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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 556 - (show annotations)
Tue Oct 26 11:06:44 2010 UTC (4 years, 6 months ago) by ph10
File size: 192559 byte(s)
Error occurred while calculating annotation data.
Fix #-comment bugs in UTF-8 mode with PCRE_NEWLINE_ANY.
1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.11 10-Oct-2010
5 ------------------------
6
7 1. (*THEN) was not working properly if there were untried alternatives prior
8 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
9 backtracked to try for "b" instead of moving to the next alternative branch
10 at the same level (in this case, to look for "c"). The Perl documentation
11 is clear that when (*THEN) is backtracked onto, it goes to the "next
12 alternative in the innermost enclosing group".
13
14 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
15 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
16 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
17 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
18 (*THEN).
19
20 3. If \s appeared in a character class, it removed the VT character from
21 the class, even if it had been included by some previous item, for example
22 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
23 of \s, but is part of the POSIX "space" class.)
24
25 4. A partial match never returns an empty string (because you can always
26 match an empty string at the end of the subject); however the checking for
27 an empty string was starting at the "start of match" point. This has been
28 changed to the "earliest inspected character" point, because the returned
29 data for a partial match starts at this character. This means that, for
30 example, /(?<=abc)def/ gives a partial match for the subject "abc"
31 (previously it gave "no match").
32
33 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
34 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
35 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
36 has an implication that the given string is incomplete (because a partial
37 match is preferred over a full match). For this reason, these items now
38 give a partial match in this situation. [Aside: previously, the one case
39 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
40 match rather than a full match, which was wrong by the old rules, but is
41 now correct.]
42
43 6. There was a bug in the handling of #-introduced comments, recognized when
44 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
45 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
46 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
47 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
48 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
49 places in pcre_compile().
50
51 7. Related to (6) above, when pcre_compile() was skipping #-introduced
52 comments when looking ahead for named forward references to subpatterns,
53 the only newline sequence it recognized was NL. It now handles newlines
54 according to the set newline convention.
55
56
57 Version 8.10 25-Jun-2010
58 ------------------------
59
60 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
61 THEN.
62
63 2. (*ACCEPT) was not working when inside an atomic group.
64
65 3. Inside a character class, \B is treated as a literal by default, but
66 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
67 causes the error). The code is unchanged, but I tidied the documentation.
68
69 4. Inside a character class, PCRE always treated \R and \X as literals,
70 whereas Perl faults them if its -w option is set. I have changed PCRE so
71 that it faults them when PCRE_EXTRA is set.
72
73 5. Added support for \N, which always matches any character other than
74 newline. (It is the same as "." when PCRE_DOTALL is not set.)
75
76 6. When compiling pcregrep with newer versions of gcc which may have
77 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
78 declared with attribute warn_unused_result" were given. Just casting the
79 result to (void) does not stop the warnings; a more elaborate fudge is
80 needed. I've used a macro to implement this.
81
82 7. Minor change to pcretest.c to avoid a compiler warning.
83
84 8. Added four artifical Unicode properties to help with an option to make
85 \s etc use properties (see next item). The new properties are: Xan
86 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
87
88 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
89 use Unicode properties. (*UCP) at the start of a pattern can be used to set
90 this option. Modified pcretest to add /W to test this facility. Added
91 REG_UCP to make it available via the POSIX interface.
92
93 10. Added --line-buffered to pcregrep.
94
95 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
96 studied, and the match started with a letter with a code point greater than
97 127 whose first byte was different to the first byte of the other case of
98 the letter, the other case of this starting letter was not recognized
99 (#976).
100
101 12. If a pattern that was studied started with a repeated Unicode property
102 test, for example, \p{Nd}+, there was the theoretical possibility of
103 setting up an incorrect bitmap of starting bytes, but fortunately it could
104 not have actually happened in practice until change 8 above was made (it
105 added property types that matched character-matching opcodes).
106
107 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
108 possible starting bytes for non-anchored patterns.
109
110 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
111 \R, and also a number of cases that involve Unicode properties, both
112 explicit and implicit when PCRE_UCP is set.
113
114 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
115 input, it could crash or give wrong results if characters with values
116 greater than 0xc0 were present in the subject string. (Detail: it assumed
117 UTF-8 input when processing these items.)
118
119 16. Added a lot of (int) casts to avoid compiler warnings in systems where
120 size_t is 64-bit (#991).
121
122 17. Added a check for running out of memory when PCRE is compiled with
123 --disable-stack-for-recursion (#990).
124
125 18. If the last data line in a file for pcretest does not have a newline on
126 the end, a newline was missing in the output.
127
128 19. The default pcre_chartables.c file recognizes only ASCII characters (values
129 less than 128) in its various bitmaps. However, there is a facility for
130 generating tables according to the current locale when PCRE is compiled. It
131 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
132 characters, are recognized by isspace() and therefore were getting set in
133 these tables, and indeed these tables seem to approximate to ISO 8859. This
134 caused a problem in UTF-8 mode when pcre_study() was used to create a list
135 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
136 which of course cannot start UTF-8 characters. I have changed the code so
137 that only real ASCII characters (less than 128) and the correct starting
138 bytes for UTF-8 encodings are set for characters greater than 127 when in
139 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
140 altogether.)
141
142 20. Added the /T option to pcretest so as to be able to run tests with non-
143 standard character tables, thus making it possible to include the tests
144 used for 19 above in the standard set of tests.
145
146 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
147 reference to a subpattern the other side of a comment that contains an
148 opening parenthesis caused either an internal compiling error, or a
149 reference to the wrong subpattern.
150
151
152 Version 8.02 19-Mar-2010
153 ------------------------
154
155 1. The Unicode data tables have been updated to Unicode 5.2.0.
156
157 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
158 configured.
159
160 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
161 original author of that file, following a query about its status.
162
163 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
164 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
165
166 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
167 quantifier applied to a forward-referencing subroutine call, could compile
168 incorrect code or give the error "internal error: previously-checked
169 referenced subpattern not found".
170
171 6. Both MS Visual Studio and Symbian OS have problems with initializing
172 variables to point to external functions. For these systems, therefore,
173 pcre_malloc etc. are now initialized to local functions that call the
174 relevant global functions.
175
176 7. There were two entries missing in the vectors called coptable and poptable
177 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
178 I've fixed the data, and added a kludgy way of testing at compile time that
179 the lengths are correct (equal to the number of opcodes).
180
181 8. Following on from 7, I added a similar kludge to check the length of the
182 eint vector in pcreposix.c.
183
184 9. Error texts for pcre_compile() are held as one long string to avoid too
185 much relocation at load time. To find a text, the string is searched,
186 counting zeros. There was no check for running off the end of the string,
187 which could happen if a new error number was added without updating the
188 string.
189
190 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
191
192 11. \K was not working if it appeared in an atomic group or in a group that
193 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
194 \K is "not well defined" if used in an assertion. PCRE now accepts it if
195 the assertion is positive, but not if it is negative.
196
197 12. Change 11 fortuitously reduced the size of the stack frame used in the
198 "match()" function of pcre_exec.c by one pointer. Forthcoming
199 implementation of support for (*MARK) will need an extra pointer on the
200 stack; I have reserved it now, so that the stack frame size does not
201 decrease.
202
203 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
204 item in branch that calls a recursion is a subroutine call - as in the
205 second branch in the above example - was incorrectly given the compile-
206 time error "recursive call could loop indefinitely" because pcre_compile()
207 was not correctly checking the subroutine for matching a non-empty string.
208
209 14. The checks for overrunning compiling workspace could trigger after an
210 overrun had occurred. This is a "should never occur" error, but it can be
211 triggered by pathological patterns such as hundreds of nested parentheses.
212 The checks now trigger 100 bytes before the end of the workspace.
213
214 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
215
216
217 Version 8.01 19-Jan-2010
218 ------------------------
219
220 1. If a pattern contained a conditional subpattern with only one branch (in
221 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
222 computed the wrong minimum data length (which is of course zero for such
223 subpatterns). This could cause incorrect "no match" results.
224
225 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
226 the pattern is reset in the first branch, pcre_compile() failed with
227 "internal error: code overflow at offset...". This happened only when
228 the reset was to the original external option setting. (An optimization
229 abstracts leading options settings into an external setting, which was the
230 cause of this.)
231
232 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
233 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
234 assertion pattern did not match (meaning that the assertion was true), it
235 was incorrectly treated as false if the SKIP had been reached during the
236 matching. This also applied to assertions used as conditions.
237
238 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
239 assertion subpattern, including such a pattern used as a condition,
240 unpredictable results occurred, instead of the error return
241 PCRE_ERROR_DFA_UITEM.
242
243 5. The C++ GlobalReplace function was not working like Perl for the special
244 situation when an empty string is matched. It now does the fancy magic
245 stuff that is necessary.
246
247 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
248 removed. (These were left over from very, very early versions of PCRE.)
249
250 7. Some cosmetic changes to the code to make life easier when compiling it
251 as part of something else:
252
253 (a) Change DEBUG to PCRE_DEBUG.
254
255 (b) In pcre_compile(), rename the member of the "branch_chain" structure
256 called "current" as "current_branch", to prevent a collision with the
257 Linux macro when compiled as a kernel module.
258
259 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
260 prevent a collision with the Linux macro when compiled as a kernel
261 module.
262
263 8. In pcre_compile() there are some checks for integer overflows that used to
264 cast potentially large values to (double). This has been changed to that
265 when building, a check for int64_t is made, and if it is found, it is used
266 instead, thus avoiding the use of floating point arithmetic. (There is no
267 other use of FP in PCRE.) If int64_t is not found, the fallback is to
268 double.
269
270 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
271 2005 (difference between two addresses compared to an unsigned value).
272
273 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
274 custom one, because of the following reported problem in Windows:
275
276 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
277 under Win32.
278 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
279 therefore missing the function definition.
280 - The compiler thus generates a "C" signature for the test function.
281 - The linker fails to find the "C" function.
282 - PCRE fails to configure if asked to do so against libbz2.
283
284 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
285 messages were output:
286
287 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
288 rerunning libtoolize, to keep the correct libtool macros in-tree.
289 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
290
291 I have done both of these things.
292
293 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
294 most of the time, it *can* run out if it is given a pattern that contains a
295 runaway infinite recursion. I updated the discussion in the pcrestack man
296 page.
297
298 13. Now that we have gone to the x.xx style of version numbers, the minor
299 version may start with zero. Using 08 or 09 is a bad idea because users
300 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
301 interpreted as invalid octal numbers. I've updated the previous comment in
302 configure.ac, and also added a check that gives an error if 08 or 09 are
303 used.
304
305 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
306 causing partial matching to fail when the end of the subject matched \W
307 in a UTF-8 pattern where \W was quantified with a minimum of 3.
308
309 15. There were some discrepancies between the declarations in pcre_internal.h
310 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
311 their definitions. The declarations used "const uschar *" and the
312 definitions used USPTR. Even though USPTR is normally defined as "const
313 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
314 reported that: "This difference in casting confuses some C++ compilers, for
315 example, SunCC recognizes above declarations as different functions and
316 generates broken code for hbpcre." I have changed the declarations to use
317 USPTR.
318
319 16. GNU libtool is named differently on some systems. The autogen.sh script now
320 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
321 (FreeBSD).
322
323 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
324 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
325 comment: "Figure out how to create a longlong from a string: strtoll and
326 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
327 instance, but it only takes 2 args instead of 3!"
328
329 18. A subtle bug concerned with back references has been fixed by a change of
330 specification, with a corresponding code fix. A pattern such as
331 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
332 refers, was giving matches when it shouldn't. For example, xa=xaaa would
333 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
334 same bug. Such groups have to be quantified to be useful, or contained
335 inside another quantified group. (If there's no repetition, the reference
336 can never match.) The problem arises because, having left the group and
337 moved on to the rest of the pattern, a later failure that backtracks into
338 the group uses the captured value from the final iteration of the group
339 rather than the correct earlier one. I have fixed this in PCRE by forcing
340 any group that contains a reference to itself to be an atomic group; that
341 is, there cannot be any backtracking into it once it has completed. This is
342 similar to recursive and subroutine calls.
343
344
345 Version 8.00 19-Oct-09
346 ----------------------
347
348 1. The table for translating pcre_compile() error codes into POSIX error codes
349 was out-of-date, and there was no check on the pcre_compile() error code
350 being within the table. This could lead to an OK return being given in
351 error.
352
353 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
354 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
355 in a Windows environment.
356
357 3. The pcregrep --count option prints the count for each file even when it is
358 zero, as does GNU grep. However, pcregrep was also printing all files when
359 --files-with-matches was added. Now, when both options are given, it prints
360 counts only for those files that have at least one match. (GNU grep just
361 prints the file name in this circumstance, but including the count seems
362 more useful - otherwise, why use --count?) Also ensured that the
363 combination -clh just lists non-zero counts, with no names.
364
365 4. The long form of the pcregrep -F option was incorrectly implemented as
366 --fixed_strings instead of --fixed-strings. This is an incompatible change,
367 but it seems right to fix it, and I didn't think it was worth preserving
368 the old behaviour.
369
370 5. The command line items --regex=pattern and --regexp=pattern were not
371 recognized by pcregrep, which required --regex pattern or --regexp pattern
372 (with a space rather than an '='). The man page documented the '=' forms,
373 which are compatible with GNU grep; these now work.
374
375 6. No libpcreposix.pc file was created for pkg-config; there was just
376 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
377
378 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
379 when UCP support is not needed, by modifying the Python script that
380 generates it from Unicode data files. This should not matter if the module
381 is correctly used as a library, but I received one complaint about 50K of
382 unwanted data. My guess is that the person linked everything into his
383 program rather than using a library. Anyway, it does no harm.
384
385 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
386 was a minimum greater than 1 for a wide character in a possessive
387 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
388 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
389 character. Chaos in the form of incorrect output or a compiling loop could
390 result.
391
392 9. The restrictions on what a pattern can contain when partial matching is
393 requested for pcre_exec() have been removed. All patterns can now be
394 partially matched by this function. In addition, if there are at least two
395 slots in the offset vector, the offset of the earliest inspected character
396 for the match and the offset of the end of the subject are set in them when
397 PCRE_ERROR_PARTIAL is returned.
398
399 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
400 synonymous with PCRE_PARTIAL, for backwards compatibility, and
401 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
402 and may be more useful for multi-segment matching.
403
404 11. Partial matching with pcre_exec() is now more intuitive. A partial match
405 used to be given if ever the end of the subject was reached; now it is
406 given only if matching could not proceed because another character was
407 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
408 string "Z", which now yields "no match" instead of "partial match". In the
409 case of pcre_dfa_exec(), "no match" is given if every matching path for the
410 final character ended with (*FAIL).
411
412 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
413 if the pattern had a "must contain" character that was already found in the
414 earlier partial match, unless partial matching was again requested. For
415 example, with the pattern /dog.(body)?/, the "must contain" character is
416 "g". If the first part-match was for the string "dog", restarting with
417 "sbody" failed. This bug has been fixed.
418
419 13. The string returned by pcre_dfa_exec() after a partial match has been
420 changed so that it starts at the first inspected character rather than the
421 first character of the match. This makes a difference only if the pattern
422 starts with a lookbehind assertion or \b or \B (\K is not supported by
423 pcre_dfa_exec()). It's an incompatible change, but it makes the two
424 matching functions compatible, and I think it's the right thing to do.
425
426 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
427 so that the demonstration program is easily available in environments where
428 PCRE has not been installed from source.
429
430 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
431 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
432 library.
433
434 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
435 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
436 is not the first non-POSIX option to be added. Clearly some people find
437 these options useful.
438
439 17. If a caller to the POSIX matching function regexec() passes a non-zero
440 value for nmatch with a NULL value for pmatch, the value of
441 nmatch is forced to zero.
442
443 18. RunGrepTest did not have a test for the availability of the -u option of
444 the diff command, as RunTest does. It now checks in the same way as
445 RunTest, and also checks for the -b option.
446
447 19. If an odd number of negated classes containing just a single character
448 interposed, within parentheses, between a forward reference to a named
449 subpattern and the definition of the subpattern, compilation crashed with
450 an internal error, complaining that it could not find the referenced
451 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
452 [The bug was that it was starting one character too far in when skipping
453 over the character class, thus treating the ] as data rather than
454 terminating the class. This meant it could skip too much.]
455
456 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
457 /g option in pcretest when the pattern contains \K, which makes it possible
458 to have an empty string match not at the start, even when the pattern is
459 anchored. Updated pcretest and pcredemo to use this option.
460
461 21. If the maximum number of capturing subpatterns in a recursion was greater
462 than the maximum at the outer level, the higher number was returned, but
463 with unset values at the outer level. The correct (outer level) value is
464 now given.
465
466 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
467 PCRE did not set those parentheses (unlike Perl). I have now found a way to
468 make it do so. The string so far is captured, making this feature
469 compatible with Perl.
470
471 23. The tests have been re-organized, adding tests 11 and 12, to make it
472 possible to check the Perl 5.10 features against Perl 5.10.
473
474 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
475 pattern matches a fixed length string. PCRE did not allow this; now it
476 does. Neither allows recursion.
477
478 25. I finally figured out how to implement a request to provide the minimum
479 length of subject string that was needed in order to match a given pattern.
480 (It was back references and recursion that I had previously got hung up
481 on.) This code has now been added to pcre_study(); it finds a lower bound
482 to the length of subject needed. It is not necessarily the greatest lower
483 bound, but using it to avoid searching strings that are too short does give
484 some useful speed-ups. The value is available to calling programs via
485 pcre_fullinfo().
486
487 26. While implementing 25, I discovered to my embarrassment that pcretest had
488 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
489 study optimizations had never been tested with that matching function.
490 Oops. What is worse, even when it was passed study data, there was a bug in
491 pcre_dfa_exec() that meant it never actually used it. Double oops. There
492 were also very few tests of studied patterns with pcre_dfa_exec().
493
494 27. If (?| is used to create subpatterns with duplicate numbers, they are now
495 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
496 on the other side of the coin, they are no longer allowed to have different
497 names, because these cannot be distinguished in PCRE, and this has caused
498 confusion. (This is a difference from Perl.)
499
500 28. When duplicate subpattern names are present (necessarily with different
501 numbers, as required by 27 above), and a test is made by name in a
502 conditional pattern, either for a subpattern having been matched, or for
503 recursion in such a pattern, all the associated numbered subpatterns are
504 tested, and the overall condition is true if the condition is true for any
505 one of them. This is the way Perl works, and is also more like the way
506 testing by number works.
507
508
509 Version 7.9 11-Apr-09
510 ---------------------
511
512 1. When building with support for bzlib/zlib (pcregrep) and/or readline
513 (pcretest), all targets were linked against these libraries. This included
514 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
515 libraries. This caused unwanted dependencies to be created. This problem
516 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
517 pcretest is linked with readline.
518
519 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
520 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
521 moved outside it again, because FALSE and TRUE are already defined in AIX,
522 but BOOL is not.
523
524 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
525 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
526
527 4. The pcregrep documentation said spaces were inserted as well as colons (or
528 hyphens) following file names and line numbers when outputting matching
529 lines. This is not true; no spaces are inserted. I have also clarified the
530 wording for the --colour (or --color) option.
531
532 5. In pcregrep, when --colour was used with -o, the list of matching strings
533 was not coloured; this is different to GNU grep, so I have changed it to be
534 the same.
535
536 6. When --colo(u)r was used in pcregrep, only the first matching substring in
537 each matching line was coloured. Now it goes on to look for further matches
538 of any of the test patterns, which is the same behaviour as GNU grep.
539
540 7. A pattern that could match an empty string could cause pcregrep to loop; it
541 doesn't make sense to accept an empty string match in pcregrep, so I have
542 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
543 seems to be how GNU grep behaves.
544
545 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
546 start or after a newline", because the conditional assertion was not being
547 correctly handled. The rule now is that both the assertion and what follows
548 in the first alternative must satisfy the test.
549
550 9. If auto-callout was enabled in a pattern with a conditional group whose
551 condition was an assertion, PCRE could crash during matching, both with
552 pcre_exec() and pcre_dfa_exec().
553
554 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
555 used for matching.
556
557 11. Unicode property support in character classes was not working for
558 characters (bytes) greater than 127 when not in UTF-8 mode.
559
560 12. Added the -M command line option to pcretest.
561
562 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
563
564 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
565
566 16. Added comments and documentation about mis-use of no_arg in the C++
567 wrapper.
568
569 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
570 from Martin Jerabek that uses macro names for all relevant character and
571 string constants.
572
573 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
574 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
575 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
576 these, but not everybody uses configure.
577
578 19. A conditional group that had only one branch was not being correctly
579 recognized as an item that could match an empty string. This meant that an
580 enclosing group might also not be so recognized, causing infinite looping
581 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
582 with the subject "ab", where knowledge that the repeated group can match
583 nothing is needed in order to break the loop.
584
585 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
586 exec(), but without supplying a callout function, matching went wrong.
587
588 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
589 leak if the size of the offset vector was greater than 30. When the vector
590 is smaller, the saved offsets during recursion go onto a local stack
591 vector, but for larger vectors malloc() is used. It was failing to free
592 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
593 error, in fact).
594
595 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
596 heapframe that is used only when UTF-8 support is enabled. This caused no
597 problem, but was untidy.
598
599 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
600 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
601 included within another project.
602
603 24. Steven Van Ingelgem's patches to add more options to the CMake support,
604 slightly modified by me:
605
606 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
607 not building pcregrep.
608
609 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
610 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
611
612 25. Forward references, both numeric and by name, in patterns that made use of
613 duplicate group numbers, could behave incorrectly or give incorrect errors,
614 because when scanning forward to find the reference group, PCRE was not
615 taking into account the duplicate group numbers. A pattern such as
616 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
617
618 26. Changed a few more instances of "const unsigned char *" to USPTR, making
619 the feature of a custom pointer more persuasive (as requested by a user).
620
621 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
622 pcretest.c, inside #ifndefs, because it seems they are sometimes already
623 pre-defined.
624
625 28. Added support for (*UTF8) at the start of a pattern.
626
627 29. Arrange for flags added by the "release type" setting in CMake to be shown
628 in the configuration summary.
629
630
631 Version 7.8 05-Sep-08
632 ---------------------
633
634 1. Replaced UCP searching code with optimized version as implemented for Ad
635 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
636 stage table and inline lookup instead of a function, giving speed ups of 2
637 to 5 times on some simple patterns that I tested. Permission was given to
638 distribute the MultiStage2.py script that generates the tables (it's not in
639 the tarball, but is in the Subversion repository).
640
641 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
642 scripts.
643
644 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
645 a group with a zero qualifier. The result of the study could be incorrect,
646 or the function might crash, depending on the pattern.
647
648 4. Caseless matching was not working for non-ASCII characters in back
649 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
650 It now works when Unicode Property Support is available.
651
652 5. In pcretest, an escape such as \x{de} in the data was always generating
653 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
654 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
655 truncation.
656
657 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
658
659 7. Added two (int) casts to pcregrep when printing the difference of two
660 pointers, in case they are 64-bit values.
661
662 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
663 test 2 if it fails.
664
665 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
666 and a #define of that name to empty if it is not externally set. This is to
667 allow users of MSVC to set it if necessary.
668
669 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
670 the convenience functions in the pcre_get.c source file.
671
672 11. An option change at the start of a pattern that had top-level alternatives
673 could cause overwriting and/or a crash. This command provoked a crash in
674 some environments:
675
676 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
677
678 This potential security problem was recorded as CVE-2008-2371.
679
680 12. For a pattern where the match had to start at the beginning or immediately
681 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
682 pcre_dfa_exec() could read past the end of the passed subject if there was
683 no match. To help with detecting such bugs (e.g. with valgrind), I modified
684 pcretest so that it places the subject at the end of its malloc-ed buffer.
685
686 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
687 exec() might read past the end of the data buffer in UTF-8 mode.
688
689 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
690 the data contained the byte 0x85 as part of a UTF-8 character within its
691 first line. This applied both to normal and DFA matching.
692
693 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
694 /^[^d]*?$/8 failed to match "abc".
695
696 16. Added a missing copyright notice to pcrecpp_internal.h.
697
698 17. Make it more clear in the documentation that values returned from
699 pcre_exec() in ovector are byte offsets, not character counts.
700
701 18. Tidied a few places to stop certain compilers from issuing warnings.
702
703 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
704 supplied by Stefan Weber. I made a further small update for 7.8 because
705 there is a change of source arrangements: the pcre_searchfuncs.c module is
706 replaced by pcre_ucd.c.
707
708
709 Version 7.7 07-May-08
710 ---------------------
711
712 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
713 a string to a long long, pretend we don't even have a long long." This is
714 done by checking for the strtoq, strtoll, and _strtoi64 functions.
715
716 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
717 pre-7.6 versions, which defined a global no_arg variable instead of putting
718 it in the RE class. (See also #8 below.)
719
720 3. Remove a line of dead code, identified by coverity and reported by Nuno
721 Lopes.
722
723 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
724
725 (1) The include/exclude patterns were being applied to the whole pathnames
726 of files, instead of just to the final components.
727
728 (2) If there was more than one level of directory, the subdirectories were
729 skipped unless they satisfied the include/exclude conditions. This is
730 inconsistent with GNU grep (and could even be seen as contrary to the
731 pcregrep specification - which I improved to make it absolutely clear).
732 The action now is always to scan all levels of directory, and just
733 apply the include/exclude patterns to regular files.
734
735 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
736 --exclude_dir in the tests to avoid scanning .svn directories.
737
738 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
739 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
740 doesn't support NULs in patterns.
741
742 7. Added some missing "const"s to declarations of static tables in
743 pcre_compile.c and pcre_dfa_exec.c.
744
745 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
746 caused by fix #2 above. (Subsequently also a second patch to fix the
747 first patch. And a third patch - this was a messy problem.)
748
749 9. Applied Craig's patch to remove the use of push_back().
750
751 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
752 matching function regexec().
753
754 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
755 which, however, unlike Perl's \g{...}, are subroutine calls, not back
756 references. PCRE supports relative numbers with this syntax (I don't think
757 Oniguruma does).
758
759 12. Previously, a group with a zero repeat such as (...){0} was completely
760 omitted from the compiled regex. However, this means that if the group
761 was called as a subroutine from elsewhere in the pattern, things went wrong
762 (an internal error was given). Such groups are now left in the compiled
763 pattern, with a new opcode that causes them to be skipped at execution
764 time.
765
766 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
767 to the way PCRE behaves:
768
769 (a) A lone ] character is dis-allowed (Perl treats it as data).
770
771 (b) A back reference to an unmatched subpattern matches an empty string
772 (Perl fails the current match path).
773
774 (c) A data ] in a character class must be notated as \] because if the
775 first data character in a class is ], it defines an empty class. (In
776 Perl it is not possible to have an empty class.) The empty class []
777 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
778 The negative empty class [^] matches any one character, independently
779 of the DOTALL setting.
780
781 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
782 non-existent subpattern following a character class starting with ']' and
783 containing () gave an internal compiling error instead of "reference to
784 non-existent subpattern". Fortunately, when the pattern did exist, the
785 compiled code was correct. (When scanning forwards to check for the
786 existencd of the subpattern, it was treating the data ']' as terminating
787 the class, so got the count wrong. When actually compiling, the reference
788 was subsequently set up correctly.)
789
790 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
791 it was being rejected as not supported by pcre_dfa_exec(), even though
792 other assertions are supported. I have made pcre_dfa_exec() support
793 (*FAIL).
794
795 16. The implementation of 13c above involved the invention of a new opcode,
796 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
797 cannot be changed at match time, I realized I could make a small
798 improvement to matching performance by compiling OP_ALLANY instead of
799 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
800 on the OP_ANY path.
801
802 17. Compiling pcretest on Windows with readline support failed without the
803 following two fixes: (1) Make the unistd.h include conditional on
804 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
805
806 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
807 ncurses library to be included for pcretest when ReadLine support is
808 requested, but also to allow for it to be overridden. This patch came from
809 Daniel Bergström.
810
811 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
812 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
813 any errors with the current Unicode tables. Thanks to Peter Kankowski for
814 spotting this.
815
816
817 Version 7.6 28-Jan-08
818 ---------------------
819
820 1. A character class containing a very large number of characters with
821 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
822 overflow.
823
824 2. Patch to cut out the "long long" test in pcrecpp_unittest when
825 HAVE_LONG_LONG is not defined.
826
827 3. Applied Christian Ehrlicher's patch to update the CMake build files to
828 bring them up to date and include new features. This patch includes:
829
830 - Fixed PH's badly added libz and libbz2 support.
831 - Fixed a problem with static linking.
832 - Added pcredemo. [But later removed - see 7 below.]
833 - Fixed dftables problem and added an option.
834 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
835 HAVE_LONG_LONG.
836 - Added readline support for pcretest.
837 - Added an listing of the option settings after cmake has run.
838
839 4. A user submitted a patch to Makefile that makes it easy to create
840 "pcre.dll" under mingw when using Configure/Make. I added stuff to
841 Makefile.am that cause it to include this special target, without
842 affecting anything else. Note that the same mingw target plus all
843 the other distribution libraries and programs are now supported
844 when configuring with CMake (see 6 below) instead of with
845 Configure/Make.
846
847 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
848 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
849 exported in the Windows port". It has not yet been confirmed that the patch
850 solves the problem, but it does no harm.
851
852 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
853 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
854 with CMake, and also correct the comment about stack recursion.
855
856 7. Remove the automatic building of pcredemo from the ./configure system and
857 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
858 of a program that users should build themselves after PCRE is installed, so
859 building it automatically is not really right. What is more, it gave
860 trouble in some build environments.
861
862 8. Further tidies to CMakeLists.txt from Sheri and Christian.
863
864
865 Version 7.5 10-Jan-08
866 ---------------------
867
868 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
869 values in parens when parsing an RE using the C++ wrapper."
870
871 2. Negative specials like \S did not work in character classes in UTF-8 mode.
872 Characters greater than 255 were excluded from the class instead of being
873 included.
874
875 3. The same bug as (2) above applied to negated POSIX classes such as
876 [:^space:].
877
878 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
879 defined or documented. It seems to have been a typo for PCRE_STATIC, so
880 I have changed it.
881
882 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
883 first named subpattern) and a construct such as (?&a) would reference the
884 first named subpattern whose name started with "a" (in other words, the
885 length check was missing). Both these problems are fixed. "Subpattern name
886 expected" is now given for (?&) (a zero-length name), and this patch also
887 makes it give the same error for \k'' (previously it complained that that
888 was a reference to a non-existent subpattern).
889
890 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
891 this is right because (?- can be followed by option settings as well as by
892 digits. I have, however, made the messages clearer.
893
894 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
895 than the number used in the conditional) now cause a compile-time error.
896 This is actually not compatible with Perl, which accepts such patterns, but
897 treats the conditional as always being FALSE (as PCRE used to), but it
898 seems to me that giving a diagnostic is better.
899
900 8. Change "alphameric" to the more common word "alphanumeric" in comments
901 and messages.
902
903 9. Fix two occurrences of "backslash" in comments that should have been
904 "backspace".
905
906 10. Remove two redundant lines of code that can never be obeyed (their function
907 was moved elsewhere).
908
909 11. The program that makes PCRE's Unicode character property table had a bug
910 which caused it to generate incorrect table entries for sequences of
911 characters that have the same character type, but are in different scripts.
912 It amalgamated them into a single range, with the script of the first of
913 them. In other words, some characters were in the wrong script. There were
914 thirteen such cases, affecting characters in the following ranges:
915
916 U+002b0 - U+002c1
917 U+0060c - U+0060d
918 U+0061e - U+00612
919 U+0064b - U+0065e
920 U+0074d - U+0076d
921 U+01800 - U+01805
922 U+01d00 - U+01d77
923 U+01d9b - U+01dbf
924 U+0200b - U+0200f
925 U+030fc - U+030fe
926 U+03260 - U+0327f
927 U+0fb46 - U+0fbb1
928 U+10450 - U+1049d
929
930 12. The -o option (show only the matching part of a line) for pcregrep was not
931 compatible with GNU grep in that, if there was more than one match in a
932 line, it showed only the first of them. It now behaves in the same way as
933 GNU grep.
934
935 13. If the -o and -v options were combined for pcregrep, it printed a blank
936 line for every non-matching line. GNU grep prints nothing, and pcregrep now
937 does the same. The return code can be used to tell if there were any
938 non-matching lines.
939
940 14. Added --file-offsets and --line-offsets to pcregrep.
941
942 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
943 infinitely looping recursion. The bug was that positive lookaheads were not
944 being skipped when checking for a possible empty match (negative lookaheads
945 and both kinds of lookbehind were skipped).
946
947 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
948 inclusion of <windows.h> to before rather than after the definition of
949 INVALID_FILE_ATTRIBUTES (patch from David Byron).
950
951 17. Specifying a possessive quantifier with a specific limit for a Unicode
952 character property caused pcre_compile() to compile bad code, which led at
953 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
954 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
955 caused the error; without that there was no problem.
956
957 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
958
959 19. Added --enable-pcretest-libreadline.
960
961 20. In pcrecpp.cc, the variable 'count' was incremented twice in
962 RE::GlobalReplace(). As a result, the number of replacements returned was
963 double what it should be. I removed one of the increments, but Craig sent a
964 later patch that removed the other one (the right fix) and added unit tests
965 that check the return values (which was not done before).
966
967 21. Several CMake things:
968
969 (1) Arranged that, when cmake is used on Unix, the libraries end up with
970 the names libpcre and libpcreposix, not just pcre and pcreposix.
971
972 (2) The above change means that pcretest and pcregrep are now correctly
973 linked with the newly-built libraries, not previously installed ones.
974
975 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
976
977 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
978 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
979 UTF-8 newline character). The key issue is that the pattern starts .*;
980 this means that the match must be either at the beginning, or after a
981 newline. The bug was in the code for advancing after a failed match and
982 checking that the new position followed a newline. It was not taking
983 account of UTF-8 characters correctly.
984
985 23. PCRE was behaving differently from Perl in the way it recognized POSIX
986 character classes. PCRE was not treating the sequence [:...:] as a
987 character class unless the ... were all letters. Perl, however, seems to
988 allow any characters between [: and :], though of course it rejects as
989 unknown any "names" that contain non-letters, because all the known class
990 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
991 for example, whereas PCRE did not - it did not recognize a POSIX character
992 class. This seemed a bit dangerous, so the code has been changed to be
993 closer to Perl. The behaviour is not identical to Perl, because PCRE will
994 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
995 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
996 Perl does, and where it didn't before.
997
998 24. Rewrite so as to remove the single use of %n from pcregrep because in some
999 Windows environments %n is disabled by default.
1000
1001
1002 Version 7.4 21-Sep-07
1003 ---------------------
1004
1005 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1006 means that a class such as [\s] counted as "explicit reference to CR or
1007 LF". That isn't really right - the whole point of the change was to try to
1008 help when there was an actual mention of one of the two characters. So now
1009 the change happens only if \r or \n (or a literal CR or LF) character is
1010 encountered.
1011
1012 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1013 of both had grown to the point where there were only 3 bits left.
1014 Fortunately, there was spare space in the data structure, and so I have
1015 moved the internal flags into a new 16-bit field to free up more option
1016 bits.
1017
1018 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1019 but did not set the internal JCHANGED flag - either of these is enough to
1020 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1021 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1022 start sets both bits.
1023
1024 4. Added options (at build time, compile time, exec time) to change \R from
1025 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1026
1027 5. doc/pcresyntax.html was missing from the distribution.
1028
1029 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1030 compatibility, even though it is no longer used.
1031
1032 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1033 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1034 windows.h file is present (where different names are used). [This was
1035 reversed later after testing - see 16 below.]
1036
1037 8. Changed all #include <config.h> to #include "config.h". There were also
1038 some further <pcre.h> cases that I changed to "pcre.h".
1039
1040 9. When pcregrep was used with the --colour option, it missed the line ending
1041 sequence off the lines that it output.
1042
1043 10. It was pointed out to me that arrays of string pointers cause lots of
1044 relocations when a shared library is dynamically loaded. A technique of
1045 using a single long string with a table of offsets can drastically reduce
1046 these. I have refactored PCRE in four places to do this. The result is
1047 dramatic:
1048
1049 Originally: 290
1050 After changing UCP table: 187
1051 After changing error message table: 43
1052 After changing table of "verbs" 36
1053 After changing table of Posix names 22
1054
1055 Thanks to the folks working on Gregex for glib for this insight.
1056
1057 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1058 unicode-properties was also set.
1059
1060 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1061
1062 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1063 checked only for CRLF.
1064
1065 14. Added casts to pcretest.c to avoid compiler warnings.
1066
1067 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1068
1069 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1070 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1071 entirely. This removes changes made in 7 above.
1072
1073 17. The CMake files have been updated, and there is now more information about
1074 building with CMake in the NON-UNIX-USE document.
1075
1076
1077 Version 7.3 28-Aug-07
1078 ---------------------
1079
1080 1. In the rejigging of the build system that eventually resulted in 7.1, the
1081 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1082 brackets there is not right, since it causes compilers to look for an
1083 installed pcre.h, not the version that is in the source that is being
1084 compiled (which of course may be different). I have changed it back to:
1085
1086 #include "pcre.h"
1087
1088 I have a vague recollection that the change was concerned with compiling in
1089 different directories, but in the new build system, that is taken care of
1090 by the VPATH setting the Makefile.
1091
1092 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1093 when the subject happened to end in the byte 0x85 (e.g. if the last
1094 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1095 characters but of course it shouldn't be taken as a newline when it is part
1096 of another character. The bug was that, for an unlimited repeat of . in
1097 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1098 characters when looking for a newline.
1099
1100 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1101
1102 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1103 in debug output.
1104
1105 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1106 long printing in the pcrecpp unittest when running under MinGW.
1107
1108 6. ESC_K was left out of the EBCDIC table.
1109
1110 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1111 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1112 limit also applies to "virtual nesting" when a pattern is recursive, and in
1113 this case 1000 isn't so big. I have been able to remove this limit at the
1114 expense of backing off one optimization in certain circumstances. Normally,
1115 when pcre_exec() would call its internal match() function recursively and
1116 immediately return the result unconditionally, it uses a "tail recursion"
1117 feature to save stack. However, when a subpattern that can match an empty
1118 string has an unlimited repetition quantifier, it no longer makes this
1119 optimization. That gives it a stack frame in which to save the data for
1120 checking that an empty string has been matched. Previously this was taken
1121 from the 1000-entry workspace that had been reserved. So now there is no
1122 explicit limit, but more stack is used.
1123
1124 8. Applied Daniel's patches to solve problems with the import/export magic
1125 syntax that is required for Windows, and which was going wrong for the
1126 pcreposix and pcrecpp parts of the library. These were overlooked when this
1127 problem was solved for the main library.
1128
1129 9. There were some crude static tests to avoid integer overflow when computing
1130 the size of patterns that contain repeated groups with explicit upper
1131 limits. As the maximum quantifier is 65535, the maximum group length was
1132 set at 30,000 so that the product of these two numbers did not overflow a
1133 32-bit integer. However, it turns out that people want to use groups that
1134 are longer than 30,000 bytes (though not repeat them that many times).
1135 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1136 made it possible to implement the integer overflow checks in a much more
1137 dynamic way, which I have now done. The artificial limitation on group
1138 length has been removed - we now have only the limit on the total length of
1139 the compiled pattern, which depends on the LINK_SIZE setting.
1140
1141 10. Fixed a bug in the documentation for get/copy named substring when
1142 duplicate names are permitted. If none of the named substrings are set, the
1143 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1144 empty string.
1145
1146 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1147 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1148 because the ] is interpreted as the first data character and the
1149 terminating ] is not found. PCRE has been made compatible with Perl in this
1150 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1151 cause memory overwriting.
1152
1153 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1154 string has been matched (to stop an infinite loop). It was not recognizing
1155 a conditional subpattern that could match an empty string if that
1156 subpattern was within another subpattern. For example, it looped when
1157 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1158 condition was not nested. This bug has been fixed.
1159
1160 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1161 past the start of the subject in the presence of bytes with the top bit
1162 set, for example "\x8aBCD".
1163
1164 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1165 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1166
1167 14. Optimized (?!) to (*FAIL).
1168
1169 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1170 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1171 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1172 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1173 does: it's just the validity check that is more restrictive.
1174
1175 16. Inserted checks for integer overflows during escape sequence (backslash)
1176 processing, and also fixed erroneous offset values for syntax errors during
1177 backslash processing.
1178
1179 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1180 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1181
1182 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1183 caused an overrun.
1184
1185 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1186 something other than just ASCII characters) inside a group that had an
1187 unlimited repeat caused a loop at compile time (while checking to see
1188 whether the group could match an empty string).
1189
1190 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1191 [\P{Any}] did so. (Error in the code for printing property names.)
1192
1193 21. An orphan \E inside a character class could cause a crash.
1194
1195 22. A repeated capturing bracket such as (A)? could cause a wild memory
1196 reference during compilation.
1197
1198 23. There are several functions in pcre_compile() that scan along a compiled
1199 expression for various reasons (e.g. to see if it's fixed length for look
1200 behind). There were bugs in these functions when a repeated \p or \P was
1201 present in the pattern. These operators have additional parameters compared
1202 with \d, etc, and these were not being taken into account when moving along
1203 the compiled data. Specifically:
1204
1205 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1206 length.
1207
1208 (b) An item such as \pL+ within a repeated group could cause crashes or
1209 loops.
1210
1211 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1212 "reference to non-existent subpattern" error.
1213
1214 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1215
1216 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1217 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1218
1219 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1220
1221 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1222 character were causing crashes (broken optimization).
1223
1224 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1225 \p or \P) caused a compile-time loop.
1226
1227 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1228 break. For example, the unstudied pattern [\r\n]A does not match the string
1229 "\r\nA" because change 7.0/46 below moves the current point on by two
1230 characters after failing to match at the start. However, the pattern \nA
1231 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1232 the same is true. There doesn't seem any very clean way out of this, but
1233 what I have chosen to do makes the common cases work: PCRE now takes note
1234 of whether there can be an explicit match for \r or \n anywhere in the
1235 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1236 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1237 pattern has explicit CR or LF references.
1238
1239 29. Added (*CR) etc for changing newline setting at start of pattern.
1240
1241
1242 Version 7.2 19-Jun-07
1243 ---------------------
1244
1245 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1246 which is apparently normally available under Windows.
1247
1248 2. Re-jig the pcregrep tests with different newline settings in an attempt
1249 to make them independent of the local environment's newline setting.
1250
1251 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1252
1253 4. Some of the "internals" tests were previously cut out when the link size
1254 was not 2, because the output contained actual offsets. The recent new
1255 "Z" feature of pcretest means that these can be cut out, making the tests
1256 usable with all link sizes.
1257
1258 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1259 stack recursion. This gives a massive performance boost under BSD, but just
1260 a small improvement under Linux. However, it saves one field in the frame
1261 in all cases.
1262
1263 6. Added more features from the forthcoming Perl 5.10:
1264
1265 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1266 recursion call. It refers to the nth most recently opened parentheses.
1267
1268 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1269 to be opened parentheses.
1270
1271 (c) Conditions that refer to capturing parentheses can be specified
1272 relatively, for example, (?(-2)... or (?(+3)...
1273
1274 (d) \K resets the start of the current match so that everything before
1275 is not part of it.
1276
1277 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1278
1279 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1280 reference syntax.
1281
1282 (g) (?| introduces a group in which the numbering of parentheses in each
1283 alternative starts with the same number.
1284
1285 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1286
1287 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1288 PCRE_INFO_JCHANGED.
1289
1290 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1291 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1292 for detecting groups that can match an empty string.
1293
1294 9. A pattern with a very large number of alternatives (more than several
1295 hundred) was running out of internal workspace during the pre-compile
1296 phase, where pcre_compile() figures out how much memory will be needed. A
1297 bit of new cunning has reduced the workspace needed for groups with
1298 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1299 workspace instead of running out of the 4096 that are available.
1300
1301 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1302
1303 11. Applied patch from Google to remove an optimization that didn't quite work.
1304 The report of the bug said:
1305
1306 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1307 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1308 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1309
1310 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1311 it matched the wrong number of bytes.
1312
1313
1314 Version 7.1 24-Apr-07
1315 ---------------------
1316
1317 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1318 that is more "standard", making use of automake and other Autotools. There
1319 is some re-arrangement of the files and adjustment of comments consequent
1320 on this.
1321
1322 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1323 for recursive directory scanning broke on some systems because the files
1324 are not scanned in any specific order and on different systems the order
1325 was different. A call to "sort" has been inserted into RunGrepTest for the
1326 approprate test as a short-term fix. In the longer term there may be an
1327 alternative.
1328
1329 3. I had an email from Eric Raymond about problems translating some of PCRE's
1330 man pages to HTML (despite the fact that I distribute HTML pages, some
1331 people do their own conversions for various reasons). The problems
1332 concerned the use of low-level troff macros .br and .in. I have therefore
1333 removed all such uses from the man pages (some were redundant, some could
1334 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1335 HTML has been updated to handle .nf/.fi and to complain if it encounters
1336 .br or .in.
1337
1338 4. Updated comments in configure.ac that get placed in config.h.in and also
1339 arranged for config.h to be included in the distribution, with the name
1340 config.h.generic, for the benefit of those who have to compile without
1341 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1342
1343 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1344 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1345 makevp.bat for the current PCRE, using the additional files
1346 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1347
1348 6. A Windows user reported a minor discrepancy with test 2, which turned out
1349 to be caused by a trailing space on an input line that had got lost in his
1350 copy. The trailing space was an accident, so I've just removed it.
1351
1352 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1353 that is needed.
1354
1355 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1356 as "const" (a) because they are and (b) because it helps the PHP
1357 maintainers who have recently made a script to detect big data structures
1358 in the php code that should be moved to the .rodata section. I remembered
1359 to update Builducptable as well, so it won't revert if ucptable.h is ever
1360 re-created.
1361
1362 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1363 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1364 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1365 support is not required. This saves 1.5-2K of code, which is important in
1366 some applications.
1367
1368 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1369 so as not to refer to the tables, even though these functions will never be
1370 called when UTF-8 support is disabled. Otherwise there are problems with a
1371 shared library.
1372
1373 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1374
1375 (a) It was defining its arguments as char * instead of void *.
1376
1377 (b) It was assuming that all moves were upwards in memory; this was true
1378 a long time ago when I wrote it, but is no longer the case.
1379
1380 The emulated memove() is provided for those environments that have neither
1381 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1382 is clearly not the case, as these two bugs were recently reported.
1383
1384 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1385 and Detrail to create the HTML documentation, the .txt form of the man
1386 pages, and it removes trailing spaces from listed files. It also creates
1387 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1388 case, it wraps all the #defines with #ifndefs. This script should be run
1389 before "make dist".
1390
1391 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1392 with Unicode property support.
1393
1394 (a) For a maximizing quantifier, if the two different cases of the
1395 character were of different lengths in their UTF-8 codings (there are
1396 some cases like this - I found 11), and the matching function had to
1397 back up over a mixture of the two cases, it incorrectly assumed they
1398 were both the same length.
1399
1400 (b) When PCRE was configured to use the heap rather than the stack for
1401 recursion during matching, it was not correctly preserving the data for
1402 the other case of a UTF-8 character when checking ahead for a match
1403 while processing a minimizing repeat. If the check also involved
1404 matching a wide character, but failed, corruption could cause an
1405 erroneous result when trying to check for a repeat of the original
1406 character.
1407
1408 13. Some tidying changes to the testing mechanism:
1409
1410 (a) The RunTest script now detects the internal link size and whether there
1411 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1412 values substituted by "configure". (The RunGrepTest script already did
1413 this for UTF-8.) The configure.ac script no longer substitutes the
1414 relevant variables.
1415
1416 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1417 with length and offset values. This means that the output is different
1418 for different internal link sizes. Test 2 is skipped for link sizes
1419 other than 2 because of this, bypassing the problem. Unfortunately,
1420 there was also a test in test 3 (the locale tests) that used /B and
1421 failed for link sizes other than 2. Rather than cut the whole test out,
1422 I have added a new /Z option to pcretest that replaces the length and
1423 offset values with spaces. This is now used to make test 3 independent
1424 of link size. (Test 2 will be tidied up later.)
1425
1426 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1427 segmentation fault instead of returning the appropriate error message.
1428
1429 15. In multiline mode when the newline sequence was set to "any", the pattern
1430 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1431 This doesn't seem right; it now treats the CRLF combination as the line
1432 ending, and so does not match in that case. It's only a pattern such as ^$
1433 that would hit this one: something like ^ABC$ would have failed after \r
1434 and then tried again after \r\n.
1435
1436 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1437 in an attempt to make files that differ only in their line terminators
1438 compare equal. This works on Linux.
1439
1440 17. Under certain error circumstances pcregrep might try to free random memory
1441 as it exited. This is now fixed, thanks to valgrind.
1442
1443 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1444 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1445 was because its rules for how to advance for /g after matching an empty
1446 string at the end of a line did not allow for this case. They now check for
1447 it specially.
1448
1449 20. pcretest is supposed to handle patterns and data of any length, by
1450 extending its buffers when necessary. It was getting this wrong when the
1451 buffer for a data line had to be extended.
1452
1453 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1454 CRLF as a newline sequence.
1455
1456 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1457 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1458 I have nevertheless tidied it up.
1459
1460 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1461
1462 24. Added a man page for pcre-config.
1463
1464
1465 Version 7.0 19-Dec-06
1466 ---------------------
1467
1468 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1469 moving to gcc 4.1.1.
1470
1471 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1472 sys/time.h, which is documented as needed for this function. It doesn't
1473 seem to matter on Linux, but it showed up on some releases of OS X.
1474
1475 3. It seems that there are systems where bytes whose values are greater than
1476 127 match isprint() in the "C" locale. The "C" locale should be the
1477 default when a C program starts up. In most systems, only ASCII printing
1478 characters match isprint(). This difference caused the output from pcretest
1479 to vary, making some of the tests fail. I have changed pcretest so that:
1480
1481 (a) When it is outputting text in the compiled version of a pattern, bytes
1482 other than 32-126 are always shown as hex escapes.
1483
1484 (b) When it is outputting text that is a matched part of a subject string,
1485 it does the same, unless a different locale has been set for the match
1486 (using the /L modifier). In this case, it uses isprint() to decide.
1487
1488 4. Fixed a major bug that caused incorrect computation of the amount of memory
1489 required for a compiled pattern when options that changed within the
1490 pattern affected the logic of the preliminary scan that determines the
1491 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1492 that the computed length was too small. The symptoms of this bug were
1493 either the PCRE error "internal error: code overflow" from pcre_compile(),
1494 or a glibc crash with a message such as "pcretest: free(): invalid next
1495 size (fast)". Examples of patterns that provoked this bug (shown in
1496 pcretest format) are:
1497
1498 /(?-x: )/x
1499 /(?x)(?-x: \s*#\s*)/
1500 /((?i)[\x{c0}])/8
1501 /(?i:[\x{c0}])/8
1502
1503 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1504 is now done differently.
1505
1506 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1507 wrapper classes; (b) implement a new function in the C++ scanner that is
1508 more efficient than the old way of doing things because it avoids levels of
1509 recursion in the regex matching; (c) add a paragraph to the documentation
1510 for the FullMatch() function.
1511
1512 6. The escape sequence \n was being treated as whatever was defined as
1513 "newline". Not only was this contrary to the documentation, which states
1514 that \n is character 10 (hex 0A), but it also went horribly wrong when
1515 "newline" was defined as CRLF. This has been fixed.
1516
1517 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1518 was being set to -1 for the "end of line" case (supposedly a value that no
1519 character can have). Though this value is never used (the check for end of
1520 line is "zero bytes in current character"), it caused compiler complaints.
1521 I've changed it to 0xffffffff.
1522
1523 8. In pcre_version.c, the version string was being built by a sequence of
1524 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1525 string (as it is for production releases) called a macro with an empty
1526 argument. The C standard says the result of this is undefined. The gcc
1527 compiler treats it as an empty string (which was what was wanted) but it is
1528 reported that Visual C gives an error. The source has been hacked around to
1529 avoid this problem.
1530
1531 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1532 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1533 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1534 of them did).
1535
1536 10. Originally, pcretest opened its input and output without "b"; then I was
1537 told that "b" was needed in some environments, so it was added for release
1538 5.0 to both the input and output. (It makes no difference on Unix-like
1539 systems.) Later I was told that it is wrong for the input on Windows. I've
1540 now abstracted the modes into two macros, to make it easier to fiddle with
1541 them, and removed "b" from the input mode under Windows.
1542
1543 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1544
1545 12. Added -help and --help to pcretest as an official way of being reminded
1546 of the options.
1547
1548 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1549 and pcrecpp.cc because they annoy compilers at high warning levels.
1550
1551 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1552
1553 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1554 scripts are not C programs :-) and which was not noticed because it works
1555 on Linux.
1556
1557 16. pcretest is supposed to handle any length of pattern and data line (as one
1558 line or as a continued sequence of lines) by extending its input buffer if
1559 necessary. This feature was broken for very long pattern lines, leading to
1560 a string of junk being passed to pcre_compile() if the pattern was longer
1561 than about 50K.
1562
1563 17. I have done a major re-factoring of the way pcre_compile() computes the
1564 amount of memory needed for a compiled pattern. Previously, there was code
1565 that made a preliminary scan of the pattern in order to do this. That was
1566 OK when PCRE was new, but as the facilities have expanded, it has become
1567 harder and harder to keep it in step with the real compile phase, and there
1568 have been a number of bugs (see for example, 4 above). I have now found a
1569 cunning way of running the real compile function in a "fake" mode that
1570 enables it to compute how much memory it would need, while actually only
1571 ever using a few hundred bytes of working memory and without too many
1572 tests of the mode. This should make future maintenance and development
1573 easier. A side effect of this work is that the limit of 200 on the nesting
1574 depth of parentheses has been removed (though this was never a serious
1575 limitation, I suspect). However, there is a downside: pcre_compile() now
1576 runs more slowly than before (30% or more, depending on the pattern). I
1577 hope this isn't a big issue. There is no effect on runtime performance.
1578
1579 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1580 newline (only possible for the last line of a file) and it was a
1581 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1582
1583 19. Added additional timing features to pcretest. (1) The -tm option now times
1584 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1585 separate command line item, by a number that specifies the number of
1586 repeats to use when timing. The default is 50000; this gives better
1587 precision, but takes uncomfortably long for very large patterns.
1588
1589 20. Extended pcre_study() to be more clever in cases where a branch of a
1590 subpattern has no definite first character. For example, (a*|b*)[cd] would
1591 previously give no result from pcre_study(). Now it recognizes that the
1592 first character must be a, b, c, or d.
1593
1594 21. There was an incorrect error "recursive call could loop indefinitely" if
1595 a subpattern (or the entire pattern) that was being tested for matching an
1596 empty string contained only one non-empty item after a nested subpattern.
1597 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1598 incorrectly, because the \d was being skipped in the check.
1599
1600 22. The pcretest program now has a new pattern option /B and a command line
1601 option -b, which is equivalent to adding /B to every pattern. This causes
1602 it to show the compiled bytecode, without the additional information that
1603 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1604 is the same as /B/I).
1605
1606 23. A new optimization is now able automatically to treat some sequences such
1607 as a*b as a*+b. More specifically, if something simple (such as a character
1608 or a simple class like \d) has an unlimited quantifier, and is followed by
1609 something that cannot possibly match the quantified thing, the quantifier
1610 is automatically "possessified".
1611
1612 24. A recursive reference to a subpattern whose number was greater than 39
1613 went wrong under certain circumstances in UTF-8 mode. This bug could also
1614 have affected the operation of pcre_study().
1615
1616 25. Realized that a little bit of performance could be had by replacing
1617 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1618
1619 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1620
1621 27. Possessive quantifiers such as a++ were previously implemented by turning
1622 them into atomic groups such as ($>a+). Now they have their own opcodes,
1623 which improves performance. This includes the automatically created ones
1624 from 23 above.
1625
1626 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1627 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1628 the first matched character to be a colon. This applied both to named and
1629 numbered groups.
1630
1631 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1632
1633 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1634 building PCRE on Windows easier, so I have included it in the distribution.
1635
1636 31. There is now a check in pcretest against a ridiculously large number being
1637 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1638 loop, the loop is abandoned.
1639
1640 32. Forward references to subpatterns in conditions such as (?(2)...) where
1641 subpattern 2 is defined later cause pcre_compile() to search forwards in
1642 the pattern for the relevant set of parentheses. This search went wrong
1643 when there were unescaped parentheses in a character class, parentheses
1644 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1645
1646 33. "Subroutine" calls and backreferences were previously restricted to
1647 referencing subpatterns earlier in the regex. This restriction has now
1648 been removed.
1649
1650 34. Added a number of extra features that are going to be in Perl 5.10. On the
1651 whole, these are just syntactic alternatives for features that PCRE had
1652 previously implemented using the Python syntax or my own invention. The
1653 other formats are all retained for compatibility.
1654
1655 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1656 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1657 also .NET compatible.
1658
1659 (b) A recursion or subroutine call to a named group can now be defined as
1660 (?&name) as well as (?P>name).
1661
1662 (c) A backreference to a named group can now be defined as \k<name> or
1663 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1664 5.10, are also .NET compatible.
1665
1666 (d) A conditional reference to a named group can now use the syntax
1667 (?(<name>) or (?('name') as well as (?(name).
1668
1669 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1670 groups (named and numbered) that are never evaluated inline, but can be
1671 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1672 is always false. There may be only one alternative in such a group.
1673
1674 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1675 as the simple (?(R). The condition is true only if the most recent
1676 recursion is that of the given number or name. It does not search out
1677 through the entire recursion stack.
1678
1679 (g) The escape \gN or \g{N} has been added, where N is a positive or
1680 negative number, specifying an absolute or relative reference.
1681
1682 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1683 some "unreachable code" warnings.
1684
1685 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1686 things, this adds five new scripts.
1687
1688 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1689 There were also incompatibilities regarding the handling of \Q..\E inside
1690 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1691 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1692
1693 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1694 matches an empty string, and forcibly breaks the loop. There were bugs in
1695 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1696 against aaaa the result was just "a" rather than "aaaa", for example. Two
1697 separate and independent bugs (that affected different cases) have been
1698 fixed.
1699
1700 39. Refactored the code to abolish the use of different opcodes for small
1701 capturing bracket numbers. This is a tidy that I avoided doing when I
1702 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1703 The new approach is not only tidier, it makes it possible to reduce the
1704 memory needed to fix the previous bug (38).
1705
1706 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1707 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1708 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1709 mode.
1710
1711 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1712 report.
1713
1714 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1715 copy construction and assignment in the C++ wrapper.
1716
1717 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1718 couple of bugs that could have given wrong results in the "--newline=crlf"
1719 case.
1720
1721 44. Added a number of casts and did some reorganization of signed/unsigned int
1722 variables following suggestions from Dair Grant. Also renamed the variable
1723 "this" as "item" because it is a C++ keyword.
1724
1725 45. Arranged for dftables to add
1726
1727 #include "pcre_internal.h"
1728
1729 to pcre_chartables.c because without it, gcc 4.x may remove the array
1730 definition from the final binary if PCRE is built into a static library and
1731 dead code stripping is activated.
1732
1733 46. For an unanchored pattern, if a match attempt fails at the start of a
1734 newline sequence, and the newline setting is CRLF or ANY, and the next two
1735 characters are CRLF, advance by two characters instead of one.
1736
1737
1738 Version 6.7 04-Jul-06
1739 ---------------------
1740
1741 1. In order to handle tests when input lines are enormously long, pcretest has
1742 been re-factored so that it automatically extends its buffers when
1743 necessary. The code is crude, but this _is_ just a test program. The
1744 default size has been increased from 32K to 50K.
1745
1746 2. The code in pcre_study() was using the value of the re argument before
1747 testing it for NULL. (Of course, in any sensible call of the function, it
1748 won't be NULL.)
1749
1750 3. The memmove() emulation function in pcre_internal.h, which is used on
1751 systems that lack both memmove() and bcopy() - that is, hardly ever -
1752 was missing a "static" storage class specifier.
1753
1754 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1755 containing an extended class (one that cannot be represented by a bitmap
1756 because it contains high-valued characters or Unicode property items, e.g.
1757 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1758 pattern, but PCRE should not loop if you do not (it no longer does).
1759 [Detail: two cases were found: (a) a repeated subpattern containing an
1760 extended class; (b) a recursive reference to a subpattern that followed a
1761 previous extended class. It wasn't skipping over the extended class
1762 correctly when UTF-8 mode was not set.]
1763
1764 5. A negated single-character class was not being recognized as fixed-length
1765 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1766 compile error "lookbehind assertion is not fixed length".
1767
1768 6. The RunPerlTest auxiliary script was showing an unexpected difference
1769 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1770 write a Perl script that can interpret lines of an input file either as
1771 byte characters or as UTF-8, which is what "perltest" was being required to
1772 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1773 can't do is switch easily at run time between having the "use utf8;" pragma
1774 or not. In the end, I fudged it by using the RunPerlTest script to insert
1775 "use utf8;" explicitly for the UTF-8 tests.
1776
1777 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1778 the end of the subject string, contrary to the documentation and to what
1779 Perl does. This was true of both matching functions. Now it matches only at
1780 the start of the subject and immediately after *internal* newlines.
1781
1782 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1783 a pointer to an int instead of a pointer to an unsigned long int. This
1784 caused problems on 64-bit systems.
1785
1786 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1787 instance of the 'standard' template library not being so standard".
1788
1789 10. There was no check on the number of named subpatterns nor the maximum
1790 length of a subpattern name. The product of these values is used to compute
1791 the size of the memory block for a compiled pattern. By supplying a very
1792 long subpattern name and a large number of named subpatterns, the size
1793 computation could be caused to overflow. This is now prevented by limiting
1794 the length of names to 32 characters, and the number of named subpatterns
1795 to 10,000.
1796
1797 11. Subpatterns that are repeated with specific counts have to be replicated in
1798 the compiled pattern. The size of memory for this was computed from the
1799 length of the subpattern and the repeat count. The latter is limited to
1800 65535, but there was no limit on the former, meaning that integer overflow
1801 could in principle occur. The compiled length of a repeated subpattern is
1802 now limited to 30,000 bytes in order to prevent this.
1803
1804 12. Added the optional facility to have named substrings with the same name.
1805
1806 13. Added the ability to use a named substring as a condition, using the
1807 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1808 are numbers (not recommended). Forward references are permitted.
1809
1810 14. Added forward references in named backreferences (if you see what I mean).
1811
1812 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1813 pattern could run off the end of the subject. For example, the pattern
1814 "(?s)(.{1,5})"8 did this with the subject "ab".
1815
1816 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1817 PCRE_CASELESS was set when matching characters that were quantified with ?
1818 or *.
1819
1820 17. A character class other than a single negated character that had a minimum
1821 but no maximum quantifier - for example [ab]{6,} - was not handled
1822 correctly by pce_dfa_exec(). It would match only one character.
1823
1824 18. A valid (though odd) pattern that looked like a POSIX character
1825 class but used an invalid character after [ (for example [[,abc,]]) caused
1826 pcre_compile() to give the error "Failed: internal error: code overflow" or
1827 in some cases to crash with a glibc free() error. This could even happen if
1828 the pattern terminated after [[ but there just happened to be a sequence of
1829 letters, a binary zero, and a closing ] in the memory that followed.
1830
1831 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1832 over the years. Originally (before any Unicode support), just the bottom 8
1833 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1834 output from "man perlunicode" includes this:
1835
1836 The regular expression compiler produces polymorphic opcodes. That
1837 is, the pattern adapts to the data and automatically switches to
1838 the Unicode character scheme when presented with Unicode data--or
1839 instead uses a traditional byte scheme when presented with byte
1840 data.
1841
1842 Sadly, a wide octal escape does not cause a switch, and in a string with
1843 no other multibyte characters, these octal escapes are treated as before.
1844 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1845 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1846 Unicode string.
1847
1848 I have not perpetrated such confusion in PCRE. Up till now, it took just
1849 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1850 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1851 translate to the appropriate multibyte character.
1852
1853 29. Applied some refactoring to reduce the number of warnings from Microsoft
1854 and Borland compilers. This has included removing the fudge introduced
1855 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1856 a warning about an unused variable.
1857
1858 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1859 characters since release 4.0, because Perl (from release 5.004) does not.
1860 [Or at least, is documented not to: some releases seem to be in conflict
1861 with the documentation.] However, when a pattern was studied with
1862 pcre_study() and all its branches started with \s, PCRE still included VT
1863 as a possible starting character. Of course, this did no harm; it just
1864 caused an unnecessary match attempt.
1865
1866 22. Removed a now-redundant internal flag bit that recorded the fact that case
1867 dependency changed within the pattern. This was once needed for "required
1868 byte" processing, but is no longer used. This recovers a now-scarce options
1869 bit. Also moved the least significant internal flag bit to the most-
1870 significant bit of the word, which was not previously used (hangover from
1871 the days when it was an int rather than a uint) to free up another bit for
1872 the future.
1873
1874 23. Added support for CRLF line endings as well as CR and LF. As well as the
1875 default being selectable at build time, it can now be changed at runtime
1876 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1877 specify that it is scanning data with non-default line endings.
1878
1879 24. Changed the definition of CXXLINK to make it agree with the definition of
1880 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1881
1882 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1883 recursions. This makes a big different to stack usage for some patterns.
1884
1885 26. If a subpattern containing a named recursion or subroutine reference such
1886 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1887 the space required for the compiled pattern went wrong and gave too small a
1888 value. Depending on the environment, this could lead to "Failed: internal
1889 error: code overflow at offset 49" or "glibc detected double free or
1890 corruption" errors.
1891
1892 27. Applied patches from Google (a) to support the new newline modes and (b) to
1893 advance over multibyte UTF-8 characters in GlobalReplace.
1894
1895 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1896 difference for some implementation of PCRE in some Windows version.
1897
1898 29. Added some extra testing facilities to pcretest:
1899
1900 \q<number> in a data line sets the "match limit" value
1901 \Q<number> in a data line sets the "match recursion limt" value
1902 -S <number> sets the stack size, where <number> is in megabytes
1903
1904 The -S option isn't available for Windows.
1905
1906
1907 Version 6.6 06-Feb-06
1908 ---------------------
1909
1910 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1911 in pcreposix.h. I have copied the definition from pcre.h.
1912
1913 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1914 because pcre.h is no longer a built file.
1915
1916 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1917 not normally included in the compiled code.
1918
1919
1920 Version 6.5 01-Feb-06
1921 ---------------------
1922
1923 1. When using the partial match feature with pcre_dfa_exec(), it was not
1924 anchoring the second and subsequent partial matches at the new starting
1925 point. This could lead to incorrect results. For example, with the pattern
1926 /1234/, partially matching against "123" and then "a4" gave a match.
1927
1928 2. Changes to pcregrep:
1929
1930 (a) All non-match returns from pcre_exec() were being treated as failures
1931 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1932 error message is output. Some extra information is given for the
1933 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1934 probably the only errors that are likely to be caused by users (by
1935 specifying a regex that has nested indefinite repeats, for instance).
1936 If there are more than 20 of these errors, pcregrep is abandoned.
1937
1938 (b) A binary zero was treated as data while matching, but terminated the
1939 output line if it was written out. This has been fixed: binary zeroes
1940 are now no different to any other data bytes.
1941
1942 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1943 used to set a locale for matching. The --locale=xxxx long option has
1944 been added (no short equivalent) to specify a locale explicitly on the
1945 pcregrep command, overriding the environment variables.
1946
1947 (d) When -B was used with -n, some line numbers in the output were one less
1948 than they should have been.
1949
1950 (e) Added the -o (--only-matching) option.
1951
1952 (f) If -A or -C was used with -c (count only), some lines of context were
1953 accidentally printed for the final match.
1954
1955 (g) Added the -H (--with-filename) option.
1956
1957 (h) The combination of options -rh failed to suppress file names for files
1958 that were found from directory arguments.
1959
1960 (i) Added the -D (--devices) and -d (--directories) options.
1961
1962 (j) Added the -F (--fixed-strings) option.
1963
1964 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1965
1966 (l) Added the --colo(u)r option.
1967
1968 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1969 is not present by default.
1970
1971 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1972 items such as (?R) or (?1), when the recursion could match a number of
1973 alternatives. If it matched one of the alternatives, but subsequently,
1974 outside the recursion, there was a failure, the code tried to back up into
1975 the recursion. However, because of the way PCRE is implemented, this is not
1976 possible, and the result was an incorrect result from the match.
1977
1978 In order to prevent this happening, the specification of recursion has
1979 been changed so that all such subpatterns are automatically treated as
1980 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1981
1982 4. I had overlooked the fact that, in some locales, there are characters for
1983 which isalpha() is true but neither isupper() nor islower() are true. In
1984 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1985 and ordfeminine) are like this. This affected the treatment of \w and \W
1986 when they appeared in character classes, but not when they appeared outside
1987 a character class. The bit map for "word" characters is now created
1988 separately from the results of isalnum() instead of just taking it from the
1989 upper, lower, and digit maps. (Plus the underscore character, of course.)
1990
1991 5. The above bug also affected the handling of POSIX character classes such as
1992 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1993 permanent tables. Instead, the bit maps for such a class were previously
1994 created as the appropriate unions of the upper, lower, and digit bitmaps.
1995 Now they are created by subtraction from the [[:word:]] class, which has
1996 its own bitmap.
1997
1998 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1999 It is created by subtracting the vertical space characters (\x09, \x0a,
2000 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2001 subtraction was done in the overall bitmap for a character class, meaning
2002 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2003 be recognized. This bug has been fixed.
2004
2005 7. Patches from the folks at Google:
2006
2007 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2008 real life, but is still worth protecting against".
2009
2010 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2011 regular expressions".
2012
2013 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2014 have it.
2015
2016 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2017 "configure" and the latter not, in order to fix a problem somebody had
2018 with compiling the Arg class on HP-UX.
2019
2020 (e) Improve the error-handling of the C++ wrapper a little bit.
2021
2022 (f) New tests for checking recursion limiting.
2023
2024 8. The pcre_memmove() function, which is used only if the environment does not
2025 have a standard memmove() function (and is therefore rarely compiled),
2026 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2027 returning a result (though PCRE never actually uses the result).
2028
2029 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2030 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2031 returned instead of calling malloc() with an overflowing number that would
2032 most likely cause subsequent chaos.
2033
2034 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2035
2036 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2037 with this option is matched, the nmatch and pmatch options of regexec() are
2038 ignored.
2039
2040 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2041 provided in case anyone wants to the the POSIX interface with UTF-8
2042 strings.
2043
2044 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2045 C++ linking (needed for some HP-UX environments).
2046
2047 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2048 (unused parameter) and in the pcre_printint() function (omitted "default"
2049 switch label when the default is to do nothing).
2050
2051 15. Added some code to make it possible, when PCRE is compiled as a C++
2052 library, to replace subject pointers for pcre_exec() with a smart pointer
2053 class, thus making it possible to process discontinuous strings.
2054
2055 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2056 much the same function. They were added by different people who were trying
2057 to make PCRE easy to compile on non-Unix systems. It has been suggested
2058 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2059 for compiling on Windows systems. I have therefore replaced it with
2060 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2061 defaults to "extern" for C or "extern C" for C++, which works fine on
2062 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2063 SCOPE with something explicit in config.h. In addition:
2064
2065 (a) pcreposix.h still had just "extern" instead of either of these macros;
2066 I have replaced it with PCRE_DATA_SCOPE.
2067
2068 (b) Functions such as _pcre_xclass(), which are internal to the library,
2069 but external in the C sense, all had PCRE_EXPORT in their definitions.
2070 This is apparently wrong for the Windows case, so I have removed it.
2071 (It makes no difference on Unix-like systems.)
2072
2073 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2074 of recursive calls to match(). This is different to MATCH_LIMIT because
2075 that limits the total number of calls to match(), not all of which increase
2076 the depth of recursion. Limiting the recursion depth limits the amount of
2077 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2078 when PCRE is compiled, and changed at run time. A patch from Google adds
2079 this functionality to the C++ interface.
2080
2081 18. Changes to the handling of Unicode character properties:
2082
2083 (a) Updated the table to Unicode 4.1.0.
2084
2085 (b) Recognize characters that are not in the table as "Cn" (undefined).
2086
2087 (c) I revised the way the table is implemented to a much improved format
2088 which includes recognition of ranges. It now supports the ranges that
2089 are defined in UnicodeData.txt, and it also amalgamates other
2090 characters into ranges. This has reduced the number of entries in the
2091 table from around 16,000 to around 3,000, thus reducing its size
2092 considerably. I realized I did not need to use a tree structure after
2093 all - a binary chop search is just as efficient. Having reduced the
2094 number of entries, I extended their size from 6 bytes to 8 bytes to
2095 allow for more data.
2096
2097 (d) Added support for Unicode script names via properties such as \p{Han}.
2098
2099 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2100 matching that character.
2101
2102 20. When matching a repeated Unicode property with a minimum greater than zero,
2103 (for example \pL{2,}), PCRE could look past the end of the subject if it
2104 reached it while seeking the minimum number of characters. This could
2105 happen only if some of the characters were more than one byte long, because
2106 there is a check for at least the minimum number of bytes.
2107
2108 21. Refactored the implementation of \p and \P so as to be more general, to
2109 allow for more different types of property in future. This has changed the
2110 compiled form incompatibly. Anybody with saved compiled patterns that use
2111 \p or \P will have to recompile them.
2112
2113 22. Added "Any" and "L&" to the supported property types.
2114
2115 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2116 but give a compile time error if the value is greater than 0xff.
2117
2118 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2119 accidentally not being installed or uninstalled.
2120
2121 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2122 made were to insert the current release number. This seemed silly, because
2123 it made things harder for people building PCRE on systems that don't run
2124 "configure". I have turned pcre.h into a distributed file, no longer built
2125 by "configure", with the version identification directly included. There is
2126 no longer a pcre.h.in file.
2127
2128 However, this change necessitated a change to the pcre-config script as
2129 well. It is built from pcre-config.in, and one of the substitutions was the
2130 release number. I have updated configure.ac so that ./configure now finds
2131 the release number by grepping pcre.h.
2132
2133 26. Added the ability to run the tests under valgrind.
2134
2135
2136 Version 6.4 05-Sep-05
2137 ---------------------
2138
2139 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2140 "--" to be printed when multiple files were scanned, even when none of the
2141 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2142 consider it to be a bug, and have restored the previous behaviour.
2143
2144 2. A couple of code tidies to get rid of compiler warnings.
2145
2146 3. The pcretest program used to cheat by referring to symbols in the library
2147 whose names begin with _pcre_. These are internal symbols that are not
2148 really supposed to be visible externally, and in some environments it is
2149 possible to suppress them. The cheating is now confined to including
2150 certain files from the library's source, which is a bit cleaner.
2151
2152 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2153 file's purpose clearer.
2154
2155 5. Reorganized pcre_ucp_findchar().
2156
2157
2158 Version 6.3 15-Aug-05
2159 ---------------------
2160
2161 1. The file libpcre.pc.in did not have general read permission in the tarball.
2162
2163 2. There were some problems when building without C++ support:
2164
2165 (a) If C++ support was not built, "make install" and "make test" still
2166 tried to test it.
2167
2168 (b) There were problems when the value of CXX was explicitly set. Some
2169 changes have been made to try to fix these, and ...
2170
2171 (c) --disable-cpp can now be used to explicitly disable C++ support.
2172
2173 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2174 backslash in a target when C++ was disabled. This confuses some
2175 versions of "make", apparently. Using an intermediate variable solves
2176 this. (Same for CPP_LOBJ.)
2177
2178 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2179 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2180 necessary on certain architectures.
2181
2182 4. Added a setting of -export-symbols-regex to the link command to remove
2183 those symbols that are exported in the C sense, but actually are local
2184 within the library, and not documented. Their names all begin with
2185 "_pcre_". This is not a perfect job, because (a) we have to except some
2186 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2187 available (and never for static libraries). I have made a note to try to
2188 find a way round (a) in the future.
2189
2190
2191 Version 6.2 01-Aug-05
2192 ---------------------
2193
2194 1. There was no test for integer overflow of quantifier values. A construction
2195 such as {1111111111111111} would give undefined results. What is worse, if
2196 a minimum quantifier for a parenthesized subpattern overflowed and became
2197 negative, the calculation of the memory size went wrong. This could have
2198 led to memory overwriting.
2199
2200 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2201
2202 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2203 operating environments where this matters.
2204
2205 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2206 PCRE options from within the C++ wrapper.
2207
2208 5. Named capturing subpatterns were not being correctly counted when a pattern
2209 was compiled. This caused two problems: (a) If there were more than 100
2210 such subpatterns, the calculation of the memory needed for the whole
2211 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2212 back references of the form \12, where the number was greater than 9, were
2213 not recognized as back references, even though there were sufficient
2214 previous subpatterns.
2215
2216 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2217 versions of gcc, e.g. 2.95.4.
2218
2219
2220 Version 6.1 21-Jun-05
2221 ---------------------
2222
2223 1. There was one reference to the variable "posix" in pcretest.c that was not
2224 surrounded by "#if !defined NOPOSIX".
2225
2226 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2227 the cross-check on the old pcre_info() function, for the benefit of the
2228 cut-down version of PCRE that is currently imported into Exim.
2229
2230 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2231 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2232 patterns that start (?i) but I don't think that matters. The use of (?i) is
2233 just an example; this all applies to the other options as well.
2234
2235 4. Since libtool seems to echo the compile commands it is issuing, the output
2236 from "make" can be reduced a bit by putting "@" in front of each libtool
2237 compile command.
2238
2239 5. Patch from the folks at Google for configure.in to be a bit more thorough
2240 in checking for a suitable C++ installation before trying to compile the
2241 C++ stuff. This should fix a reported problem when a compiler was present,
2242 but no suitable headers.
2243
2244 6. The man pages all had just "PCRE" as their title. I have changed them to
2245 be the relevant file name. I have also arranged that these names are
2246 retained in the file doc/pcre.txt, which is a concatenation in text format
2247 of all the man pages except the little individual ones for each function.
2248
2249 7. The NON-UNIX-USE file had not been updated for the different set of source
2250 files that come with release 6. I also added a few comments about the C++
2251 wrapper.
2252
2253
2254 Version 6.0 07-Jun-05
2255 ---------------------
2256
2257 1. Some minor internal re-organization to help with my DFA experiments.
2258
2259 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2260 didn't matter for the library itself when fully configured, but did matter
2261 when compiling without UCP support, or within Exim, where the ucp files are
2262 not imported.
2263
2264 3. Refactoring of the library code to split up the various functions into
2265 different source modules. The addition of the new DFA matching code (see
2266 below) to a single monolithic source would have made it really too
2267 unwieldy, quite apart from causing all the code to be include in a
2268 statically linked application, when only some functions are used. This is
2269 relevant even without the DFA addition now that patterns can be compiled in
2270 one application and matched in another.
2271
2272 The downside of splitting up is that there have to be some external
2273 functions and data tables that are used internally in different modules of
2274 the library but which are not part of the API. These have all had their
2275 names changed to start with "_pcre_" so that they are unlikely to clash
2276 with other external names.
2277
2278 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2279 a different (DFA) algorithm. Although it is slower than the original
2280 function, it does have some advantages for certain types of matching
2281 problem.
2282
2283 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2284 including restarting after a partial match.
2285
2286 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2287 defined when compiling for Windows was sent to me. I have put it into the
2288 code, though I have no means of testing or verifying it.
2289
2290 7. Added the pcre_refcount() auxiliary function.
2291
2292 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2293 match before or at the first newline in the subject string. In pcretest,
2294 the /f option on a pattern can be used to set this.
2295
2296 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2297 would behave wrongly. This has been present in PCRE since release 4.0.
2298
2299 10. A number of changes to the pcregrep command:
2300
2301 (a) Refactored how -x works; insert ^(...)$ instead of setting
2302 PCRE_ANCHORED and checking the length, in preparation for adding
2303 something similar for -w.
2304
2305 (b) Added the -w (match as a word) option.
2306
2307 (c) Refactored the way lines are read and buffered so as to have more
2308 than one at a time available.
2309
2310 (d) Implemented a pcregrep test script.
2311
2312 (e) Added the -M (multiline match) option. This allows patterns to match
2313 over several lines of the subject. The buffering ensures that at least
2314 8K, or the rest of the document (whichever is the shorter) is available
2315 for matching (and similarly the previous 8K for lookbehind assertions).
2316
2317 (f) Changed the --help output so that it now says
2318
2319 -w, --word-regex(p)
2320
2321 instead of two lines, one with "regex" and the other with "regexp"
2322 because that confused at least one person since the short forms are the
2323 same. (This required a bit of code, as the output is generated
2324 automatically from a table. It wasn't just a text change.)
2325
2326 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2327 option but starts with a hyphen. Could be a pattern or a path name
2328 starting with a hyphen, for instance.
2329
2330 (h) "-" can be given as a file name to represent stdin.
2331
2332 (i) When file names are being printed, "(standard input)" is used for
2333 the standard input, for compatibility with GNU grep. Previously
2334 "<stdin>" was used.
2335
2336 (j) The option --label=xxx can be used to supply a name to be used for
2337 stdin when file names are being printed. There is no short form.
2338
2339 (k) Re-factored the options decoding logic because we are going to add
2340 two more options that take data. Such options can now be given in four
2341 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2342
2343 (l) Added the -A, -B, and -C options for requesting that lines of context
2344 around matches be printed.
2345
2346 (m) Added the -L option to print the names of files that do not contain
2347 any matching lines, that is, the complement of -l.
2348
2349 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2350 continue to scan other files.
2351
2352 (o) The -s option was incorrectly implemented. For compatibility with other
2353 greps, it now suppresses the error message for a non-existent or non-
2354 accessible file (but not the return code). There is a new option called
2355 -q that suppresses the output of matching lines, which was what -s was
2356 previously doing.
2357
2358 (p) Added --include and --exclude options to specify files for inclusion
2359 and exclusion when recursing.
2360
2361 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2362 Hopefully, it now does.
2363
2364 12. Missing cast in pcre_study().
2365
2366 13. Added an "uninstall" target to the makefile.
2367
2368 14. Replaced "extern" in the function prototypes in Makefile.in with
2369 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2370 world, but is set differently for Windows.
2371
2372 15. Added a second compiling function called pcre_compile2(). The only
2373 difference is that it has an extra argument, which is a pointer to an
2374 integer error code. When there is a compile-time failure, this is set
2375 non-zero, in addition to the error test pointer being set to point to an
2376 error message. The new argument may be NULL if no error number is required
2377 (but then you may as well call pcre_compile(), which is now just a
2378 wrapper). This facility is provided because some applications need a
2379 numeric error indication, but it has also enabled me to tidy up the way
2380 compile-time errors are handled in the POSIX wrapper.
2381
2382 16. Added VPATH=.libs to the makefile; this should help when building with one
2383 prefix path and installing with another. (Or so I'm told by someone who
2384 knows more about this stuff than I do.)
2385
2386 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2387 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2388 match everything, including newlines. This is not POSIX-compatible, but
2389 somebody wanted the feature. From pcretest it can be activated by using
2390 both the P and the s flags.
2391
2392 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2393
2394 19. libpcre.pc was being incorrectly installed as executable.
2395
2396 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2397 it now also looks for '\r' so that it will work unmodified on Windows.
2398
2399 21. Added Google's contributed C++ wrapper to the distribution.
2400
2401 22. Added some untidy missing memory free() calls in pcretest, to keep
2402 Electric Fence happy when testing.
2403
2404
2405
2406 Version 5.0 13-Sep-04
2407 ---------------------
2408
2409 1. Internal change: literal characters are no longer packed up into items
2410 containing multiple characters in a single byte-string. Each character
2411 is now matched using a separate opcode. However, there may be more than one
2412 byte in the character in UTF-8 mode.
2413
2414 2. The pcre_callout_block structure has two new fields: pattern_position and
2415 next_item_length. These contain the offset in the pattern to the next match
2416 item, and its length, respectively.
2417
2418 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2419 insertion of callouts before each pattern item. Added the /C option to
2420 pcretest to make use of this.
2421
2422 4. On the advice of a Windows user, the lines
2423
2424 #if defined(_WIN32) || defined(WIN32)
2425 _setmode( _fileno( stdout ), 0x8000 );
2426 #endif /* defined(_WIN32) || defined(WIN32) */
2427
2428 have been added to the source of pcretest. This apparently does useful
2429 magic in relation to line terminators.
2430
2431 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2432 for the benefit of those environments where the "b" makes a difference.
2433
2434 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2435 to know about it. I have put a hack into configure.in that adds in code
2436 to set GCC=yes if CC=icc. This seems to end up at a point in the
2437 generated configure script that is early enough to affect the setting of
2438 compiler options, which is what is needed, but I have no means of testing
2439 whether it really works. (The user who reported this had patched the
2440 generated configure script, which of course I cannot do.)
2441
2442 LATER: After change 22 below (new libtool files), the configure script
2443 seems to know about icc (and also ecc). Therefore, I have commented out
2444 this hack in configure.in.
2445
2446 7. Added support for pkg-config (2 patches were sent in).
2447
2448 8. Negated POSIX character classes that used a combination of internal tables
2449 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2450 [[:^ascii]]. Typically, they would match almost any characters. The other
2451 POSIX classes were not broken in this way.
2452
2453 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2454 to find the match, as PCRE was deluded into thinking that the match had to
2455 start at the start point or following a newline. The same bug applied to
2456 patterns with negative forward assertions or any backward assertions
2457 preceding ".*" at the start, unless the pattern required a fixed first
2458 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2459
2460 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2461 starting at the last subject character, bytes beyond the end of the subject
2462 string were read.
2463
2464 11. Renamed the variable "class" as "classbits" to make life easier for C++
2465 users. (Previously there was a macro definition, but it apparently wasn't
2466 enough.)
2467
2468 12. Added the new field "tables" to the extra data so that tables can be passed
2469 in at exec time, or the internal tables can be re-selected. This allows
2470 a compiled regex to be saved and re-used at a later time by a different
2471 program that might have everything at different addresses.
2472
2473 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2474 -R library as well as a -L library.
2475
2476 14. The debugging options of pcretest (-d on the command line or D on a
2477 pattern) showed incorrect output for anything following an extended class
2478 that contained multibyte characters and which was followed by a quantifier.
2479
2480 15. Added optional support for general category Unicode character properties
2481 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2482 support. It adds about 90K to the size of the library. The meanings of the
2483 inbuilt class escapes such as \d and \s have NOT been changed.
2484
2485 16. Updated pcredemo.c to include calls to free() to release the memory for the
2486 compiled pattern.
2487
2488 17. The generated file chartables.c was being created in the source directory
2489 instead of in the building directory. This caused the build to fail if the
2490 source directory was different from the building directory, and was
2491 read-only.
2492
2493 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2494 file. No doubt somebody will tell me if they don't make sense... Also added
2495 Dan Mooney's comments about building on OpenVMS.
2496
2497 19. Added support for partial matching via the PCRE_PARTIAL option for
2498 pcre_exec() and the \P data escape in pcretest.
2499
2500 20. Extended pcretest with 3 new pattern features:
2501
2502 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2503 write the compiled pattern to the file whose name is "rest-of-line".
2504 This is a straight binary dump of the data, with the saved pointer to
2505 the character tables forced to be NULL. The study data, if any, is
2506 written too. After writing, pcretest reads a new pattern.
2507
2508 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2509 compiled pattern from the given file. There must not be any
2510 occurrences of "<" in the file name (pretty unlikely); if there are,
2511 pcretest will instead treat the initial "<" as a pattern delimiter.
2512 After reading in the pattern, pcretest goes on to read data lines as
2513 usual.
2514
2515 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2516 and 16-bit fields in a compiled pattern, to simulate a pattern that
2517 was compiled on a host of opposite endianness.
2518
2519 21. The pcre-exec() function can now cope with patterns that were compiled on
2520 hosts of opposite endianness, with this restriction:
2521
2522 As for any compiled expression that is saved and used later, the tables
2523 pointer field cannot be preserved; the extra_data field in the arguments
2524 to pcre_exec() should be used to pass in a tables address if a value
2525 other than the default internal tables were used at compile time.
2526
2527 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2528 now diagnosed as an error. Previously, most of the time, a negative number
2529 would have been treated as zero, but if in addition "ovector" was passed as
2530 NULL, a crash could occur.
2531
2532 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2533 new versions from the libtool 1.5 distribution (the last one is a copy of
2534 a file called libtool.m4). This seems to have fixed the need to patch
2535 "configure" to support Darwin 1.3 (which I used to do). However, I still
2536 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2537 workstation).
2538
2539 24. Changed the PCRE licence to be the more standard "BSD" licence.
2540
2541
2542 Version 4.5 01-Dec-03
2543 ---------------------
2544
2545 1. There has been some re-arrangement of the code for the match() function so
2546 that it can be compiled in a version that does not call itself recursively.
2547 Instead, it keeps those local variables that need separate instances for
2548 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2549 needs to "recurse". Keeping track of where control must go is done by means
2550 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2551 hide most of the details from the main code, and operates only if
2552 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2553 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2554 operating.
2555
2556 To make it easier for callers to provide specially tailored get/free
2557 functions for this usage, two new functions, pcre_stack_malloc, and
2558 pcre_stack_free, are used. They are always called in strict stacking order,
2559 and the size of block requested is always the same.
2560
2561 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2562 PCRE has been compiled to use the stack or the heap for recursion. The
2563 -C option of pcretest uses this to show which version is compiled.
2564
2565 A new data escape \S, is added to pcretest; it causes the amounts of store
2566 obtained and freed by both kinds of malloc/free at match time to be added
2567 to the output.
2568
2569 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2570 what's available on my current Linux desktop machine.
2571
2572 3. When matching a UTF-8 string, the test for a valid string at the start has
2573 been extended. If start_offset is not zero, PCRE now checks that it points
2574 to a byte that is the start of a UTF-8 character. If not, it returns
2575 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2576 this is necessary because there may be backward assertions in the pattern.
2577 When matching the same subject several times, it may save resources to use
2578 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2579
2580 4. The code for checking the validity of UTF-8 strings has been tightened so
2581 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2582 containing "overlong sequences".
2583
2584 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2585 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2586 should have been "&", but it just so happened that all the cases this let
2587 through by mistake were picked up later in the function.
2588
2589 6. I had used a variable called "isblank" - this is a C99 function, causing
2590 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2591
2592 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2593 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2594 so the version is shown; (c) stop "make test" echoing "./RunTest".
2595
2596 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2597
2598 9. The prototype for memmove() for systems that don't have it was using
2599 size_t, but the inclusion of the header that defines size_t was later. I've
2600 moved the #includes for the C headers earlier to avoid this.
2601
2602 10. Added some adjustments to the code to make it easier to compiler on certain
2603 special systems:
2604
2605 (a) Some "const" qualifiers were missing.
2606 (b) Added the macro EXPORT before all exported functions; by default this
2607 is defined to be empty.
2608 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2609 that it reads its output file name as an argument instead of writing
2610 to the standard output and assuming this can be redirected.
2611
2612 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2613 class containing characters with values greater than 255, PCRE compilation
2614 went into a loop.
2615
2616 12. A recursive reference to a subpattern that was within another subpattern
2617 that had a minimum quantifier of zero caused PCRE to crash. For example,
2618 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2619 recursion. If the recursively-called subpattern itself had a zero repeat,
2620 that was OK.
2621
2622 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2623 buffer into which it was copied (for escape processing) was still set at
2624 1024, so long lines caused crashes.
2625
2626 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2627 "internal error: code overflow...". This applied to any character class
2628 that was followed by a possessive quantifier.
2629
2630 15. Modified the Makefile to add libpcre.la as a prerequisite for
2631 libpcreposix.la because I was told this is needed for a parallel build to
2632 work.
2633
2634 16. If a pattern that contained .* following optional items at the start was
2635 studied, the wrong optimizing data was generated, leading to matching
2636 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2637 matching string must start with a or b or c. The correct conclusion for
2638 this pattern is that a match can start with any character.
2639
2640
2641 Version 4.4 13-Aug-03
2642 ---------------------
2643
2644 1. In UTF-8 mode, a character class containing characters with values between
2645 127 and 255 was not handled correctly if the compiled pattern was studied.
2646 In fixing this, I have also improved the studying algorithm for such
2647 classes (slightly).
2648
2649 2. Three internal functions had redundant arguments passed to them. Removal
2650 might give a very teeny performance improvement.
2651
2652 3. Documentation bug: the value of the capture_top field in a callout is *one
2653 more than* the number of the hightest numbered captured substring.
2654
2655 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2656 in incorrectly linking with a previously installed version. They now link
2657 explicitly with libpcre.la.
2658
2659 5. configure.in no longer needs to recognize Cygwin specially.
2660
2661 6. A problem in pcre.in for Windows platforms is fixed.
2662
2663 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2664 pcretest, it used to include the size of the study block as part of its
2665 output. Unfortunately, the structure contains a field that has a different
2666 size on different hardware architectures. This meant that the tests that
2667 showed this size failed. As the block is currently always of a fixed size,
2668 this information isn't actually particularly useful in pcretest output, so
2669 I have just removed it.
2670
2671 8. Three pre-processor statements accidentally did not start in column 1.
2672 Sadly, there are *still* compilers around that complain, even though
2673 standard C has not required this for well over a decade. Sigh.
2674
2675 9. In pcretest, the code for checking callouts passed small integers in the
2676 callout_data field, which is a void * field. However, some picky compilers
2677 complained about the casts involved for this on 64-bit systems. Now
2678 pcretest passes the address of the small integer instead, which should get
2679 rid of the warnings.
2680
2681 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2682 both compile and run time, and gives an error if an invalid UTF-8 sequence
2683 is found. There is a option for disabling this check in cases where the
2684 string is known to be correct and/or the maximum performance is wanted.
2685
2686 11. In response to a bug report, I changed one line in Makefile.in from
2687
2688 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2689 to
2690 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2691
2692 to look similar to other lines, but I have no way of telling whether this
2693 is the right thing to do, as I do not use Windows. No doubt I'll get told
2694 if it's wrong...
2695
2696
2697 Version 4.3 21-May-03
2698 ---------------------
2699
2700 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2701 Makefile.
2702
2703 2. Some refactoring to improve the quality of the code:
2704
2705 (i) The utf8_table... variables are now declared "const".
2706
2707 (ii) The code for \cx, which used the "case flipping" table to upper case
2708 lower case letters, now just substracts 32. This is ASCII-specific,
2709 but the whole concept of \cx is ASCII-specific, so it seems
2710 reasonable.
2711
2712 (iii) PCRE was using its character types table to recognize decimal and
2713 hexadecimal digits in the pattern. This is silly, because it handles
2714 only 0-9, a-f, and A-F, but the character types table is locale-
2715 specific, which means strange things might happen. A private
2716 table is now used for this - though it costs 256 bytes, a table is
2717 much faster than multiple explicit tests. Of course, the standard
2718 character types table is still used for matching digits in subject
2719 strings against \d.
2720
2721 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2722 ending in _t are). So I've renamed it as ESC_tee.
2723
2724 3. The first argument for regexec() in the POSIX wrapper should have been
2725 defined as "const".
2726
2727 4. Changed pcretest to use malloc() for its buffers so that they can be
2728 Electric Fenced for debugging.
2729
2730 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2731 to read one or more bytes before the start of the subject string. Often this
2732 had no effect on PCRE's behaviour, but in some circumstances it could
2733 provoke a segmentation fault.
2734
2735 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2736 to try to read one or more bytes before the start of the subject string.
2737
2738 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2739 UTF-8 support could misbehave in various ways if the subject string
2740 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2741 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2742 back over UTF-8 characters.)
2743
2744
2745 Version 4.2 14-Apr-03
2746 ---------------------
2747
2748 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2749
2750 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2751 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2752 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2753 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2754 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2755 and BUILD_EXEEXT
2756 Note: automatic setting of the BUILD variables is not yet working
2757 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2758 compile-time but not at link-time
2759 [LINK]: use for linking executables only
2760 make different versions for Windows and non-Windows
2761 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2762 libraries
2763 [LINK_FOR_BUILD]: new variable
2764 [OBJEXT]: use throughout
2765 [EXEEXT]: use throughout
2766 <winshared>: new target
2767 <wininstall>: new target
2768 <dftables.o>: use native compiler
2769 <dftables>: use native linker
2770 <install>: handle Windows platform correctly
2771 <clean>: ditto
2772 <check>: ditto
2773 copy DLL to top builddir before testing
2774
2775 As part of these changes, -no-undefined was removed again. This was reported
2776 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2777 in any case.
2778
2779 3. Some tidies to get rid of compiler warnings:
2780
2781 . In the match_data structure, match_limit was an unsigned long int, whereas
2782 match_call_count was an int. I've made them both unsigned long ints.
2783
2784 . In pcretest the fact that a const uschar * doesn't automatically cast to
2785 a void * provoked a warning.
2786
2787 . Turning on some more compiler warnings threw up some "shadow" variables
2788 and a few more missing casts.
2789
2790 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2791 option, a class that contained a single character with a value between 128
2792 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2793
2794 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2795 option, a class that contained several characters, but with at least one
2796 whose value was between 128 and 255 caused PCRE to crash.
2797
2798
2799 Version 4.1 12-Mar-03
2800 ---------------------
2801
2802 1. Compiling with gcc -pedantic found a couple of places where casts were
2803 needed, and a string in dftables.c that was longer than standard compilers are
2804 required to support.
2805
2806 2. Compiling with Sun's compiler found a few more places where the code could
2807 be tidied up in order to avoid warnings.
2808
2809 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2810 first of these names is deprecated in the latest Autoconf in favour of the name
2811 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2812 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2813 analogy I have changed it to CFLAGS_FOR_BUILD.
2814
2815 4. Added -no-undefined to the linking command in the Makefile, because this is
2816 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2817 linking step for the pcreposix library.
2818
2819 5. PCRE was failing to diagnose the case of two named groups with the same
2820 name.
2821
2822 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2823 literal character that is needed in the subject for a match, and scans along to
2824 ensure that it is present before embarking on the full matching process. This
2825 saves time in cases of nested unlimited repeats that are never going to match.
2826 Problem: the scan can take a lot of time if the subject is very long (e.g.
2827 megabytes), thus penalizing straightforward matches. It is now done only if the
2828 amount of subject to be scanned is less than 1000 bytes.
2829
2830 7. A lesser problem with the same optimization is that it was recording the
2831 first character of an anchored pattern as "needed", thus provoking a search
2832 right along the subject, even when the first match of the pattern was going to
2833 fail. The "needed" character is now not set for anchored patterns, unless it
2834 follows something in the pattern that is of non-fixed length. Thus, it still
2835 fulfils its original purpose of finding quick non-matches in cases of nested
2836 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2837
2838
2839 Version 4.0 17-Feb-03
2840 ---------------------
2841
2842 1. If a comment in an extended regex that started immediately after a meta-item
2843 extended to the end of string, PCRE compiled incorrect data. This could lead to
2844 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2845
2846 2. Moved to autoconf 2.53 and libtool 1.4.2.
2847
2848 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2849 the special perltest8 script is no longer needed - all the tests can be run
2850 from a single perltest script.
2851
2852 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2853 by \s. It has now been removed in PCRE. This means it isn't recognized as
2854 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2855 class [:space:] *does* include VT, thereby creating a mess.
2856
2857 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2858 space and tab.
2859
2860 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2861 its new features into the main test script, reducing the number of scripts.
2862
2863 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2864 were backward compatible, and made the (?i) apply to the whole pattern, as if
2865 /i were given. Now it behaves more logically, and applies the option setting
2866 only to what follows. PCRE has been changed to follow suit. However, if it
2867 finds options settings right at the start of the pattern, it extracts them into
2868 the global options, as before. Thus, they show up in the info data.
2869
2870 8. Added support for the \Q...\E escape sequence. Characters in between are
2871 treated as literals. This is slightly different from Perl in that $ and @ are
2872 also handled as literals inside the quotes. In Perl, they will cause variable
2873 interpolation. Note the following examples:
2874
2875 Pattern PCRE matches Perl matches
2876
2877 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2878 \Qabc\$xyz\E abc\$xyz abc\$xyz
2879 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2880
2881 For compatibility with Perl, \Q...\E sequences are recognized inside character
2882 classes as well as outside them.
2883
2884 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2885 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2886 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2887 signed/unsigned warnings.
2888
2889 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2890 option for pcretest, so I've replaced it by a simple function that does just
2891 that job.
2892
2893 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2894 "pcregrep -".
2895
2896 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2897 Java package. This provides some syntactic sugar for simple cases of what my
2898 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2899 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2900 item, you can use this simplified notation. Note that only makes sense with
2901 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2902 greediness, whatever the setting of the PCRE_UNGREEDY option.
2903
2904 13. A change of greediness default within a pattern was not taking effect at
2905 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2906 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2907 was abstracted outside.
2908
2909 14. PCRE now supports the \G assertion. It is true when the current matching
2910 position is at the start point of the match. This differs from \A when the
2911 starting offset is non-zero. Used with the /g option of pcretest (or similar
2912 code), it works in the same way as it does for Perl's /g option. If all
2913 alternatives of a regex begin with \G, the expression is anchored to the start
2914 match position, and the "anchored" flag is set in the compiled expression.
2915
2916 15. Some bugs concerning the handling of certain option changes within patterns
2917 have been fixed. These applied to options other than (?ims). For example,
2918 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2919 been the other way round. Some of this was related to change 7 above.
2920
2921 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2922 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2923 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2924 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2925
2926 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2927 mode. Unlike ".", it always matches newline, whatever the setting of
2928 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2929 assertions. Perl allows it, but it doesn't (in general) work because it can't
2930 calculate the length of the lookbehind. At least, that's the case for Perl
2931 5.8.0 - I've been told they are going to document that it doesn't work in
2932 future.
2933
2934 18. Added an error diagnosis for escapes that PCRE does not support: these are
2935 \L, \l, \N, \P, \p, \U, \u, and \X.
2936
2937 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2938 reading past the end of the pattern in cases such as /[abcd/.
2939
2940 20. PCRE was getting more memory than necessary for patterns with classes that
2941 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2942
2943 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2944 compiling PCRE for use with Virtual Pascal.
2945
2946 22. Small fix to the Makefile to make it work properly if the build is done
2947 outside the source tree.
2948
2949 23. Added a new extension: a condition to go with recursion. If a conditional
2950 subpattern starts with (?(R) the "true" branch is used if recursion has
2951 happened, whereas the "false" branch is used only at the top level.
2952
2953 24. When there was a very long string of literal characters (over 255 bytes
2954 without UTF support, over 250 bytes with UTF support), the computation of how
2955 much memory was required could be incorrect, leading to segfaults or other
2956 strange effects.
2957
2958 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2959 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2960 there was a subsequent back reference to those brackets. This meant that, for
2961 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2962 possible to check for precisely this case. All we can do is abandon the
2963 optimization if .* occurs inside capturing brackets when there are any back
2964 references whatsoever. (See below for a better fix that came later.)
2965
2966 26. The handling of the optimization for finding the first character of a
2967 non-anchored pattern, and for finding a character that is required later in the
2968 match were failing in some cases. This didn't break the matching; it just
2969 failed to optimize when it could. The way this is done has been re-implemented.
2970
2971 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2972
2973 28. Added a new feature that provides some of the functionality that Perl
2974 provides with (?{...}). The facility is termed a "callout". The way it is done
2975 in PCRE is for the caller to provide an optional function, by setting
2976 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2977 global variable. By default it is unset, which disables all calling out. To get
2978 the function called, the regex must include (?C) at appropriate points. This
2979 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2980 This provides a means of identifying different callout points. When PCRE
2981 reaches such a point in the regex, if pcre_callout has been set, the external
2982 function is called. It is provided with data in a structure called
2983 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2984 matching continues; if it returns a non-zero value, the match at the current
2985 point fails. However, backtracking will occur if possible. [This was changed
2986 later and other features added - see item 49 below.]
2987
2988 29. pcretest is upgraded to test the callout functionality. It provides a
2989 callout function that displays information. By default, it shows the start of
2990 the match and the current position in the text. There are some new data escapes
2991 to vary what happens:
2992
2993 \C+ in addition, show current contents of captured substrings
2994 \C- do not supply a callout function
2995 \C!n return 1 when callout number n is reached
2996 \C!n!m return 1 when callout number n is reached for the mth time
2997
2998 30. If pcregrep was called with the -l option and just a single file name, it
2999 output "<stdin>" if a match was found, instead of the file name.
3000
3001 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3002 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3003 pcre_exec(). This saves a malloc/free per call. The default value of
3004 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3005 when configuring.
3006
3007 32. The default maximum size of a compiled pattern is 64K. There have been a
3008 few cases of people hitting this limit. The code now uses macros to handle the
3009 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3010 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3011 configuring. Tests 2 and 5 work only with 2-byte links because they output
3012 debugging information about compiled patterns.
3013
3014 33. Internal code re-arrangements:
3015
3016 (a) Moved the debugging function for printing out a compiled regex into
3017 its own source file (printint.c) and used #include to pull it into
3018 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3019 separate copies.
3020
3021 (b) Defined the list of op-code names for debugging as a macro in
3022 internal.h so that it is next to the definition of the opcodes.
3023
3024 (c) Defined a table of op-code lengths for simpler skipping along compiled
3025 code. This is again a macro in internal.h so that it is next to the
3026 definition of the opcodes.
3027
3028 34. Added support for recursive calls to individual subpatterns, along the
3029 lines of Robin Houston's patch (but implemented somewhat differently).
3030
3031 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3032 allow it to read and process whole directories in Win32. This code was
3033 contributed by Lionel Fourquaux; it has not been tested by me.
3034
3035 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3036 used to name a group. Names consist of alphanumerics and underscores, and must
3037 be unique. Back references use the syntax (?P=name) and recursive calls use
3038 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3039 numbers. The function pcre_fullinfo() can be used after compilation to extract
3040 a name/number map. There are three relevant calls:
3041
3042 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3043 PCRE_INFO_NAMECOUNT yields the number of entries
3044 PCRE_INFO_NAMETABLE yields a pointer to the map.
3045
3046 The map is a vector of fixed-size entries. The size of each entry depends on
3047 the length of the longest name used. The first two bytes of each entry are the
3048 group number, most significant byte first. There follows the corresponding
3049 name, zero terminated. The names are in alphabetical order.
3050
3051 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3052 case instead of 255. Making it the same both with and without UTF-8 support
3053 means that the same test output works with both.
3054
3055 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3056 calling malloc() with a zero argument.
3057
3058 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3059 optimization. I've improved things by keeping a bitmap of backreferences with
3060 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3061 fact referenced, the optimization can be applied. It is unlikely that a
3062 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3063 the match to follow \n) will appear inside brackets with a number greater than
3064 31, but if it does, any back reference > 31 suppresses the optimization.
3065
3066 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3067 of disabling numbered capturing parentheses. Any opening parenthesis that is
3068 not followed by ? behaves as if it were followed by ?: but named parentheses
3069 can still be used for capturing (and they will acquire numbers in the usual
3070 way).
3071
3072 41. Redesigned the return codes from the match() function into yes/no/error so
3073 that errors can be passed back from deep inside the nested calls. A malloc
3074 failure while inside a recursive subpattern call now causes the
3075 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3076
3077 42. It is now possible to set a limit on the number of times the match()
3078 function is called in a call to pcre_exec(). This facility makes it possible to
3079 limit the amount of recursion and backtracking, though not in a directly
3080 obvious way, because the match() function is used in a number of different
3081 circumstances. The count starts from zero for each position in the subject
3082 string (for non-anchored patterns). The default limit is, for compatibility, a
3083 large number, namely 10 000 000. You can change this in two ways:
3084
3085 (a) When configuring PCRE before making, you can use --with-match-limit=n
3086 to set a default value for the compiled library.
3087
3088 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3089 a different value is set. See 45 below.
3090
3091 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3092
3093 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3094 of things that can be changed at compile time. The first argument specifies
3095 what is wanted and the second points to where the information is to be placed.
3096 The current list of available information is:
3097
3098 PCRE_CONFIG_UTF8
3099
3100 The output is an integer that is set to one if UTF-8 support is available;
3101 otherwise it is set to zero.
3102
3103 PCRE_CONFIG_NEWLINE
3104
3105 The output is an integer that it set to the value of the code that is used for
3106 newline. It is either LF (10) or CR (13).
3107
3108 PCRE_CONFIG_LINK_SIZE
3109
3110 The output is an integer that contains the number of bytes used for internal
3111 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3112
3113 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3114
3115 The output is an integer that contains the threshold above which the POSIX
3116 interface uses malloc() for output vectors. See item 31 above.
3117
3118 PCRE_CONFIG_MATCH_LIMIT
3119
3120 The output is an unsigned integer that contains the default limit of the number
3121 of match() calls in a pcre_exec() execution. See 42 above.
3122
3123 44. pcretest has been upgraded by the addition of the -C option. This causes it
3124 to extract all the available output from the new pcre_config() function, and to
3125 output it. The program then exits immediately.
3126
3127 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3128 order to support additional features. One way would have been to define
3129 pcre_exec2() (for example) with extra arguments, but this would not have been
3130 extensible, and would also have required all calls to the original function to
3131 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3132 is used for passing in "extra" data from pcre_study().
3133
3134 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3135 contains the following fields:
3136
3137 flags a bitmap indicating which of the following fields are set
3138 study_data opaque data from pcre_study()
3139 match_limit a way of specifying a limit on match() calls for a specific
3140 call to pcre_exec()
3141 callout_data data for callouts (see 49 below)
3142
3143 The flag bits are also defined in pcre.h, and are
3144
3145 PCRE_EXTRA_STUDY_DATA
3146 PCRE_EXTRA_MATCH_LIMIT
3147 PCRE_EXTRA_CALLOUT_DATA
3148
3149 The pcre_study() function now returns one of these new pcre_extra blocks, with
3150 the actual study data pointed to by the study_data field, and the
3151 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3152 before. That is, this change is entirely upwards-compatible and requires no
3153 change to existing code.
3154
3155 If you want to pass in additional data to pcre_exec(), you can either place it
3156 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3157 block.
3158
3159 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3160 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3161 times with different match limits, until it finds the minimum value needed for
3162 pcre_exec() to complete. The value is then output. This can be instructive; for
3163 most simple matches the number is quite small, but for pathological cases it
3164 gets very large very quickly.
3165
3166 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3167 returns the size of the data block pointed to by the study_data field in a
3168 pcre_extra block, that is, the value that was passed as the argument to
3169 pcre_malloc() when PCRE was getting memory in which to place the information
3170 created by pcre_study(). The fourth argument should point to a size_t variable.
3171 pcretest has been extended so that this information is shown after a successful
3172 pcre_study() call when information about the compiled regex is being displayed.
3173
3174 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3175 because what follows is always an absolute path. (Later: it turns out that this
3176 is more than cosmetic for MinGW, because it doesn't like empty path
3177 components.)
3178
3179 49. Some changes have been made to the callout feature (see 28 above):
3180
3181 (i) A callout function now has three choices for what it returns:
3182
3183 0 => success, carry on matching
3184 > 0 => failure at this point, but backtrack if possible
3185 < 0 => serious error, return this value from pcre_exec()
3186
3187 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3188 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3189 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3190 use by callout functions. It will never be used by PCRE itself.
3191
3192 (ii) The pcre_extra structure (see 45 above) has a void * field called
3193 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3194 pcre_callout_block structure has a field of the same name. The contents of
3195 the field passed in the pcre_extra structure are passed to the callout
3196 function in the corresponding field in the callout block. This makes it
3197 easier to use the same callout-containing regex from multiple threads. For
3198 testing, the pcretest program has a new data escape
3199
3200 \C*n pass the number n (may be negative) as callout_data
3201
3202 If the callout function in pcretest receives a non-zero value as
3203 callout_data, it returns that value.
3204
3205 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3206 there were some redundant $(CFLAGS) in commands that are now specified as
3207 $(LINK), which already includes $(CFLAGS).
3208
3209 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3210 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3211 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3212 one-byte characters throughout. Note that case-insensitive matching applies
3213 only to characters whose values are less than 256. PCRE doesn't support the
3214 notion of cases for higher-valued characters.
3215
3216 (i) A character class whose characters are all within 0-255 is handled as
3217 a bit map, and the map is inverted for negative classes. Previously, a
3218 character > 255 always failed to match such a class; however it should
3219 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3220
3221 (ii) A negated character class with a single character < 255 is coded as
3222 "not this character" (OP_NOT). This wasn't working properly when the test
3223 character was multibyte, either singly or repeated.
3224
3225 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3226 mode, for example: \x{100}{2,3}.
3227
3228 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3229 singly or repeated) now correctly test multibyte characters. However,
3230 PCRE doesn't recognize any characters with values greater than 255 as
3231 digits, spaces, or word characters. Such characters always match \D, \S,
3232 and \W, and never match \d, \s, or \w.
3233
3234 (v) Classes may now contain characters and character ranges with values
3235 greater than 255. For example: [ab\x{100}-\x{400}].
3236
3237 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3238 PCRE in UTF-8 mode.
3239
3240 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3241 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3242 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3243 value.)
3244
3245 53. The single man page has become too large. I have therefore split it up into
3246 a number of separate man pages. These also give rise to individual HTML pages;
3247 these are now put in a separate directory, and there is an index.html page that
3248 lists them all. Some hyperlinking between the pages has been installed.
3249
3250 54. Added convenience functions for handling named capturing parentheses.
3251
3252 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3253 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3254 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3255 are faulted.
3256
3257 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3258 calling configure. These values are used when compiling the dftables.c program
3259 which is run to generate the source of the default character tables. They
3260 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3261 you will need to set these values.
3262
3263 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3264
3265
3266 Version 3.9 02-Jan-02
3267 ---------------------
3268
3269 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3270
3271 2. If --disable-static was given, the building process failed when trying to
3272 build pcretest and pcregrep. (For some reason it was using libtool to compile
3273 them, which is not right, as they aren't part of the library.)
3274
3275
3276 Version 3.8 18-Dec-01
3277 ---------------------
3278
3279 1. The experimental UTF-8 code was completely screwed up. It was packing the
3280 bytes in the wrong order. How dumb can you get?
3281
3282
3283 Version 3.7 29-Oct-01
3284 ---------------------
3285
3286 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3287 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3288 this didn't happen under Solaris 8, where I normally test things.
3289
3290 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3291 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3292 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3293 this fix an hour or so after the initial 3.7 release.)
3294
3295
3296 Version 3.6 23-Oct-01
3297 ---------------------
3298
3299 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3300 offsets passed as NULL with zero offset count.
3301
3302 2. The config.guess and config.sub files had not been updated when I moved to
3303 the latest autoconf.
3304
3305
3306 Version 3.5 15-Aug-01
3307 ---------------------
3308
3309 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3310 had been forgotten.
3311
3312 2. By using declared but undefined structures, we can avoid using "void"
3313 definitions in pcre.h while keeping the internal definitions of the structures
3314 private.
3315
3316 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3317 user point of view, this means that both static and shared libraries are built
3318 by default, but this can be individually controlled. More of the work of
3319 handling this static/shared cases is now inside libtool instead of PCRE's make
3320 file.
3321
3322 4. The pcretest utility is now installed along with pcregrep because it is
3323 useful for users (to test regexs) and by doing this, it automatically gets
3324 relinked by libtool. The documentation has been turned into a man page, so
3325 there are now .1, .txt, and .html versions in /doc.
3326
3327 5. Upgrades to pcregrep:
3328 (i) Added long-form option names like gnu grep.
3329 (ii) Added --help to list all options with an explanatory phrase.
3330 (iii) Added -r, --recursive to recurse into sub-directories.
3331 (iv) Added -f, --file to read patterns from a file.
3332
3333 6. pcre_exec() was referring to its "code" argument before testing that
3334 argument for NULL (and giving an error if it was NULL).
3335
3336 7. Upgraded Makefile.in to allow for compiling in a different directory from
3337 the source directory.
3338
3339 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3340 options bits, the pointer it was passed was to an int instead of to an unsigned
3341 long int. This mattered only on 64-bit systems.
3342
3343 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3344 generated) instead of pcre.in, which it its source. Also made the same change
3345 in several of the .c files.
3346
3347 10. A new release of gcc defines printf() as a macro, which broke pcretest
3348 because it had an ifdef in the middle of a string argument for printf(). Fixed
3349 by using separate calls to printf().
3350
3351 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3352 script, to force use of CR or LF instead of \n in the source. On non-Unix
3353 systems, the value can be set in config.h.
3354
3355 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3356 absolute limit. Changed the text of the error message to make this clear, and
3357 likewise updated the man page.
3358
3359 13. The limit of 99 on the number of capturing subpatterns has been removed.
3360 The new limit is 65535, which I hope will not be a "real" limit.
3361
3362
3363 Version 3.4 22-Aug-00
3364 ---------------------
3365
3366 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3367
3368 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3369
3370
3371 Version 3.3 01-Aug-00
3372 ---------------------
3373
3374 1. If an octal character was given, but the value was greater than \377, it
3375 was not getting masked to the least significant bits, as documented. This could
3376 lead to crashes in some systems.
3377
3378 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3379 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3380
3381 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3382 These just pass their arguments on to (pcre_free)(), but they are provided
3383 because some uses of PCRE bind it to non-C systems that can call its functions,
3384 but cannot call free() or pcre_free() directly.
3385
3386 4. Add "make test" as a synonym for "make check". Corrected some comments in
3387 the Makefile.
3388
3389 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3390 Makefile.
3391
3392 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3393 command called pgrep for grepping around the active processes.
3394
3395 7. Added the beginnings of support for UTF-8 character strings.
3396
3397 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3398 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3399 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3400 out for the ar command.)
3401
3402
3403 Version 3.2 12-May-00
3404 ---------------------
3405
3406 This is purely a bug fixing release.
3407
3408 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3409 of ZA. This was just one example of several cases that could provoke this bug,
3410 which was introduced by change 9 of version 2.00. The code for breaking
3411 infinite loops after an iteration that matches an empty string was't working
3412 correctly.
3413
3414 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3415 when matched against abbab (for example). After matching an empty string, it
3416 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3417 caused it to match further down the string than it should.
3418
3419 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3420 was there because it doesn't seem to be needed, and it causes trouble on some
3421 systems, as it is not a Standard C header. It has been removed.
3422
3423 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3424 were reported on the Macintosh. The changes were from
3425
3426 while ((c = *(++ptr)) != 0 && c != '\n');
3427 to
3428 while ((c = *(++ptr)) != 0 && c != '\n') ;
3429
3430 Totally extraordinary, but if that's what it takes...
3431
3432 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3433 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3434 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3435 assumes the way PCRE uses memmove() (always moving upwards).
3436
3437 6. PCRE is being used in one environment where strchr() is not available. There
3438 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3439 faster code anyway.
3440
3441
3442 Version 3.1 09-Feb-00
3443 ---------------------
3444
3445 The only change in this release is the fixing of some bugs in Makefile.in for
3446 the "install" target:
3447
3448 (1) It was failing to install pcreposix.h.
3449
3450 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3451
3452
3453 Version 3.0 01-Feb-00
3454 ---------------------
3455
3456 1. Add support for the /+ modifier to perltest (to output $` like it does in
3457 pcretest).
3458
3459 2. Add support for the /g modifier to perltest.
3460
3461 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3462 matches null strings.
3463
3464 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3465 pattern. Perl treats empty patterns specially - it reuses the most recent
3466 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3467 effect.
3468
3469 5. The POSIX interface was broken in that it was just handing over the POSIX
3470 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3471 required a bigger vector, with some working space on the end. This means that
3472 the POSIX wrapper now has to get and free some memory, and copy the results.
3473
3474 6. Added some simple autoconf support, placing the test data and the
3475 documentation in separate directories, re-organizing some of the
3476 information files, and making it build pcre-config (a GNU standard). Also added
3477 libtool support for building PCRE as a shared library, which is now the
3478 default.
3479
3480 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3481 09 are not valid octal constants. Single digits will be used for minor values
3482 less than 10.
3483
3484 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3485 existing programs that set these in the POSIX interface can use PCRE without
3486 modification.
3487
3488 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3489 return all that pcre_info() returns, plus additional data. The pcre_info()
3490 function is retained for compatibility, but is considered to be obsolete.
3491
3492 10. Added experimental recursion feature (?R) to handle one common case that
3493 Perl 5.6 will be able to do with (?p{...}).
3494
3495 11. Added support for POSIX character classes like [:alpha:], which Perl is
3496 adopting.
3497
3498
3499 Version 2.08 31-Aug-99
3500 ----------------------
3501
3502 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3503 trying to match at the startoffset position, but instead was moving forward to
3504 the next newline as if a previous match had failed.
3505
3506 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3507 and could get into a loop if a null string was matched other than at the start
3508 of the subject.
3509
3510 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3511 be distinguished at compile time, and for completeness also added PCRE_DATE.
3512
3513 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3514 in GnuWin32 environments.
3515
3516
3517 Version 2.07 29-Jul-99
3518 ----------------------
3519
3520 1. The documentation is now supplied in plain text form and HTML as well as in
3521 the form of man page sources.
3522
3523 2. C++ compilers don't like assigning (void *) values to other pointer types.
3524 In particular this affects malloc(). Although there is no problem in Standard
3525 C, I've put in casts to keep C++ compilers happy.
3526
3527 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3528 should be (const char *).
3529
3530 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3531 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3532 However, I haven't made this a standard facility. The documentation doesn't
3533 mention it, and the Makefile doesn't support it.
3534
3535 5. The Makefile now contains an "install" target, with editable destinations at
3536 the top of the file. The pcretest program is not installed.
3537
3538 6. pgrep -V now gives the PCRE version number and date.
3539
3540 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3541 causing the entire string to be ignored, instead of just the last character.
3542
3543 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3544 non-matching string, it can take a very, very long time, even for strings of
3545 quite modest length, because of the nested recursion. PCRE now does better in
3546 some of these cases. It does this by remembering the last required literal
3547 character in the pattern, and pre-searching the subject to ensure it is present
3548 before running the real match. In other words, it applies a heuristic to detect
3549 some types of certain failure quickly, and in the above example, if presented
3550 with a string that has no trailing " it gives "no match" very quickly.
3551
3552 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3553 other alternatives are tried instead.
3554
3555
3556 Version 2.06 09-Jun-99
3557 ----------------------
3558
3559 1. Change pcretest's output for amount of store used to show just the code
3560 space, because the remainder (the data block) varies in size between 32-bit and
3561 64-bit systems.
3562
3563 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3564 start matching at. This allows lookbehinds to work when searching for multiple
3565 occurrences in a string.
3566
3567 3. Added additional options to pcretest for testing multiple occurrences:
3568
3569 /+ outputs the rest of the string that follows a match
3570 /g loops for multiple occurrences, using the new startoffset argument
3571 /G loops for multiple occurrences by passing an incremented pointer
3572
3573 4. PCRE wasn't doing the "first character" optimization for patterns starting
3574 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3575 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3576 the letter 'x'. On long subject strings, this gives a significant speed-up.
3577
3578
3579 Version 2.05 21-Apr-99
3580 ----------------------
3581
3582 1. Changed the type of magic_number from int to long int so that it works
3583 properly on 16-bit systems.
3584
3585 2. Fixed a bug which caused patterns starting with .* not to work correctly
3586 when the subject string contained newline characters. PCRE was assuming
3587 anchoring for such patterns in all cases, which is not correct because .* will
3588 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3589 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3590 must be retried after every newline in the subject.
3591
3592
3593 Version 2.04 18-Feb-99
3594 ----------------------
3595
3596 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3597 computation of the store needed to hold the pattern was incorrect (too large).
3598 If such patterns were nested a few deep, this could multiply and become a real
3599 problem.
3600
3601 2. Added /M option to pcretest to show the memory requirement of a specific
3602 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3603
3604 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3605 compiled in such a way that the backtracking after subsequent failure was
3606 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3607 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3608
3609
3610 Version 2.03 02-Feb-99
3611 ----------------------
3612
3613 1. Fixed typo and small mistake in man page.
3614
3615 2. Added 4th condition (GPL supersedes if conflict) and created separate
3616 LICENCE file containing the conditions.
3617
3618 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3619 Perl, that is the internal \ allows the delimiter to be included in the
3620 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3621 the final delimiter, add \ to the end of the pattern (to test the error).
3622
3623 4. Added the convenience functions for extracting substrings after a successful
3624 match. Updated pcretest to make it able to test these functions.
3625
3626
3627 Version 2.02 14-Jan-99
3628 ----------------------
3629
3630 1. Initialized the working variables associated with each extraction so that
3631 their saving and restoring doesn't refer to uninitialized store.
3632
3633 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3634 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3635 fix the problem.
3636
3637 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3638 calls, and wasn't printing the correct value for compiling calls. Increased the
3639 default value of LOOPREPEAT, and the number of significant figures in the
3640 times.
3641
3642 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3643
3644 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3645 a building problem on Windows NT with a FAT file system.
3646
3647
3648 Version 2.01 21-Oct-98
3649 ----------------------
3650
3651 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3652 to character tables built by pcre_maketables() in the current locale. If NULL
3653 is passed, the default tables are used.
3654
3655
3656 Version 2.00 24-Sep-98
3657 ----------------------
3658
3659 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3660 it any more.
3661
3662 2. Allow quantification of (?>) groups, and make it work correctly.
3663
3664 3. The first character computation wasn't working for (?>) groups.
3665
3666 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3667 end of the subject) and add 5.005's \z, which really does match only at the
3668 very end of the subject.
3669
3670 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3671
3672 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3673 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3674 localized options. All options to pcre_study() were also removed.
3675
3676 7. Add other new features from 5.005:
3677
3678 $(?<= positive lookbehind
3679 $(?<! negative lookbehind
3680 (?imsx-imsx) added the unsetting capability
3681 such a setting is global if at outer level; local otherwise
3682 (?imsx-imsx:) non-capturing groups with option setting
3683 (?(cond)re|re) conditional pattern matching
3684
3685 A backreference to itself in a repeated group matches the previous
3686 captured string.
3687
3688 8. General tidying up of studying (both automatic and via "study")
3689 consequential on the addition of new assertions.
3690
3691 9. As in 5.005, unlimited repeated groups that could match an empty substring
3692 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3693 runtime if any iteration does actually match an empty substring.
3694
3695 10. Include the RunTest script in the distribution.
3696
3697 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3698 discrepancies, some of which were old and were also with respect to 5.004. They
3699 have now been fixed.
3700
3701
3702 Version 1.09 28-Apr-98
3703 ----------------------
3704
3705 1. A negated single character class followed by a quantifier with a minimum
3706 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3707 program crashes, or just wrong answers. This did not apply to negated classes
3708 containing more than one character, or to minima other than one.
3709
3710
3711 Version 1.08 27-Mar-98
3712 ----------------------
3713
3714 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3715
3716 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3717 latter must appear before anything that relies on it in the pattern.
3718
3719
3720 Version 1.07 16-Feb-98
3721 ----------------------
3722
3723 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3724 repeat of a potentially empty string).
3725
3726
3727 Version 1.06 23-Jan-98
3728 ----------------------
3729
3730 1. Added Markus Oberhumer's little patches for C++.
3731
3732 2. Literal strings longer than 255 characters were broken.
3733
3734
3735 Version 1.05 23-Dec-97
3736 ----------------------
3737
3738 1. Negated character classes containing more than one character were failing if
3739 PCRE_CASELESS was set at run time.
3740
3741
3742 Version 1.04 19-Dec-97
3743 ----------------------
3744
3745 1. Corrected the man page, where some "const" qualifiers had been omitted.
3746
3747 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3748 input syntax.
3749
3750 3. Fixed memory leak which occurred when a regex with back references was
3751 matched with an offsets vector that wasn't big enough. The temporary memory
3752 that is used in this case wasn't being freed if the match failed.
3753
3754 4. Tidied pcretest to ensure it frees memory that it gets.
3755
3756 5. Temporary memory was being obtained in the case where the passed offsets
3757 vector was exactly big enough.
3758
3759 6. Corrected definition of offsetof() from change 5 below.
3760
3761 7. I had screwed up change 6 below and broken the rules for the use of
3762 setjmp(). Now fixed.
3763
3764
3765 Version 1.03 18-Dec-97
3766 ----------------------
3767
3768 1. A erroneous regex with a missing opening parenthesis was correctly
3769 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3770 on some systems.
3771
3772 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3773 it was reported that one broken compiler failed on the former because "code" is
3774 also an independent variable.
3775
3776 3. The erroneous regex a[]b caused an array overrun reference.
3777
3778 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3779 fail on data ending with that character. (It was going on too far, and checking
3780 the next character, typically a binary zero.) This was specific to the
3781 optimized code for single-character negative classes.
3782
3783 5. Added a contributed patch from the TIN world which does the following:
3784
3785 + Add an undef for memmove, in case the the system defines a macro for it.
3786
3787 + Add a definition of offsetof(), in case there isn't one. (I don't know
3788 the reason behind this - offsetof() is part of the ANSI standard - but
3789 it does no harm).
3790
3791 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3792 most of the places where whitespace preceded '#'. I have given up and
3793 allowed the remaining 2 cases to be at the margin.
3794
3795 + Rename some variables in pcre to eliminate shadowing. This seems very
3796 pedantic, but does no harm, of course.
3797
3798 6. Moved the call to setjmp() into its own function, to get rid of warnings
3799 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3800
3801 7. Constructs such as \d{8,} were compiling into the equivalent of
3802 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3803 outcome, but in this particular case used more store than had been allocated,
3804 which caused the bug to be discovered because it threw up an internal error.
3805
3806 8. The debugging code in both pcre and pcretest for outputting the compiled
3807 form of a regex was going wrong in the case of back references followed by
3808 curly-bracketed repeats.
3809
3810
3811 Version 1.02 12-Dec-97
3812 ----------------------
3813
3814 1. Typos in pcre.3 and comments in the source fixed.
3815
3816 2. Applied a contributed patch to get rid of places where it used to remove
3817 'const' from variables, and fixed some signed/unsigned and uninitialized
3818 variable warnings.
3819
3820 3. Added the "runtest" target to Makefile.
3821
3822 4. Set default compiler flag to -O2 rather than just -O.
3823
3824
3825 Version 1.01 19-Nov-97
3826 ----------------------
3827
3828 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3829 like /([ab]*)*/, that is, for classes with more than one character in them.
3830
3831 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3832 as /((?>a*))*/ (a PCRE_EXTRA facility).
3833
3834
3835 Version 1.00 18-Nov-97
3836 ----------------------
3837
3838 1. Added compile-time macros to support systems such as SunOS4 which don't have
3839 memmove() or strerror() but have other things that can be used instead.
3840
3841 2. Arranged that "make clean" removes the executables.
3842
3843
3844 Version 0.99 27-Oct-97
3845 ----------------------
3846
3847 1. Fixed bug in code for optimizing classes with only one character. It was
3848 initializing a 32-byte map regardless, which could cause it to run off the end
3849 of the memory it had got.
3850
3851 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3852
3853
3854 Version 0.98 22-Oct-97
3855 ----------------------
3856
3857 1. Fixed bug in code for handling temporary memory usage when there are more
3858 back references than supplied space in the ovector. This could cause segfaults.
3859
3860
3861 Version 0.97 21-Oct-97
3862 ----------------------
3863
3864 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3865
3866 2. Optimized negated single characters not to use a bit map.
3867
3868 3. Brought error texts together as macro definitions; clarified some of them;
3869 fixed one that was wrong - it said "range out of order" when it meant "invalid
3870 escape sequence".
3871
3872 4. Changed some char * arguments to const char *.
3873
3874 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3875
3876 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3877 pcretest.
3878
3879
3880 Version 0.96 16-Oct-97
3881 ----------------------
3882
3883 1. Added a simple "pgrep" utility to the distribution.
3884
3885 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3886 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3887 where "ddd" means "one or more decimal digits".
3888
3889 3. Fixed serious bug. If a pattern had a back reference, but the call to
3890 pcre_exec() didn't supply a large enough ovector to record the related
3891 identifying subpattern, the match always failed. PCRE now remembers the number
3892 of the largest back reference, and gets some temporary memory in which to save
3893 the offsets during matching if necessary, in order to ensure that
3894 backreferences always work.
3895
3896 4. Increased the compatibility with Perl in a number of ways:
3897
3898 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3899 to request this handling. The option can be set at compile or exec time.
3900
3901 (b) $ matches before a terminating newline by default; an option
3902 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3903 mode). The option can be set at compile or exec time.
3904
3905 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3906 the same as Perl's. If the decimal number it represents is less than 10
3907 or there aren't that many previous left capturing parentheses, an octal
3908 escape is read. Inside a character class, it's always an octal escape,
3909 even if it is a single digit.
3910
3911 (d) An escaped but undefined alphabetic character is taken as a literal,
3912 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3913 escapes.
3914
3915 (e) {0} is now permitted. (The previous item is removed from the compiled
3916 pattern).
3917
3918 5. Changed all the names of code files so that the basic parts are no longer
3919 than 10 characters, and abolished the teeny "globals.c" file.
3920
3921 6. Changed the handling of character classes; they are now done with a 32-byte
3922 bit map always.
3923
3924 7. Added the -d and /D options to pcretest to make it possible to look at the
3925 internals of compilation without having to recompile pcre.
3926
3927
3928 Version 0.95 23-Sep-97
3929 ----------------------
3930
3931 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3932 \x20 at the start of a run of normal characters. These were being treated as
3933 real characters, instead of the source characters being re-checked.
3934
3935
3936 Version 0.94 18-Sep-97
3937 ----------------------
3938
3939 1. The functions are now thread-safe, with the caveat that the global variables
3940 containing pointers to malloc() and free() or alternative functions are the
3941 same for all threads.
3942
3943 2. Get pcre_study() to generate a bitmap of initial characters for non-
3944 anchored patterns when this is possible, and use it if passed to pcre_exec().
3945
3946
3947 Version 0.93 15-Sep-97
3948 ----------------------
3949
3950 1. /(b)|(:+)/ was computing an incorrect first character.
3951
3952 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3953 but not actually doing anything yet.
3954
3955 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3956 as Perl does (e.g. [-az] or [az-]).
3957
3958 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3959 all possible positions.
3960
3961 5. Split up into different modules to avoid including unneeded functions in a
3962 compiled binary. However, compile and exec are still in one module. The "study"
3963 function is split off.
3964
3965 6. The character tables are now in a separate module whose source is generated
3966 by an auxiliary program - but can then be edited by hand if required. There are
3967 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3968 toupper() in the code.
3969
3970 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3971 make them global. Abolish the function for setting them, as the caller can now
3972 set them directly.
3973
3974
3975 Version 0.92 11-Sep-97
3976 ----------------------
3977
3978 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3979 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3980
3981 2. Caseless matching was not working in character classes if the characters in
3982 the pattern were in upper case.
3983
3984 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3985
3986 4. Make PCRE_ANCHORED public and accept as a compile option.
3987
3988 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3989 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3990 pass them.
3991
3992 6. Give an error if bad option bits passed at compile or run time.
3993
3994 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3995 pcretest to cause it to pass that flag.
3996
3997 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3998 options, and the first character, if set.
3999
4000 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4001
4002
4003 Version 0.91 10-Sep-97
4004 ----------------------
4005
4006 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4007 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4008
4009 2. PCRE was looping on encountering an indefinitely repeated back reference to
4010 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4011 Perl does - treats the match as successful.
4012
4013 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5