is upgraded to test the callout functionality. It provides a
2915 callout function that displays information. By default, it shows the start of
2916 the match and the current position in the text. There are some new data escapes
2917 to vary what happens:
2918
2919 \C+ in addition, show current contents of captured substrings
2920 \C- do not supply a callout function
2921 \C!n return 1 when callout number n is reached
2922 \C!n!m return 1 when callout number n is reached for the mth time
2923
2924 30. If pcregrep was called with the -l option and just a single file name, it
2925 output "<stdin>" if a match was found, instead of the file name.
2926
2927 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2928 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2929 pcre_exec(). This saves a malloc/free per call. The default value of
2930 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2931 when configuring.
2932
2933 32. The default maximum size of a compiled pattern is 64K. There have been a
2934 few cases of people hitting this limit. The code now uses macros to handle the
2935 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2936 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2937 configuring. Tests 2 and 5 work only with 2-byte links because they output
2938 debugging information about compiled patterns.
2939
2940 33. Internal code re-arrangements:
2941
2942 (a) Moved the debugging function for printing out a compiled regex into
2943 its own source file (printint.c) and used #include to pull it into
2944 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2945 separate copies.
2946
2947 (b) Defined the list of op-code names for debugging as a macro in
2948 internal.h so that it is next to the definition of the opcodes.
2949
2950 (c) Defined a table of op-code lengths for simpler skipping along compiled
2951 code. This is again a macro in internal.h so that it is next to the
2952 definition of the opcodes.
2953
2954 34. Added support for recursive calls to individual subpatterns, along the
2955 lines of Robin Houston's patch (but implemented somewhat differently).
2956
2957 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2958 allow it to read and process whole directories in Win32. This code was
2959 contributed by Lionel Fourquaux; it has not been tested by me.
2960
2961 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2962 used to name a group. Names consist of alphanumerics and underscores, and must
2963 be unique. Back references use the syntax (?P=name) and recursive calls use
2964 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2965 numbers. The function pcre_fullinfo() can be used after compilation to extract
2966 a name/number map. There are three relevant calls:
2967
2968 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2969 PCRE_INFO_NAMECOUNT yields the number of entries
2970 PCRE_INFO_NAMETABLE yields a pointer to the map.
2971
2972 The map is a vector of fixed-size entries. The size of each entry depends on
2973 the length of the longest name used. The first two bytes of each entry are the
2974 group number, most significant byte first. There follows the corresponding
2975 name, zero terminated. The names are in alphabetical order.
2976
2977 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2978 case instead of 255. Making it the same both with and without UTF-8 support
2979 means that the same test output works with both.
2980
2981 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2982 calling malloc() with a zero argument.
2983
2984 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2985 optimization. I've improved things by keeping a bitmap of backreferences with
2986 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2987 fact referenced, the optimization can be applied. It is unlikely that a
2988 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2989 the match to follow \n) will appear inside brackets with a number greater than
2990 31, but if it does, any back reference > 31 suppresses the optimization.
2991
2992 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2993 of disabling numbered capturing parentheses. Any opening parenthesis that is
2994 not followed by ? behaves as if it were followed by ?: but named parentheses
2995 can still be used for capturing (and they will acquire numbers in the usual
2996 way).
2997
2998 41. Redesigned the return codes from the match() function into yes/no/error so
2999 that errors can be passed back from deep inside the nested calls. A malloc
3000 failure while inside a recursive subpattern call now causes the
3001 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3002
3003 42. It is now possible to set a limit on the number of times the match()
3004 function is called in a call to pcre_exec(). This facility makes it possible to
3005 limit the amount of recursion and backtracking, though not in a directly
3006 obvious way, because the match() function is used in a number of different
3007 circumstances. The count starts from zero for each position in the subject
3008 string (for non-anchored patterns). The default limit is, for compatibility, a
3009 large number, namely 10 000 000. You can change this in two ways:
3010
3011 (a) When configuring PCRE before making, you can use --with-match-limit=n
3012 to set a default value for the compiled library.
3013
3014 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3015 a different value is set. See 45 below.
3016
3017 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3018
3019 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3020 of things that can be changed at compile time. The first argument specifies
3021 what is wanted and the second points to where the information is to be placed.
3022 The current list of available information is:
3023
3024 PCRE_CONFIG_UTF8
3025
3026 The output is an integer that is set to one if UTF-8 support is available;
3027 otherwise it is set to zero.
3028
3029 PCRE_CONFIG_NEWLINE
3030
3031 The output is an integer that it set to the value of the code that is used for
3032 newline. It is either LF (10) or CR (13).
3033
3034 PCRE_CONFIG_LINK_SIZE
3035
3036 The output is an integer that contains the number of bytes used for internal
3037 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3038
3039 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3040
3041 The output is an integer that contains the threshold above which the POSIX
3042 interface uses malloc() for output vectors. See item 31 above.
3043
3044 PCRE_CONFIG_MATCH_LIMIT
3045
3046 The output is an unsigned integer that contains the default limit of the number
3047 of match() calls in a pcre_exec() execution. See 42 above.
3048
3049 44. pcretest has been upgraded by the addition of the -C option. This causes it
3050 to extract all the available output from the new pcre_config() function, and to
3051 output it. The program then exits immediately.
3052
3053 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3054 order to support additional features. One way would have been to define
3055 pcre_exec2() (for example) with extra arguments, but this would not have been
3056 extensible, and would also have required all calls to the original function to
3057 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3058 is used for passing in "extra" data from pcre_study().
3059
3060 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3061 contains the following fields:
3062
3063 flags a bitmap indicating which of the following fields are set
3064 study_data opaque data from pcre_study()
3065 match_limit a way of specifying a limit on match() calls for a specific
3066 call to pcre_exec()
3067 callout_data data for callouts (see 49 below)
3068
3069 The flag bits are also defined in pcre.h, and are
3070
3071 PCRE_EXTRA_STUDY_DATA
3072 PCRE_EXTRA_MATCH_LIMIT
3073 PCRE_EXTRA_CALLOUT_DATA
3074
3075 The pcre_study() function now returns one of these new pcre_extra blocks, with
3076 the actual study data pointed to by the study_data field, and the
3077 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3078 before. That is, this change is entirely upwards-compatible and requires no
3079 change to existing code.
3080
3081 If you want to pass in additional data to pcre_exec(), you can either place it
3082 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3083 block.
3084
3085 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3086 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3087 times with different match limits, until it finds the minimum value needed for
3088 pcre_exec() to complete. The value is then output. This can be instructive; for
3089 most simple matches the number is quite small, but for pathological cases it
3090 gets very large very quickly.
3091
3092 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3093 returns the size of the data block pointed to by the study_data field in a
3094 pcre_extra block, that is, the value that was passed as the argument to
3095 pcre_malloc() when PCRE was getting memory in which to place the information
3096 created by pcre_study(). The fourth argument should point to a size_t variable.
3097 pcretest has been extended so that this information is shown after a successful
3098 pcre_study() call when information about the compiled regex is being displayed.
3099
3100 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3101 because what follows is always an absolute path. (Later: it turns out that this
3102 is more than cosmetic for MinGW, because it doesn't like empty path
3103 components.)
3104
3105 49. Some changes have been made to the callout feature (see 28 above):
3106
3107 (i) A callout function now has three choices for what it returns:
3108
3109 0 => success, carry on matching
3110 > 0 => failure at this point, but backtrack if possible
3111 < 0 => serious error, return this value from pcre_exec()
3112
3113 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3114 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3115 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3116 use by callout functions. It will never be used by PCRE itself.
3117
3118 (ii) The pcre_extra structure (see 45 above) has a void * field called
3119 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3120 pcre_callout_block structure has a field of the same name. The contents of
3121 the field passed in the pcre_extra structure are passed to the callout
3122 function in the corresponding field in the callout block. This makes it
3123 easier to use the same callout-containing regex from multiple threads. For
3124 testing, the pcretest program has a new data escape
3125
3126 \C*n pass the number n (may be negative) as callout_data
3127
3128 If the callout function in pcretest receives a non-zero value as
3129 callout_data, it returns that value.
3130
3131 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3132 there were some redundant $(CFLAGS) in commands that are now specified as
3133 $(LINK), which already includes $(CFLAGS).
3134
3135 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3136 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3137 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3138 one-byte characters throughout. Note that case-insensitive matching applies
3139 only to characters whose values are less than 256. PCRE doesn't support the
3140 notion of cases for higher-valued characters.
3141
3142 (i) A character class whose characters are all within 0-255 is handled as
3143 a bit map, and the map is inverted for negative classes. Previously, a
3144 character > 255 always failed to match such a class; however it should
3145 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3146
3147 (ii) A negated character class with a single character < 255 is coded as
3148 "not this character" (OP_NOT). This wasn't working properly when the test
3149 character was multibyte, either singly or repeated.
3150
3151 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3152 mode, for example: \x{100}{2,3}.
3153
3154 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3155 singly or repeated) now correctly test multibyte characters. However,
3156 PCRE doesn't recognize any characters with values greater than 255 as
3157 digits, spaces, or word characters. Such characters always match \D, \S,
3158 and \W, and never match \d, \s, or \w.
3159
3160 (v) Classes may now contain characters and character ranges with values
3161 greater than 255. For example: [ab\x{100}-\x{400}].
3162
3163 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3164 PCRE in UTF-8 mode.
3165
3166 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3167 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3168 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3169 value.)
3170
3171 53. The single man page has become too large. I have therefore split it up into
3172 a number of separate man pages. These also give rise to individual HTML pages;
3173 these are now put in a separate directory, and there is an index.html page that
3174 lists them all. Some hyperlinking between the pages has been installed.
3175
3176 54. Added convenience functions for handling named capturing parentheses.
3177
3178 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3179 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3180 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3181 are faulted.
3182
3183 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3184 calling configure. These values are used when compiling the dftables.c program
3185 which is run to generate the source of the default character tables. They
3186 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3187 you will need to set these values.
3188
3189 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3190
3191
3192 Version 3.9 02-Jan-02
3193 ---------------------
3194
3195 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3196
3197 2. If --disable-static was given, the building process failed when trying to
3198 build pcretest and pcregrep. (For some reason it was using libtool to compile
3199 them, which is not right, as they aren't part of the library.)
3200
3201
3202 Version 3.8 18-Dec-01
3203 ---------------------
3204
3205 1. The experimental UTF-8 code was completely screwed up. It was packing the
3206 bytes in the wrong order. How dumb can you get?
3207
3208
3209 Version 3.7 29-Oct-01
3210 ---------------------
3211
3212 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3213 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3214 this didn't happen under Solaris 8, where I normally test things.
3215
3216 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3217 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3218 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3219 this fix an hour or so after the initial 3.7 release.)
3220
3221
3222 Version 3.6 23-Oct-01
3223 ---------------------
3224
3225 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3226 offsets passed as NULL with zero offset count.
3227
3228 2. The config.guess and config.sub files had not been updated when I moved to
3229 the latest autoconf.
3230
3231
3232 Version 3.5 15-Aug-01
3233 ---------------------
3234
3235 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3236 had been forgotten.
3237
3238 2. By using declared but undefined structures, we can avoid using "void"
3239 definitions in pcre.h while keeping the internal definitions of the structures
3240 private.
3241
3242 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3243 user point of view, this means that both static and shared libraries are built
3244 by default, but this can be individually controlled. More of the work of
3245 handling this static/shared cases is now inside libtool instead of PCRE's make
3246 file.
3247
3248 4. The pcretest utility is now installed along with pcregrep because it is
3249 useful for users (to test regexs) and by doing this, it automatically gets
3250 relinked by libtool. The documentation has been turned into a man page, so
3251 there are now .1, .txt, and .html versions in /doc.
3252
3253 5. Upgrades to pcregrep:
3254 (i) Added long-form option names like gnu grep.
3255 (ii) Added --help to list all options with an explanatory phrase.
3256 (iii) Added -r, --recursive to recurse into sub-directories.
3257 (iv) Added -f, --file to read patterns from a file.
3258
3259 6. pcre_exec() was referring to its "code" argument before testing that
3260 argument for NULL (and giving an error if it was NULL).
3261
3262 7. Upgraded Makefile.in to allow for compiling in a different directory from
3263 the source directory.
3264
3265 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3266 options bits, the pointer it was passed was to an int instead of to an unsigned
3267 long int. This mattered only on 64-bit systems.
3268
3269 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3270 generated) instead of pcre.in, which it its source. Also made the same change
3271 in several of the .c files.
3272
3273 10. A new release of gcc defines printf() as a macro, which broke pcretest
3274 because it had an ifdef in the middle of a string argument for printf(). Fixed
3275 by using separate calls to printf().
3276
3277 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3278 script, to force use of CR or LF instead of \n in the source. On non-Unix
3279 systems, the value can be set in config.h.
3280
3281 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3282 absolute limit. Changed the text of the error message to make this clear, and
3283 likewise updated the man page.
3284
3285 13. The limit of 99 on the number of capturing subpatterns has been removed.
3286 The new limit is 65535, which I hope will not be a "real" limit.
3287
3288
3289 Version 3.4 22-Aug-00
3290 ---------------------
3291
3292 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3293
3294 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3295
3296
3297 Version 3.3 01-Aug-00
3298 ---------------------
3299
3300 1. If an octal character was given, but the value was greater than \377, it
3301 was not getting masked to the least significant bits, as documented. This could
3302 lead to crashes in some systems.
3303
3304 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3305 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3306
3307 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3308 These just pass their arguments on to (pcre_free)(), but they are provided
3309 because some uses of PCRE bind it to non-C systems that can call its functions,
3310 but cannot call free() or pcre_free() directly.
3311
3312 4. Add "make test" as a synonym for "make check". Corrected some comments in
3313 the Makefile.
3314
3315 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3316 Makefile.
3317
3318 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3319 command called pgrep for grepping around the active processes.
3320
3321 7. Added the beginnings of support for UTF-8 character strings.
3322
3323 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3324 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3325 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3326 out for the ar command.)
3327
3328
3329 Version 3.2 12-May-00
3330 ---------------------
3331
3332 This is purely a bug fixing release.
3333
3334 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3335 of ZA. This was just one example of several cases that could provoke this bug,
3336 which was introduced by change 9 of version 2.00. The code for breaking
3337 infinite loops after an iteration that matches an empty string was't working
3338 correctly.
3339
3340 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3341 when matched against abbab (for example). After matching an empty string, it
3342 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3343 caused it to match further down the string than it should.
3344
3345 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3346 was there because it doesn't seem to be needed, and it causes trouble on some
3347 systems, as it is not a Standard C header. It has been removed.
3348
3349 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3350 were reported on the Macintosh. The changes were from
3351
3352 while ((c = *(++ptr)) != 0 && c != '\n');
3353 to
3354 while ((c = *(++ptr)) != 0 && c != '\n') ;
3355
3356 Totally extraordinary, but if that's what it takes...
3357
3358 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3359 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3360 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3361 assumes the way PCRE uses memmove() (always moving upwards).
3362
3363 6. PCRE is being used in one environment where strchr() is not available. There
3364 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3365 faster code anyway.
3366
3367
3368 Version 3.1 09-Feb-00
3369 ---------------------
3370
3371 The only change in this release is the fixing of some bugs in Makefile.in for
3372 the "install" target:
3373
3374 (1) It was failing to install pcreposix.h.
3375
3376 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3377
3378
3379 Version 3.0 01-Feb-00
3380 ---------------------
3381
3382 1. Add support for the /+ modifier to perltest (to output $` like it does in
3383 pcretest).
3384
3385 2. Add support for the /g modifier to perltest.
3386
3387 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3388 matches null strings.
3389
3390 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3391 pattern. Perl treats empty patterns specially - it reuses the most recent
3392 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3393 effect.
3394
3395 5. The POSIX interface was broken in that it was just handing over the POSIX
3396 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3397 required a bigger vector, with some working space on the end. This means that
3398 the POSIX wrapper now has to get and free some memory, and copy the results.
3399
3400 6. Added some simple autoconf support, placing the test data and the
3401 documentation in separate directories, re-organizing some of the
3402 information files, and making it build pcre-config (a GNU standard). Also added
3403 libtool support for building PCRE as a shared library, which is now the
3404 default.
3405
3406 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3407 09 are not valid octal constants. Single digits will be used for minor values
3408 less than 10.
3409
3410 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3411 existing programs that set these in the POSIX interface can use PCRE without
3412 modification.
3413
3414 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3415 return all that pcre_info() returns, plus additional data. The pcre_info()
3416 function is retained for compatibility, but is considered to be obsolete.
3417
3418 10. Added experimental recursion feature (?R) to handle one common case that
3419 Perl 5.6 will be able to do with (?p{...}).
3420
3421 11. Added support for POSIX character classes like [:alpha:], which Perl is
3422 adopting.
3423
3424
3425 Version 2.08 31-Aug-99
3426 ----------------------
3427
3428 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3429 trying to match at the startoffset position, but instead was moving forward to
3430 the next newline as if a previous match had failed.
3431
3432 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3433 and could get into a loop if a null string was matched other than at the start
3434 of the subject.
3435
3436 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3437 be distinguished at compile time, and for completeness also added PCRE_DATE.
3438
3439 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3440 in GnuWin32 environments.
3441
3442
3443 Version 2.07 29-Jul-99
3444 ----------------------
3445
3446 1. The documentation is now supplied in plain text form and HTML as well as in
3447 the form of man page sources.
3448
3449 2. C++ compilers don't like assigning (void *) values to other pointer types.
3450 In particular this affects malloc(). Although there is no problem in Standard
3451 C, I've put in casts to keep C++ compilers happy.
3452
3453 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3454 should be (const char *).
3455
3456 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3457 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3458 However, I haven't made this a standard facility. The documentation doesn't
3459 mention it, and the Makefile doesn't support it.
3460
3461 5. The Makefile now contains an "install" target, with editable destinations at
3462 the top of the file. The pcretest program is not installed.
3463
3464 6. pgrep -V now gives the PCRE version number and date.
3465
3466 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3467 causing the entire string to be ignored, instead of just the last character.
3468
3469 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3470 non-matching string, it can take a very, very long time, even for strings of
3471 quite modest length, because of the nested recursion. PCRE now does better in
3472 some of these cases. It does this by remembering the last required literal
3473 character in the pattern, and pre-searching the subject to ensure it is present
3474 before running the real match. In other words, it applies a heuristic to detect
3475 some types of certain failure quickly, and in the above example, if presented
3476 with a string that has no trailing " it gives "no match" very quickly.
3477
3478 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3479 other alternatives are tried instead.
3480
3481
3482 Version 2.06 09-Jun-99
3483 ----------------------
3484
3485 1. Change pcretest's output for amount of store used to show just the code
3486 space, because the remainder (the data block) varies in size between 32-bit and
3487 64-bit systems.
3488
3489 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3490 start matching at. This allows lookbehinds to work when searching for multiple
3491 occurrences in a string.
3492
3493 3. Added additional options to pcretest for testing multiple occurrences:
3494
3495 /+ outputs the rest of the string that follows a match
3496 /g loops for multiple occurrences, using the new startoffset argument
3497 /G loops for multiple occurrences by passing an incremented pointer
3498
3499 4. PCRE wasn't doing the "first character" optimization for patterns starting
3500 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3501 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3502 the letter 'x'. On long subject strings, this gives a significant speed-up.
3503
3504
3505 Version 2.05 21-Apr-99
3506 ----------------------
3507
3508 1. Changed the type of magic_number from int to long int so that it works
3509 properly on 16-bit systems.
3510
3511 2. Fixed a bug which caused patterns starting with .* not to work correctly
3512 when the subject string contained newline characters. PCRE was assuming
3513 anchoring for such patterns in all cases, which is not correct because .* will
3514 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3515 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3516 must be retried after every newline in the subject.
3517
3518
3519 Version 2.04 18-Feb-99
3520 ----------------------
3521
3522 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3523 computation of the store needed to hold the pattern was incorrect (too large).
3524 If such patterns were nested a few deep, this could multiply and become a real
3525 problem.
3526
3527 2. Added /M option to pcretest to show the memory requirement of a specific
3528 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3529
3530 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3531 compiled in such a way that the backtracking after subsequent failure was
3532 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3533 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3534
3535
3536 Version 2.03 02-Feb-99
3537 ----------------------
3538
3539 1. Fixed typo and small mistake in man page.
3540
3541 2. Added 4th condition (GPL supersedes if conflict) and created separate
3542 LICENCE file containing the conditions.
3543
3544 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3545 Perl, that is the internal \ allows the delimiter to be included in the
3546 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3547 the final delimiter, add \ to the end of the pattern (to test the error).
3548
3549 4. Added the convenience functions for extracting substrings after a successful
3550 match. Updated pcretest to make it able to test these functions.
3551
3552
3553 Version 2.02 14-Jan-99
3554 ----------------------
3555
3556 1. Initialized the working variables associated with each extraction so that
3557 their saving and restoring doesn't refer to uninitialized store.
3558
3559 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3560 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3561 fix the problem.
3562
3563 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3564 calls, and wasn't printing the correct value for compiling calls. Increased the
3565 default value of LOOPREPEAT, and the number of significant figures in the
3566 times.
3567
3568 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3569
3570 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3571 a building problem on Windows NT with a FAT file system.
3572
3573
3574 Version 2.01 21-Oct-98
3575 ----------------------
3576
3577 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3578 to character tables built by pcre_maketables() in the current locale. If NULL
3579 is passed, the default tables are used.
3580
3581
3582 Version 2.00 24-Sep-98
3583 ----------------------
3584
3585 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3586 it any more.
3587
3588 2. Allow quantification of (?>) groups, and make it work correctly.
3589
3590 3. The first character computation wasn't working for (?>) groups.
3591
3592 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3593 end of the subject) and add 5.005's \z, which really does match only at the
3594 very end of the subject.
3595
3596 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3597
3598 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3599 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3600 localized options. All options to pcre_study() were also removed.
3601
3602 7. Add other new features from 5.005:
3603
3604 $(?<= positive lookbehind
3605 $(?<! negative lookbehind
3606 (?imsx-imsx) added the unsetting capability
3607 such a setting is global if at outer level; local otherwise
3608 (?imsx-imsx:) non-capturing groups with option setting
3609 (?(cond)re|re) conditional pattern matching
3610
3611 A backreference to itself in a repeated group matches the previous
3612 captured string.
3613
3614 8. General tidying up of studying (both automatic and via "study")
3615 consequential on the addition of new assertions.
3616
3617 9. As in 5.005, unlimited repeated groups that could match an empty substring
3618 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3619 runtime if any iteration does actually match an empty substring.
3620
3621 10. Include the RunTest script in the distribution.
3622
3623 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3624 discrepancies, some of which were old and were also with respect to 5.004. They
3625 have now been fixed.
3626
3627
3628 Version 1.09 28-Apr-98
3629 ----------------------
3630
3631 1. A negated single character class followed by a quantifier with a minimum
3632 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3633 program crashes, or just wrong answers. This did not apply to negated classes
3634 containing more than one character, or to minima other than one.
3635
3636
3637 Version 1.08 27-Mar-98
3638 ----------------------
3639
3640 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3641
3642 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3643 latter must appear before anything that relies on it in the pattern.
3644
3645
3646 Version 1.07 16-Feb-98
3647 ----------------------
3648
3649 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3650 repeat of a potentially empty string).
3651
3652
3653 Version 1.06 23-Jan-98
3654 ----------------------
3655
3656 1. Added Markus Oberhumer's little patches for C++.
3657
3658 2. Literal strings longer than 255 characters were broken.
3659
3660
3661 Version 1.05 23-Dec-97
3662 ----------------------
3663
3664 1. Negated character classes containing more than one character were failing if
3665 PCRE_CASELESS was set at run time.
3666
3667
3668 Version 1.04 19-Dec-97
3669 ----------------------
3670
3671 1. Corrected the man page, where some "const" qualifiers had been omitted.
3672
3673 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3674 input syntax.
3675
3676 3. Fixed memory leak which occurred when a regex with back references was
3677 matched with an offsets vector that wasn't big enough. The temporary memory
3678 that is used in this case wasn't being freed if the match failed.
3679
3680 4. Tidied pcretest to ensure it frees memory that it gets.
3681
3682 5. Temporary memory was being obtained in the case where the passed offsets
3683 vector was exactly big enough.
3684
3685 6. Corrected definition of offsetof() from change 5 below.
3686
3687 7. I had screwed up change 6 below and broken the rules for the use of
3688 setjmp(). Now fixed.
3689
3690
3691 Version 1.03 18-Dec-97
3692 ----------------------
3693
3694 1. A erroneous regex with a missing opening parenthesis was correctly
3695 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3696 on some systems.
3697
3698 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3699 it was reported that one broken compiler failed on the former because "code" is
3700 also an independent variable.
3701
3702 3. The erroneous regex a[]b caused an array overrun reference.
3703
3704 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3705 fail on data ending with that character. (It was going on too far, and checking
3706 the next character, typically a binary zero.) This was specific to the
3707 optimized code for single-character negative classes.
3708
3709 5. Added a contributed patch from the TIN world which does the following:
3710
3711 + Add an undef for memmove, in case the the system defines a macro for it.
3712
3713 + Add a definition of offsetof(), in case there isn't one. (I don't know
3714 the reason behind this - offsetof() is part of the ANSI standard - but
3715 it does no harm).
3716
3717 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3718 most of the places where whitespace preceded '#'. I have given up and
3719 allowed the remaining 2 cases to be at the margin.
3720
3721 + Rename some variables in pcre to eliminate shadowing. This seems very
3722 pedantic, but does no harm, of course.
3723
3724 6. Moved the call to setjmp() into its own function, to get rid of warnings
3725 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3726
3727 7. Constructs such as \d{8,} were compiling into the equivalent of
3728 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3729 outcome, but in this particular case used more store than had been allocated,
3730 which caused the bug to be discovered because it threw up an internal error.
3731
3732 8. The debugging code in both pcre and pcretest for outputting the compiled
3733 form of a regex was going wrong in the case of back references followed by
3734 curly-bracketed repeats.
3735
3736
3737 Version 1.02 12-Dec-97
3738 ----------------------
3739
3740 1. Typos in pcre.3 and comments in the source fixed.
3741
3742 2. Applied a contributed patch to get rid of places where it used to remove
3743 'const' from variables, and fixed some signed/unsigned and uninitialized
3744 variable warnings.
3745
3746 3. Added the "runtest" target to Makefile.
3747
3748 4. Set default compiler flag to -O2 rather than just -O.
3749
3750
3751 Version 1.01 19-Nov-97
3752 ----------------------
3753
3754 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3755 like /([ab]*)*/, that is, for classes with more than one character in them.
3756
3757 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3758 as /((?>a*))*/ (a PCRE_EXTRA facility).
3759
3760
3761 Version 1.00 18-Nov-97
3762 ----------------------
3763
3764 1. Added compile-time macros to support systems such as SunOS4 which don't have
3765 memmove() or strerror() but have other things that can be used instead.
3766
3767 2. Arranged that "make clean" removes the executables.
3768
3769
3770 Version 0.99 27-Oct-97
3771 ----------------------
3772
3773 1. Fixed bug in code for optimizing classes with only one character. It was
3774 initializing a 32-byte map regardless, which could cause it to run off the end
3775 of the memory it had got.
3776
3777 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3778
3779
3780 Version 0.98 22-Oct-97
3781 ----------------------
3782
3783 1. Fixed bug in code for handling temporary memory usage when there are more
3784 back references than supplied space in the ovector. This could cause segfaults.
3785
3786
3787 Version 0.97 21-Oct-97
3788 ----------------------
3789
3790 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3791
3792 2. Optimized negated single characters not to use a bit map.
3793
3794 3. Brought error texts together as macro definitions; clarified some of them;
3795 fixed one that was wrong - it said "range out of order" when it meant "invalid
3796 escape sequence".
3797
3798 4. Changed some char * arguments to const char *.
3799
3800 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3801
3802 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3803 pcretest.
3804
3805
3806 Version 0.96 16-Oct-97
3807 ----------------------
3808
3809 1. Added a simple "pgrep" utility to the distribution.
3810
3811 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3812 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3813 where "ddd" means "one or more decimal digits".
3814
3815 3. Fixed serious bug. If a pattern had a back reference, but the call to
3816 pcre_exec() didn't supply a large enough ovector to record the related
3817 identifying subpattern, the match always failed. PCRE now remembers the number
3818 of the largest back reference, and gets some temporary memory in which to save
3819 the offsets during matching if necessary, in order to ensure that
3820 backreferences always work.
3821
3822 4. Increased the compatibility with Perl in a number of ways:
3823
3824 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3825 to request this handling. The option can be set at compile or exec time.
3826
3827 (b) $ matches before a terminating newline by default; an option
3828 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3829 mode). The option can be set at compile or exec time.
3830
3831 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3832 the same as Perl's. If the decimal number it represents is less than 10
3833 or there aren't that many previous left capturing parentheses, an octal
3834 escape is read. Inside a character class, it's always an octal escape,
3835 even if it is a single digit.
3836
3837 (d) An escaped but undefined alphabetic character is taken as a literal,
3838 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3839 escapes.
3840
3841 (e) {0} is now permitted. (The previous item is removed from the compiled
3842 pattern).
3843
3844 5. Changed all the names of code files so that the basic parts are no longer
3845 than 10 characters, and abolished the teeny "globals.c" file.
3846
3847 6. Changed the handling of character classes; they are now done with a 32-byte
3848 bit map always.
3849
3850 7. Added the -d and /D options to pcretest to make it possible to look at the
3851 internals of compilation without having to recompile pcre.
3852
3853
3854 Version 0.95 23-Sep-97
3855 ----------------------
3856
3857 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3858 \x20 at the start of a run of normal characters. These were being treated as
3859 real characters, instead of the source characters being re-checked.
3860
3861
3862 Version 0.94 18-Sep-97
3863 ----------------------
3864
3865 1. The functions are now thread-safe, with the caveat that the global variables
3866 containing pointers to malloc() and free() or alternative functions are the
3867 same for all threads.
3868
3869 2. Get pcre_study() to generate a bitmap of initial characters for non-
3870 anchored patterns when this is possible, and use it if passed to pcre_exec().
3871
3872
3873 Version 0.93 15-Sep-97
3874 ----------------------
3875
3876 1. /(b)|(:+)/ was computing an incorrect first character.
3877
3878 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3879 but not actually doing anything yet.
3880
3881 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3882 as Perl does (e.g. [-az] or [az-]).
3883
3884 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3885 all possible positions.
3886
3887 5. Split up into different modules to avoid including unneeded functions in a
3888 compiled binary. However, compile and exec are still in one module. The "study"
3889 function is split off.
3890
3891 6. The character tables are now in a separate module whose source is generated
3892 by an auxiliary program - but can then be edited by hand if required. There are
3893 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3894 toupper() in the code.
3895
3896 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3897 make them global. Abolish the function for setting them, as the caller can now
3898 set them directly.
3899
3900
3901 Version 0.92 11-Sep-97
3902 ----------------------
3903
3904 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3905 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3906
3907 2. Caseless matching was not working in character classes if the characters in
3908 the pattern were in upper case.
3909
3910 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3911
3912 4. Make PCRE_ANCHORED public and accept as a compile option.
3913
3914 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3915 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3916 pass them.
3917
3918 6. Give an error if bad option bits passed at compile or run time.
3919
3920 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3921 pcretest to cause it to pass that flag.
3922
3923 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3924 options, and the first character, if set.
3925
3926 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3927
3928
3929 Version 0.91 10-Sep-97
3930 ----------------------
3931
3932 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3933 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3934
3935 2. PCRE was looping on encountering an indefinitely repeated back reference to
3936 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3937 Perl does - treats the match as successful.
3938
3939 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5