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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 558 - (show annotations)
Tue Oct 26 15:26:45 2010 UTC (4 years, 6 months ago) by ph10
File size: 192788 byte(s)
Error occurred while calculating annotation data.
Fix missing code for missing strtoul() and strerror(). 
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 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
57 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
58 cater for a lack of strerror(). These oversights have been fixed.
59
60
61 Version 8.10 25-Jun-2010
62 ------------------------
63
64 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
65 THEN.
66
67 2. (*ACCEPT) was not working when inside an atomic group.
68
69 3. Inside a character class, \B is treated as a literal by default, but
70 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
71 causes the error). The code is unchanged, but I tidied the documentation.
72
73 4. Inside a character class, PCRE always treated \R and \X as literals,
74 whereas Perl faults them if its -w option is set. I have changed PCRE so
75 that it faults them when PCRE_EXTRA is set.
76
77 5. Added support for \N, which always matches any character other than
78 newline. (It is the same as "." when PCRE_DOTALL is not set.)
79
80 6. When compiling pcregrep with newer versions of gcc which may have
81 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
82 declared with attribute warn_unused_result" were given. Just casting the
83 result to (void) does not stop the warnings; a more elaborate fudge is
84 needed. I've used a macro to implement this.
85
86 7. Minor change to pcretest.c to avoid a compiler warning.
87
88 8. Added four artifical Unicode properties to help with an option to make
89 \s etc use properties (see next item). The new properties are: Xan
90 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
91
92 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
93 use Unicode properties. (*UCP) at the start of a pattern can be used to set
94 this option. Modified pcretest to add /W to test this facility. Added
95 REG_UCP to make it available via the POSIX interface.
96
97 10. Added --line-buffered to pcregrep.
98
99 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
100 studied, and the match started with a letter with a code point greater than
101 127 whose first byte was different to the first byte of the other case of
102 the letter, the other case of this starting letter was not recognized
103 (#976).
104
105 12. If a pattern that was studied started with a repeated Unicode property
106 test, for example, \p{Nd}+, there was the theoretical possibility of
107 setting up an incorrect bitmap of starting bytes, but fortunately it could
108 not have actually happened in practice until change 8 above was made (it
109 added property types that matched character-matching opcodes).
110
111 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
112 possible starting bytes for non-anchored patterns.
113
114 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
115 \R, and also a number of cases that involve Unicode properties, both
116 explicit and implicit when PCRE_UCP is set.
117
118 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
119 input, it could crash or give wrong results if characters with values
120 greater than 0xc0 were present in the subject string. (Detail: it assumed
121 UTF-8 input when processing these items.)
122
123 16. Added a lot of (int) casts to avoid compiler warnings in systems where
124 size_t is 64-bit (#991).
125
126 17. Added a check for running out of memory when PCRE is compiled with
127 --disable-stack-for-recursion (#990).
128
129 18. If the last data line in a file for pcretest does not have a newline on
130 the end, a newline was missing in the output.
131
132 19. The default pcre_chartables.c file recognizes only ASCII characters (values
133 less than 128) in its various bitmaps. However, there is a facility for
134 generating tables according to the current locale when PCRE is compiled. It
135 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
136 characters, are recognized by isspace() and therefore were getting set in
137 these tables, and indeed these tables seem to approximate to ISO 8859. This
138 caused a problem in UTF-8 mode when pcre_study() was used to create a list
139 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
140 which of course cannot start UTF-8 characters. I have changed the code so
141 that only real ASCII characters (less than 128) and the correct starting
142 bytes for UTF-8 encodings are set for characters greater than 127 when in
143 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
144 altogether.)
145
146 20. Added the /T option to pcretest so as to be able to run tests with non-
147 standard character tables, thus making it possible to include the tests
148 used for 19 above in the standard set of tests.
149
150 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
151 reference to a subpattern the other side of a comment that contains an
152 opening parenthesis caused either an internal compiling error, or a
153 reference to the wrong subpattern.
154
155
156 Version 8.02 19-Mar-2010
157 ------------------------
158
159 1. The Unicode data tables have been updated to Unicode 5.2.0.
160
161 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
162 configured.
163
164 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
165 original author of that file, following a query about its status.
166
167 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
168 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
169
170 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
171 quantifier applied to a forward-referencing subroutine call, could compile
172 incorrect code or give the error "internal error: previously-checked
173 referenced subpattern not found".
174
175 6. Both MS Visual Studio and Symbian OS have problems with initializing
176 variables to point to external functions. For these systems, therefore,
177 pcre_malloc etc. are now initialized to local functions that call the
178 relevant global functions.
179
180 7. There were two entries missing in the vectors called coptable and poptable
181 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
182 I've fixed the data, and added a kludgy way of testing at compile time that
183 the lengths are correct (equal to the number of opcodes).
184
185 8. Following on from 7, I added a similar kludge to check the length of the
186 eint vector in pcreposix.c.
187
188 9. Error texts for pcre_compile() are held as one long string to avoid too
189 much relocation at load time. To find a text, the string is searched,
190 counting zeros. There was no check for running off the end of the string,
191 which could happen if a new error number was added without updating the
192 string.
193
194 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
195
196 11. \K was not working if it appeared in an atomic group or in a group that
197 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
198 \K is "not well defined" if used in an assertion. PCRE now accepts it if
199 the assertion is positive, but not if it is negative.
200
201 12. Change 11 fortuitously reduced the size of the stack frame used in the
202 "match()" function of pcre_exec.c by one pointer. Forthcoming
203 implementation of support for (*MARK) will need an extra pointer on the
204 stack; I have reserved it now, so that the stack frame size does not
205 decrease.
206
207 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
208 item in branch that calls a recursion is a subroutine call - as in the
209 second branch in the above example - was incorrectly given the compile-
210 time error "recursive call could loop indefinitely" because pcre_compile()
211 was not correctly checking the subroutine for matching a non-empty string.
212
213 14. The checks for overrunning compiling workspace could trigger after an
214 overrun had occurred. This is a "should never occur" error, but it can be
215 triggered by pathological patterns such as hundreds of nested parentheses.
216 The checks now trigger 100 bytes before the end of the workspace.
217
218 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
219
220
221 Version 8.01 19-Jan-2010
222 ------------------------
223
224 1. If a pattern contained a conditional subpattern with only one branch (in
225 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
226 computed the wrong minimum data length (which is of course zero for such
227 subpatterns). This could cause incorrect "no match" results.
228
229 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
230 the pattern is reset in the first branch, pcre_compile() failed with
231 "internal error: code overflow at offset...". This happened only when
232 the reset was to the original external option setting. (An optimization
233 abstracts leading options settings into an external setting, which was the
234 cause of this.)
235
236 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
237 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
238 assertion pattern did not match (meaning that the assertion was true), it
239 was incorrectly treated as false if the SKIP had been reached during the
240 matching. This also applied to assertions used as conditions.
241
242 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
243 assertion subpattern, including such a pattern used as a condition,
244 unpredictable results occurred, instead of the error return
245 PCRE_ERROR_DFA_UITEM.
246
247 5. The C++ GlobalReplace function was not working like Perl for the special
248 situation when an empty string is matched. It now does the fancy magic
249 stuff that is necessary.
250
251 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
252 removed. (These were left over from very, very early versions of PCRE.)
253
254 7. Some cosmetic changes to the code to make life easier when compiling it
255 as part of something else:
256
257 (a) Change DEBUG to PCRE_DEBUG.
258
259 (b) In pcre_compile(), rename the member of the "branch_chain" structure
260 called "current" as "current_branch", to prevent a collision with the
261 Linux macro when compiled as a kernel module.
262
263 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
264 prevent a collision with the Linux macro when compiled as a kernel
265 module.
266
267 8. In pcre_compile() there are some checks for integer overflows that used to
268 cast potentially large values to (double). This has been changed to that
269 when building, a check for int64_t is made, and if it is found, it is used
270 instead, thus avoiding the use of floating point arithmetic. (There is no
271 other use of FP in PCRE.) If int64_t is not found, the fallback is to
272 double.
273
274 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
275 2005 (difference between two addresses compared to an unsigned value).
276
277 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
278 custom one, because of the following reported problem in Windows:
279
280 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
281 under Win32.
282 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
283 therefore missing the function definition.
284 - The compiler thus generates a "C" signature for the test function.
285 - The linker fails to find the "C" function.
286 - PCRE fails to configure if asked to do so against libbz2.
287
288 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
289 messages were output:
290
291 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
292 rerunning libtoolize, to keep the correct libtool macros in-tree.
293 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
294
295 I have done both of these things.
296
297 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
298 most of the time, it *can* run out if it is given a pattern that contains a
299 runaway infinite recursion. I updated the discussion in the pcrestack man
300 page.
301
302 13. Now that we have gone to the x.xx style of version numbers, the minor
303 version may start with zero. Using 08 or 09 is a bad idea because users
304 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
305 interpreted as invalid octal numbers. I've updated the previous comment in
306 configure.ac, and also added a check that gives an error if 08 or 09 are
307 used.
308
309 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
310 causing partial matching to fail when the end of the subject matched \W
311 in a UTF-8 pattern where \W was quantified with a minimum of 3.
312
313 15. There were some discrepancies between the declarations in pcre_internal.h
314 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
315 their definitions. The declarations used "const uschar *" and the
316 definitions used USPTR. Even though USPTR is normally defined as "const
317 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
318 reported that: "This difference in casting confuses some C++ compilers, for
319 example, SunCC recognizes above declarations as different functions and
320 generates broken code for hbpcre." I have changed the declarations to use
321 USPTR.
322
323 16. GNU libtool is named differently on some systems. The autogen.sh script now
324 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
325 (FreeBSD).
326
327 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
328 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
329 comment: "Figure out how to create a longlong from a string: strtoll and
330 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
331 instance, but it only takes 2 args instead of 3!"
332
333 18. A subtle bug concerned with back references has been fixed by a change of
334 specification, with a corresponding code fix. A pattern such as
335 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
336 refers, was giving matches when it shouldn't. For example, xa=xaaa would
337 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
338 same bug. Such groups have to be quantified to be useful, or contained
339 inside another quantified group. (If there's no repetition, the reference
340 can never match.) The problem arises because, having left the group and
341 moved on to the rest of the pattern, a later failure that backtracks into
342 the group uses the captured value from the final iteration of the group
343 rather than the correct earlier one. I have fixed this in PCRE by forcing
344 any group that contains a reference to itself to be an atomic group; that
345 is, there cannot be any backtracking into it once it has completed. This is
346 similar to recursive and subroutine calls.
347
348
349 Version 8.00 19-Oct-09
350 ----------------------
351
352 1. The table for translating pcre_compile() error codes into POSIX error codes
353 was out-of-date, and there was no check on the pcre_compile() error code
354 being within the table. This could lead to an OK return being given in
355 error.
356
357 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
358 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
359 in a Windows environment.
360
361 3. The pcregrep --count option prints the count for each file even when it is
362 zero, as does GNU grep. However, pcregrep was also printing all files when
363 --files-with-matches was added. Now, when both options are given, it prints
364 counts only for those files that have at least one match. (GNU grep just
365 prints the file name in this circumstance, but including the count seems
366 more useful - otherwise, why use --count?) Also ensured that the
367 combination -clh just lists non-zero counts, with no names.
368
369 4. The long form of the pcregrep -F option was incorrectly implemented as
370 --fixed_strings instead of --fixed-strings. This is an incompatible change,
371 but it seems right to fix it, and I didn't think it was worth preserving
372 the old behaviour.
373
374 5. The command line items --regex=pattern and --regexp=pattern were not
375 recognized by pcregrep, which required --regex pattern or --regexp pattern
376 (with a space rather than an '='). The man page documented the '=' forms,
377 which are compatible with GNU grep; these now work.
378
379 6. No libpcreposix.pc file was created for pkg-config; there was just
380 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
381
382 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
383 when UCP support is not needed, by modifying the Python script that
384 generates it from Unicode data files. This should not matter if the module
385 is correctly used as a library, but I received one complaint about 50K of
386 unwanted data. My guess is that the person linked everything into his
387 program rather than using a library. Anyway, it does no harm.
388
389 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
390 was a minimum greater than 1 for a wide character in a possessive
391 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
392 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
393 character. Chaos in the form of incorrect output or a compiling loop could
394 result.
395
396 9. The restrictions on what a pattern can contain when partial matching is
397 requested for pcre_exec() have been removed. All patterns can now be
398 partially matched by this function. In addition, if there are at least two
399 slots in the offset vector, the offset of the earliest inspected character
400 for the match and the offset of the end of the subject are set in them when
401 PCRE_ERROR_PARTIAL is returned.
402
403 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
404 synonymous with PCRE_PARTIAL, for backwards compatibility, and
405 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
406 and may be more useful for multi-segment matching.
407
408 11. Partial matching with pcre_exec() is now more intuitive. A partial match
409 used to be given if ever the end of the subject was reached; now it is
410 given only if matching could not proceed because another character was
411 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
412 string "Z", which now yields "no match" instead of "partial match". In the
413 case of pcre_dfa_exec(), "no match" is given if every matching path for the
414 final character ended with (*FAIL).
415
416 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
417 if the pattern had a "must contain" character that was already found in the
418 earlier partial match, unless partial matching was again requested. For
419 example, with the pattern /dog.(body)?/, the "must contain" character is
420 "g". If the first part-match was for the string "dog", restarting with
421 "sbody" failed. This bug has been fixed.
422
423 13. The string returned by pcre_dfa_exec() after a partial match has been
424 changed so that it starts at the first inspected character rather than the
425 first character of the match. This makes a difference only if the pattern
426 starts with a lookbehind assertion or \b or \B (\K is not supported by
427 pcre_dfa_exec()). It's an incompatible change, but it makes the two
428 matching functions compatible, and I think it's the right thing to do.
429
430 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
431 so that the demonstration program is easily available in environments where
432 PCRE has not been installed from source.
433
434 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
435 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
436 library.
437
438 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
439 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
440 is not the first non-POSIX option to be added. Clearly some people find
441 these options useful.
442
443 17. If a caller to the POSIX matching function regexec() passes a non-zero
444 value for nmatch with a NULL value for pmatch, the value of
445 nmatch is forced to zero.
446
447 18. RunGrepTest did not have a test for the availability of the -u option of
448 the diff command, as RunTest does. It now checks in the same way as
449 RunTest, and also checks for the -b option.
450
451 19. If an odd number of negated classes containing just a single character
452 interposed, within parentheses, between a forward reference to a named
453 subpattern and the definition of the subpattern, compilation crashed with
454 an internal error, complaining that it could not find the referenced
455 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
456 [The bug was that it was starting one character too far in when skipping
457 over the character class, thus treating the ] as data rather than
458 terminating the class. This meant it could skip too much.]
459
460 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
461 /g option in pcretest when the pattern contains \K, which makes it possible
462 to have an empty string match not at the start, even when the pattern is
463 anchored. Updated pcretest and pcredemo to use this option.
464
465 21. If the maximum number of capturing subpatterns in a recursion was greater
466 than the maximum at the outer level, the higher number was returned, but
467 with unset values at the outer level. The correct (outer level) value is
468 now given.
469
470 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
471 PCRE did not set those parentheses (unlike Perl). I have now found a way to
472 make it do so. The string so far is captured, making this feature
473 compatible with Perl.
474
475 23. The tests have been re-organized, adding tests 11 and 12, to make it
476 possible to check the Perl 5.10 features against Perl 5.10.
477
478 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
479 pattern matches a fixed length string. PCRE did not allow this; now it
480 does. Neither allows recursion.
481
482 25. I finally figured out how to implement a request to provide the minimum
483 length of subject string that was needed in order to match a given pattern.
484 (It was back references and recursion that I had previously got hung up
485 on.) This code has now been added to pcre_study(); it finds a lower bound
486 to the length of subject needed. It is not necessarily the greatest lower
487 bound, but using it to avoid searching strings that are too short does give
488 some useful speed-ups. The value is available to calling programs via
489 pcre_fullinfo().
490
491 26. While implementing 25, I discovered to my embarrassment that pcretest had
492 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
493 study optimizations had never been tested with that matching function.
494 Oops. What is worse, even when it was passed study data, there was a bug in
495 pcre_dfa_exec() that meant it never actually used it. Double oops. There
496 were also very few tests of studied patterns with pcre_dfa_exec().
497
498 27. If (?| is used to create subpatterns with duplicate numbers, they are now
499 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
500 on the other side of the coin, they are no longer allowed to have different
501 names, because these cannot be distinguished in PCRE, and this has caused
502 confusion. (This is a difference from Perl.)
503
504 28. When duplicate subpattern names are present (necessarily with different
505 numbers, as required by 27 above), and a test is made by name in a
506 conditional pattern, either for a subpattern having been matched, or for
507 recursion in such a pattern, all the associated numbered subpatterns are
508 tested, and the overall condition is true if the condition is true for any
509 one of them. This is the way Perl works, and is also more like the way
510 testing by number works.
511
512
513 Version 7.9 11-Apr-09
514 ---------------------
515
516 1. When building with support for bzlib/zlib (pcregrep) and/or readline
517 (pcretest), all targets were linked against these libraries. This included
518 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
519 libraries. This caused unwanted dependencies to be created. This problem
520 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
521 pcretest is linked with readline.
522
523 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
524 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
525 moved outside it again, because FALSE and TRUE are already defined in AIX,
526 but BOOL is not.
527
528 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
529 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
530
531 4. The pcregrep documentation said spaces were inserted as well as colons (or
532 hyphens) following file names and line numbers when outputting matching
533 lines. This is not true; no spaces are inserted. I have also clarified the
534 wording for the --colour (or --color) option.
535
536 5. In pcregrep, when --colour was used with -o, the list of matching strings
537 was not coloured; this is different to GNU grep, so I have changed it to be
538 the same.
539
540 6. When --colo(u)r was used in pcregrep, only the first matching substring in
541 each matching line was coloured. Now it goes on to look for further matches
542 of any of the test patterns, which is the same behaviour as GNU grep.
543
544 7. A pattern that could match an empty string could cause pcregrep to loop; it
545 doesn't make sense to accept an empty string match in pcregrep, so I have
546 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
547 seems to be how GNU grep behaves.
548
549 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
550 start or after a newline", because the conditional assertion was not being
551 correctly handled. The rule now is that both the assertion and what follows
552 in the first alternative must satisfy the test.
553
554 9. If auto-callout was enabled in a pattern with a conditional group whose
555 condition was an assertion, PCRE could crash during matching, both with
556 pcre_exec() and pcre_dfa_exec().
557
558 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
559 used for matching.
560
561 11. Unicode property support in character classes was not working for
562 characters (bytes) greater than 127 when not in UTF-8 mode.
563
564 12. Added the -M command line option to pcretest.
565
566 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
567
568 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
569
570 16. Added comments and documentation about mis-use of no_arg in the C++
571 wrapper.
572
573 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
574 from Martin Jerabek that uses macro names for all relevant character and
575 string constants.
576
577 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
578 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
579 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
580 these, but not everybody uses configure.
581
582 19. A conditional group that had only one branch was not being correctly
583 recognized as an item that could match an empty string. This meant that an
584 enclosing group might also not be so recognized, causing infinite looping
585 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
586 with the subject "ab", where knowledge that the repeated group can match
587 nothing is needed in order to break the loop.
588
589 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
590 exec(), but without supplying a callout function, matching went wrong.
591
592 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
593 leak if the size of the offset vector was greater than 30. When the vector
594 is smaller, the saved offsets during recursion go onto a local stack
595 vector, but for larger vectors malloc() is used. It was failing to free
596 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
597 error, in fact).
598
599 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
600 heapframe that is used only when UTF-8 support is enabled. This caused no
601 problem, but was untidy.
602
603 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
604 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
605 included within another project.
606
607 24. Steven Van Ingelgem's patches to add more options to the CMake support,
608 slightly modified by me:
609
610 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
611 not building pcregrep.
612
613 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
614 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
615
616 25. Forward references, both numeric and by name, in patterns that made use of
617 duplicate group numbers, could behave incorrectly or give incorrect errors,
618 because when scanning forward to find the reference group, PCRE was not
619 taking into account the duplicate group numbers. A pattern such as
620 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
621
622 26. Changed a few more instances of "const unsigned char *" to USPTR, making
623 the feature of a custom pointer more persuasive (as requested by a user).
624
625 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
626 pcretest.c, inside #ifndefs, because it seems they are sometimes already
627 pre-defined.
628
629 28. Added support for (*UTF8) at the start of a pattern.
630
631 29. Arrange for flags added by the "release type" setting in CMake to be shown
632 in the configuration summary.
633
634
635 Version 7.8 05-Sep-08
636 ---------------------
637
638 1. Replaced UCP searching code with optimized version as implemented for Ad
639 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
640 stage table and inline lookup instead of a function, giving speed ups of 2
641 to 5 times on some simple patterns that I tested. Permission was given to
642 distribute the MultiStage2.py script that generates the tables (it's not in
643 the tarball, but is in the Subversion repository).
644
645 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
646 scripts.
647
648 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
649 a group with a zero qualifier. The result of the study could be incorrect,
650 or the function might crash, depending on the pattern.
651
652 4. Caseless matching was not working for non-ASCII characters in back
653 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
654 It now works when Unicode Property Support is available.
655
656 5. In pcretest, an escape such as \x{de} in the data was always generating
657 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
658 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
659 truncation.
660
661 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
662
663 7. Added two (int) casts to pcregrep when printing the difference of two
664 pointers, in case they are 64-bit values.
665
666 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
667 test 2 if it fails.
668
669 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
670 and a #define of that name to empty if it is not externally set. This is to
671 allow users of MSVC to set it if necessary.
672
673 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
674 the convenience functions in the pcre_get.c source file.
675
676 11. An option change at the start of a pattern that had top-level alternatives
677 could cause overwriting and/or a crash. This command provoked a crash in
678 some environments:
679
680 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
681
682 This potential security problem was recorded as CVE-2008-2371.
683
684 12. For a pattern where the match had to start at the beginning or immediately
685 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
686 pcre_dfa_exec() could read past the end of the passed subject if there was
687 no match. To help with detecting such bugs (e.g. with valgrind), I modified
688 pcretest so that it places the subject at the end of its malloc-ed buffer.
689
690 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
691 exec() might read past the end of the data buffer in UTF-8 mode.
692
693 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
694 the data contained the byte 0x85 as part of a UTF-8 character within its
695 first line. This applied both to normal and DFA matching.
696
697 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
698 /^[^d]*?$/8 failed to match "abc".
699
700 16. Added a missing copyright notice to pcrecpp_internal.h.
701
702 17. Make it more clear in the documentation that values returned from
703 pcre_exec() in ovector are byte offsets, not character counts.
704
705 18. Tidied a few places to stop certain compilers from issuing warnings.
706
707 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
708 supplied by Stefan Weber. I made a further small update for 7.8 because
709 there is a change of source arrangements: the pcre_searchfuncs.c module is
710 replaced by pcre_ucd.c.
711
712
713 Version 7.7 07-May-08
714 ---------------------
715
716 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
717 a string to a long long, pretend we don't even have a long long." This is
718 done by checking for the strtoq, strtoll, and _strtoi64 functions.
719
720 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
721 pre-7.6 versions, which defined a global no_arg variable instead of putting
722 it in the RE class. (See also #8 below.)
723
724 3. Remove a line of dead code, identified by coverity and reported by Nuno
725 Lopes.
726
727 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
728
729 (1) The include/exclude patterns were being applied to the whole pathnames
730 of files, instead of just to the final components.
731
732 (2) If there was more than one level of directory, the subdirectories were
733 skipped unless they satisfied the include/exclude conditions. This is
734 inconsistent with GNU grep (and could even be seen as contrary to the
735 pcregrep specification - which I improved to make it absolutely clear).
736 The action now is always to scan all levels of directory, and just
737 apply the include/exclude patterns to regular files.
738
739 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
740 --exclude_dir in the tests to avoid scanning .svn directories.
741
742 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
743 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
744 doesn't support NULs in patterns.
745
746 7. Added some missing "const"s to declarations of static tables in
747 pcre_compile.c and pcre_dfa_exec.c.
748
749 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
750 caused by fix #2 above. (Subsequently also a second patch to fix the
751 first patch. And a third patch - this was a messy problem.)
752
753 9. Applied Craig's patch to remove the use of push_back().
754
755 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
756 matching function regexec().
757
758 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
759 which, however, unlike Perl's \g{...}, are subroutine calls, not back
760 references. PCRE supports relative numbers with this syntax (I don't think
761 Oniguruma does).
762
763 12. Previously, a group with a zero repeat such as (...){0} was completely
764 omitted from the compiled regex. However, this means that if the group
765 was called as a subroutine from elsewhere in the pattern, things went wrong
766 (an internal error was given). Such groups are now left in the compiled
767 pattern, with a new opcode that causes them to be skipped at execution
768 time.
769
770 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
771 to the way PCRE behaves:
772
773 (a) A lone ] character is dis-allowed (Perl treats it as data).
774
775 (b) A back reference to an unmatched subpattern matches an empty string
776 (Perl fails the current match path).
777
778 (c) A data ] in a character class must be notated as \] because if the
779 first data character in a class is ], it defines an empty class. (In
780 Perl it is not possible to have an empty class.) The empty class []
781 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
782 The negative empty class [^] matches any one character, independently
783 of the DOTALL setting.
784
785 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
786 non-existent subpattern following a character class starting with ']' and
787 containing () gave an internal compiling error instead of "reference to
788 non-existent subpattern". Fortunately, when the pattern did exist, the
789 compiled code was correct. (When scanning forwards to check for the
790 existencd of the subpattern, it was treating the data ']' as terminating
791 the class, so got the count wrong. When actually compiling, the reference
792 was subsequently set up correctly.)
793
794 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
795 it was being rejected as not supported by pcre_dfa_exec(), even though
796 other assertions are supported. I have made pcre_dfa_exec() support
797 (*FAIL).
798
799 16. The implementation of 13c above involved the invention of a new opcode,
800 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
801 cannot be changed at match time, I realized I could make a small
802 improvement to matching performance by compiling OP_ALLANY instead of
803 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
804 on the OP_ANY path.
805
806 17. Compiling pcretest on Windows with readline support failed without the
807 following two fixes: (1) Make the unistd.h include conditional on
808 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
809
810 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
811 ncurses library to be included for pcretest when ReadLine support is
812 requested, but also to allow for it to be overridden. This patch came from
813 Daniel Bergström.
814
815 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
816 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
817 any errors with the current Unicode tables. Thanks to Peter Kankowski for
818 spotting this.
819
820
821 Version 7.6 28-Jan-08
822 ---------------------
823
824 1. A character class containing a very large number of characters with
825 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
826 overflow.
827
828 2. Patch to cut out the "long long" test in pcrecpp_unittest when
829 HAVE_LONG_LONG is not defined.
830
831 3. Applied Christian Ehrlicher's patch to update the CMake build files to
832 bring them up to date and include new features. This patch includes:
833
834 - Fixed PH's badly added libz and libbz2 support.
835 - Fixed a problem with static linking.
836 - Added pcredemo. [But later removed - see 7 below.]
837 - Fixed dftables problem and added an option.
838 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
839 HAVE_LONG_LONG.
840 - Added readline support for pcretest.
841 - Added an listing of the option settings after cmake has run.
842
843 4. A user submitted a patch to Makefile that makes it easy to create
844 "pcre.dll" under mingw when using Configure/Make. I added stuff to
845 Makefile.am that cause it to include this special target, without
846 affecting anything else. Note that the same mingw target plus all
847 the other distribution libraries and programs are now supported
848 when configuring with CMake (see 6 below) instead of with
849 Configure/Make.
850
851 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
852 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
853 exported in the Windows port". It has not yet been confirmed that the patch
854 solves the problem, but it does no harm.
855
856 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
857 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
858 with CMake, and also correct the comment about stack recursion.
859
860 7. Remove the automatic building of pcredemo from the ./configure system and
861 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
862 of a program that users should build themselves after PCRE is installed, so
863 building it automatically is not really right. What is more, it gave
864 trouble in some build environments.
865
866 8. Further tidies to CMakeLists.txt from Sheri and Christian.
867
868
869 Version 7.5 10-Jan-08
870 ---------------------
871
872 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
873 values in parens when parsing an RE using the C++ wrapper."
874
875 2. Negative specials like \S did not work in character classes in UTF-8 mode.
876 Characters greater than 255 were excluded from the class instead of being
877 included.
878
879 3. The same bug as (2) above applied to negated POSIX classes such as
880 [:^space:].
881
882 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
883 defined or documented. It seems to have been a typo for PCRE_STATIC, so
884 I have changed it.
885
886 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
887 first named subpattern) and a construct such as (?&a) would reference the
888 first named subpattern whose name started with "a" (in other words, the
889 length check was missing). Both these problems are fixed. "Subpattern name
890 expected" is now given for (?&) (a zero-length name), and this patch also
891 makes it give the same error for \k'' (previously it complained that that
892 was a reference to a non-existent subpattern).
893
894 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
895 this is right because (?- can be followed by option settings as well as by
896 digits. I have, however, made the messages clearer.
897
898 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
899 than the number used in the conditional) now cause a compile-time error.
900 This is actually not compatible with Perl, which accepts such patterns, but
901 treats the conditional as always being FALSE (as PCRE used to), but it
902 seems to me that giving a diagnostic is better.
903
904 8. Change "alphameric" to the more common word "alphanumeric" in comments
905 and messages.
906
907 9. Fix two occurrences of "backslash" in comments that should have been
908 "backspace".
909
910 10. Remove two redundant lines of code that can never be obeyed (their function
911 was moved elsewhere).
912
913 11. The program that makes PCRE's Unicode character property table had a bug
914 which caused it to generate incorrect table entries for sequences of
915 characters that have the same character type, but are in different scripts.
916 It amalgamated them into a single range, with the script of the first of
917 them. In other words, some characters were in the wrong script. There were
918 thirteen such cases, affecting characters in the following ranges:
919
920 U+002b0 - U+002c1
921 U+0060c - U+0060d
922 U+0061e - U+00612
923 U+0064b - U+0065e
924 U+0074d - U+0076d
925 U+01800 - U+01805
926 U+01d00 - U+01d77
927 U+01d9b - U+01dbf
928 U+0200b - U+0200f
929 U+030fc - U+030fe
930 U+03260 - U+0327f
931 U+0fb46 - U+0fbb1
932 U+10450 - U+1049d
933
934 12. The -o option (show only the matching part of a line) for pcregrep was not
935 compatible with GNU grep in that, if there was more than one match in a
936 line, it showed only the first of them. It now behaves in the same way as
937 GNU grep.
938
939 13. If the -o and -v options were combined for pcregrep, it printed a blank
940 line for every non-matching line. GNU grep prints nothing, and pcregrep now
941 does the same. The return code can be used to tell if there were any
942 non-matching lines.
943
944 14. Added --file-offsets and --line-offsets to pcregrep.
945
946 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
947 infinitely looping recursion. The bug was that positive lookaheads were not
948 being skipped when checking for a possible empty match (negative lookaheads
949 and both kinds of lookbehind were skipped).
950
951 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
952 inclusion of <windows.h> to before rather than after the definition of
953 INVALID_FILE_ATTRIBUTES (patch from David Byron).
954
955 17. Specifying a possessive quantifier with a specific limit for a Unicode
956 character property caused pcre_compile() to compile bad code, which led at
957 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
958 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
959 caused the error; without that there was no problem.
960
961 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
962
963 19. Added --enable-pcretest-libreadline.
964
965 20. In pcrecpp.cc, the variable 'count' was incremented twice in
966 RE::GlobalReplace(). As a result, the number of replacements returned was
967 double what it should be. I removed one of the increments, but Craig sent a
968 later patch that removed the other one (the right fix) and added unit tests
969 that check the return values (which was not done before).
970
971 21. Several CMake things:
972
973 (1) Arranged that, when cmake is used on Unix, the libraries end up with
974 the names libpcre and libpcreposix, not just pcre and pcreposix.
975
976 (2) The above change means that pcretest and pcregrep are now correctly
977 linked with the newly-built libraries, not previously installed ones.
978
979 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
980
981 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
982 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
983 UTF-8 newline character). The key issue is that the pattern starts .*;
984 this means that the match must be either at the beginning, or after a
985 newline. The bug was in the code for advancing after a failed match and
986 checking that the new position followed a newline. It was not taking
987 account of UTF-8 characters correctly.
988
989 23. PCRE was behaving differently from Perl in the way it recognized POSIX
990 character classes. PCRE was not treating the sequence [:...:] as a
991 character class unless the ... were all letters. Perl, however, seems to
992 allow any characters between [: and :], though of course it rejects as
993 unknown any "names" that contain non-letters, because all the known class
994 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
995 for example, whereas PCRE did not - it did not recognize a POSIX character
996 class. This seemed a bit dangerous, so the code has been changed to be
997 closer to Perl. The behaviour is not identical to Perl, because PCRE will
998 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
999 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1000 Perl does, and where it didn't before.
1001
1002 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1003 Windows environments %n is disabled by default.
1004
1005
1006 Version 7.4 21-Sep-07
1007 ---------------------
1008
1009 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1010 means that a class such as [\s] counted as "explicit reference to CR or
1011 LF". That isn't really right - the whole point of the change was to try to
1012 help when there was an actual mention of one of the two characters. So now
1013 the change happens only if \r or \n (or a literal CR or LF) character is
1014 encountered.
1015
1016 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1017 of both had grown to the point where there were only 3 bits left.
1018 Fortunately, there was spare space in the data structure, and so I have
1019 moved the internal flags into a new 16-bit field to free up more option
1020 bits.
1021
1022 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1023 but did not set the internal JCHANGED flag - either of these is enough to
1024 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1025 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1026 start sets both bits.
1027
1028 4. Added options (at build time, compile time, exec time) to change \R from
1029 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1030
1031 5. doc/pcresyntax.html was missing from the distribution.
1032
1033 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1034 compatibility, even though it is no longer used.
1035
1036 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1037 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1038 windows.h file is present (where different names are used). [This was
1039 reversed later after testing - see 16 below.]
1040
1041 8. Changed all #include <config.h> to #include "config.h". There were also
1042 some further <pcre.h> cases that I changed to "pcre.h".
1043
1044 9. When pcregrep was used with the --colour option, it missed the line ending
1045 sequence off the lines that it output.
1046
1047 10. It was pointed out to me that arrays of string pointers cause lots of
1048 relocations when a shared library is dynamically loaded. A technique of
1049 using a single long string with a table of offsets can drastically reduce
1050 these. I have refactored PCRE in four places to do this. The result is
1051 dramatic:
1052
1053 Originally: 290
1054 After changing UCP table: 187
1055 After changing error message table: 43
1056 After changing table of "verbs" 36
1057 After changing table of Posix names 22
1058
1059 Thanks to the folks working on Gregex for glib for this insight.
1060
1061 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1062 unicode-properties was also set.
1063
1064 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1065
1066 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1067 checked only for CRLF.
1068
1069 14. Added casts to pcretest.c to avoid compiler warnings.
1070
1071 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1072
1073 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1074 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1075 entirely. This removes changes made in 7 above.
1076
1077 17. The CMake files have been updated, and there is now more information about
1078 building with CMake in the NON-UNIX-USE document.
1079
1080
1081 Version 7.3 28-Aug-07
1082 ---------------------
1083
1084 1. In the rejigging of the build system that eventually resulted in 7.1, the
1085 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1086 brackets there is not right, since it causes compilers to look for an
1087 installed pcre.h, not the version that is in the source that is being
1088 compiled (which of course may be different). I have changed it back to:
1089
1090 #include "pcre.h"
1091
1092 I have a vague recollection that the change was concerned with compiling in
1093 different directories, but in the new build system, that is taken care of
1094 by the VPATH setting the Makefile.
1095
1096 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1097 when the subject happened to end in the byte 0x85 (e.g. if the last
1098 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1099 characters but of course it shouldn't be taken as a newline when it is part
1100 of another character. The bug was that, for an unlimited repeat of . in
1101 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1102 characters when looking for a newline.
1103
1104 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1105
1106 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1107 in debug output.
1108
1109 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1110 long printing in the pcrecpp unittest when running under MinGW.
1111
1112 6. ESC_K was left out of the EBCDIC table.
1113
1114 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1115 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1116 limit also applies to "virtual nesting" when a pattern is recursive, and in
1117 this case 1000 isn't so big. I have been able to remove this limit at the
1118 expense of backing off one optimization in certain circumstances. Normally,
1119 when pcre_exec() would call its internal match() function recursively and
1120 immediately return the result unconditionally, it uses a "tail recursion"
1121 feature to save stack. However, when a subpattern that can match an empty
1122 string has an unlimited repetition quantifier, it no longer makes this
1123 optimization. That gives it a stack frame in which to save the data for
1124 checking that an empty string has been matched. Previously this was taken
1125 from the 1000-entry workspace that had been reserved. So now there is no
1126 explicit limit, but more stack is used.
1127
1128 8. Applied Daniel's patches to solve problems with the import/export magic
1129 syntax that is required for Windows, and which was going wrong for the
1130 pcreposix and pcrecpp parts of the library. These were overlooked when this
1131 problem was solved for the main library.
1132
1133 9. There were some crude static tests to avoid integer overflow when computing
1134 the size of patterns that contain repeated groups with explicit upper
1135 limits. As the maximum quantifier is 65535, the maximum group length was
1136 set at 30,000 so that the product of these two numbers did not overflow a
1137 32-bit integer. However, it turns out that people want to use groups that
1138 are longer than 30,000 bytes (though not repeat them that many times).
1139 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1140 made it possible to implement the integer overflow checks in a much more
1141 dynamic way, which I have now done. The artificial limitation on group
1142 length has been removed - we now have only the limit on the total length of
1143 the compiled pattern, which depends on the LINK_SIZE setting.
1144
1145 10. Fixed a bug in the documentation for get/copy named substring when
1146 duplicate names are permitted. If none of the named substrings are set, the
1147 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1148 empty string.
1149
1150 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1151 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1152 because the ] is interpreted as the first data character and the
1153 terminating ] is not found. PCRE has been made compatible with Perl in this
1154 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1155 cause memory overwriting.
1156
1157 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1158 string has been matched (to stop an infinite loop). It was not recognizing
1159 a conditional subpattern that could match an empty string if that
1160 subpattern was within another subpattern. For example, it looped when
1161 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1162 condition was not nested. This bug has been fixed.
1163
1164 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1165 past the start of the subject in the presence of bytes with the top bit
1166 set, for example "\x8aBCD".
1167
1168 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1169 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1170
1171 14. Optimized (?!) to (*FAIL).
1172
1173 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1174 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1175 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1176 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1177 does: it's just the validity check that is more restrictive.
1178
1179 16. Inserted checks for integer overflows during escape sequence (backslash)
1180 processing, and also fixed erroneous offset values for syntax errors during
1181 backslash processing.
1182
1183 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1184 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1185
1186 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1187 caused an overrun.
1188
1189 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1190 something other than just ASCII characters) inside a group that had an
1191 unlimited repeat caused a loop at compile time (while checking to see
1192 whether the group could match an empty string).
1193
1194 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1195 [\P{Any}] did so. (Error in the code for printing property names.)
1196
1197 21. An orphan \E inside a character class could cause a crash.
1198
1199 22. A repeated capturing bracket such as (A)? could cause a wild memory
1200 reference during compilation.
1201
1202 23. There are several functions in pcre_compile() that scan along a compiled
1203 expression for various reasons (e.g. to see if it's fixed length for look
1204 behind). There were bugs in these functions when a repeated \p or \P was
1205 present in the pattern. These operators have additional parameters compared
1206 with \d, etc, and these were not being taken into account when moving along
1207 the compiled data. Specifically:
1208
1209 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1210 length.
1211
1212 (b) An item such as \pL+ within a repeated group could cause crashes or
1213 loops.
1214
1215 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1216 "reference to non-existent subpattern" error.
1217
1218 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1219
1220 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1221 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1222
1223 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1224
1225 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1226 character were causing crashes (broken optimization).
1227
1228 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1229 \p or \P) caused a compile-time loop.
1230
1231 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1232 break. For example, the unstudied pattern [\r\n]A does not match the string
1233 "\r\nA" because change 7.0/46 below moves the current point on by two
1234 characters after failing to match at the start. However, the pattern \nA
1235 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1236 the same is true. There doesn't seem any very clean way out of this, but
1237 what I have chosen to do makes the common cases work: PCRE now takes note
1238 of whether there can be an explicit match for \r or \n anywhere in the
1239 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1240 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1241 pattern has explicit CR or LF references.
1242
1243 29. Added (*CR) etc for changing newline setting at start of pattern.
1244
1245
1246 Version 7.2 19-Jun-07
1247 ---------------------
1248
1249 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1250 which is apparently normally available under Windows.
1251
1252 2. Re-jig the pcregrep tests with different newline settings in an attempt
1253 to make them independent of the local environment's newline setting.
1254
1255 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1256
1257 4. Some of the "internals" tests were previously cut out when the link size
1258 was not 2, because the output contained actual offsets. The recent new
1259 "Z" feature of pcretest means that these can be cut out, making the tests
1260 usable with all link sizes.
1261
1262 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1263 stack recursion. This gives a massive performance boost under BSD, but just
1264 a small improvement under Linux. However, it saves one field in the frame
1265 in all cases.
1266
1267 6. Added more features from the forthcoming Perl 5.10:
1268
1269 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1270 recursion call. It refers to the nth most recently opened parentheses.
1271
1272 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1273 to be opened parentheses.
1274
1275 (c) Conditions that refer to capturing parentheses can be specified
1276 relatively, for example, (?(-2)... or (?(+3)...
1277
1278 (d) \K resets the start of the current match so that everything before
1279 is not part of it.
1280
1281 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1282
1283 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1284 reference syntax.
1285
1286 (g) (?| introduces a group in which the numbering of parentheses in each
1287 alternative starts with the same number.
1288
1289 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1290
1291 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1292 PCRE_INFO_JCHANGED.
1293
1294 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1295 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1296 for detecting groups that can match an empty string.
1297
1298 9. A pattern with a very large number of alternatives (more than several
1299 hundred) was running out of internal workspace during the pre-compile
1300 phase, where pcre_compile() figures out how much memory will be needed. A
1301 bit of new cunning has reduced the workspace needed for groups with
1302 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1303 workspace instead of running out of the 4096 that are available.
1304
1305 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1306
1307 11. Applied patch from Google to remove an optimization that didn't quite work.
1308 The report of the bug said:
1309
1310 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1311 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1312 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1313
1314 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1315 it matched the wrong number of bytes.
1316
1317
1318 Version 7.1 24-Apr-07
1319 ---------------------
1320
1321 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1322 that is more "standard", making use of automake and other Autotools. There
1323 is some re-arrangement of the files and adjustment of comments consequent
1324 on this.
1325
1326 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1327 for recursive directory scanning broke on some systems because the files
1328 are not scanned in any specific order and on different systems the order
1329 was different. A call to "sort" has been inserted into RunGrepTest for the
1330 approprate test as a short-term fix. In the longer term there may be an
1331 alternative.
1332
1333 3. I had an email from Eric Raymond about problems translating some of PCRE's
1334 man pages to HTML (despite the fact that I distribute HTML pages, some
1335 people do their own conversions for various reasons). The problems
1336 concerned the use of low-level troff macros .br and .in. I have therefore
1337 removed all such uses from the man pages (some were redundant, some could
1338 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1339 HTML has been updated to handle .nf/.fi and to complain if it encounters
1340 .br or .in.
1341
1342 4. Updated comments in configure.ac that get placed in config.h.in and also
1343 arranged for config.h to be included in the distribution, with the name
1344 config.h.generic, for the benefit of those who have to compile without
1345 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1346
1347 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1348 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1349 makevp.bat for the current PCRE, using the additional files
1350 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1351
1352 6. A Windows user reported a minor discrepancy with test 2, which turned out
1353 to be caused by a trailing space on an input line that had got lost in his
1354 copy. The trailing space was an accident, so I've just removed it.
1355
1356 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1357 that is needed.
1358
1359 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1360 as "const" (a) because they are and (b) because it helps the PHP
1361 maintainers who have recently made a script to detect big data structures
1362 in the php code that should be moved to the .rodata section. I remembered
1363 to update Builducptable as well, so it won't revert if ucptable.h is ever
1364 re-created.
1365
1366 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1367 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1368 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1369 support is not required. This saves 1.5-2K of code, which is important in
1370 some applications.
1371
1372 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1373 so as not to refer to the tables, even though these functions will never be
1374 called when UTF-8 support is disabled. Otherwise there are problems with a
1375 shared library.
1376
1377 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1378
1379 (a) It was defining its arguments as char * instead of void *.
1380
1381 (b) It was assuming that all moves were upwards in memory; this was true
1382 a long time ago when I wrote it, but is no longer the case.
1383
1384 The emulated memove() is provided for those environments that have neither
1385 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1386 is clearly not the case, as these two bugs were recently reported.
1387
1388 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1389 and Detrail to create the HTML documentation, the .txt form of the man
1390 pages, and it removes trailing spaces from listed files. It also creates
1391 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1392 case, it wraps all the #defines with #ifndefs. This script should be run
1393 before "make dist".
1394
1395 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1396 with Unicode property support.
1397
1398 (a) For a maximizing quantifier, if the two different cases of the
1399 character were of different lengths in their UTF-8 codings (there are
1400 some cases like this - I found 11), and the matching function had to
1401 back up over a mixture of the two cases, it incorrectly assumed they
1402 were both the same length.
1403
1404 (b) When PCRE was configured to use the heap rather than the stack for
1405 recursion during matching, it was not correctly preserving the data for
1406 the other case of a UTF-8 character when checking ahead for a match
1407 while processing a minimizing repeat. If the check also involved
1408 matching a wide character, but failed, corruption could cause an
1409 erroneous result when trying to check for a repeat of the original
1410 character.
1411
1412 13. Some tidying changes to the testing mechanism:
1413
1414 (a) The RunTest script now detects the internal link size and whether there
1415 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1416 values substituted by "configure". (The RunGrepTest script already did
1417 this for UTF-8.) The configure.ac script no longer substitutes the
1418 relevant variables.
1419
1420 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1421 with length and offset values. This means that the output is different
1422 for different internal link sizes. Test 2 is skipped for link sizes
1423 other than 2 because of this, bypassing the problem. Unfortunately,
1424 there was also a test in test 3 (the locale tests) that used /B and
1425 failed for link sizes other than 2. Rather than cut the whole test out,
1426 I have added a new /Z option to pcretest that replaces the length and
1427 offset values with spaces. This is now used to make test 3 independent
1428 of link size. (Test 2 will be tidied up later.)
1429
1430 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1431 segmentation fault instead of returning the appropriate error message.
1432
1433 15. In multiline mode when the newline sequence was set to "any", the pattern
1434 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1435 This doesn't seem right; it now treats the CRLF combination as the line
1436 ending, and so does not match in that case. It's only a pattern such as ^$
1437 that would hit this one: something like ^ABC$ would have failed after \r
1438 and then tried again after \r\n.
1439
1440 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1441 in an attempt to make files that differ only in their line terminators
1442 compare equal. This works on Linux.
1443
1444 17. Under certain error circumstances pcregrep might try to free random memory
1445 as it exited. This is now fixed, thanks to valgrind.
1446
1447 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1448 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1449 was because its rules for how to advance for /g after matching an empty
1450 string at the end of a line did not allow for this case. They now check for
1451 it specially.
1452
1453 20. pcretest is supposed to handle patterns and data of any length, by
1454 extending its buffers when necessary. It was getting this wrong when the
1455 buffer for a data line had to be extended.
1456
1457 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1458 CRLF as a newline sequence.
1459
1460 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1461 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1462 I have nevertheless tidied it up.
1463
1464 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1465
1466 24. Added a man page for pcre-config.
1467
1468
1469 Version 7.0 19-Dec-06
1470 ---------------------
1471
1472 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1473 moving to gcc 4.1.1.
1474
1475 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1476 sys/time.h, which is documented as needed for this function. It doesn't
1477 seem to matter on Linux, but it showed up on some releases of OS X.
1478
1479 3. It seems that there are systems where bytes whose values are greater than
1480 127 match isprint() in the "C" locale. The "C" locale should be the
1481 default when a C program starts up. In most systems, only ASCII printing
1482 characters match isprint(). This difference caused the output from pcretest
1483 to vary, making some of the tests fail. I have changed pcretest so that:
1484
1485 (a) When it is outputting text in the compiled version of a pattern, bytes
1486 other than 32-126 are always shown as hex escapes.
1487
1488 (b) When it is outputting text that is a matched part of a subject string,
1489 it does the same, unless a different locale has been set for the match
1490 (using the /L modifier). In this case, it uses isprint() to decide.
1491
1492 4. Fixed a major bug that caused incorrect computation of the amount of memory
1493 required for a compiled pattern when options that changed within the
1494 pattern affected the logic of the preliminary scan that determines the
1495 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1496 that the computed length was too small. The symptoms of this bug were
1497 either the PCRE error "internal error: code overflow" from pcre_compile(),
1498 or a glibc crash with a message such as "pcretest: free(): invalid next
1499 size (fast)". Examples of patterns that provoked this bug (shown in
1500 pcretest format) are:
1501
1502 /(?-x: )/x
1503 /(?x)(?-x: \s*#\s*)/
1504 /((?i)[\x{c0}])/8
1505 /(?i:[\x{c0}])/8
1506
1507 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1508 is now done differently.
1509
1510 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1511 wrapper classes; (b) implement a new function in the C++ scanner that is
1512 more efficient than the old way of doing things because it avoids levels of
1513 recursion in the regex matching; (c) add a paragraph to the documentation
1514 for the FullMatch() function.
1515
1516 6. The escape sequence \n was being treated as whatever was defined as
1517 "newline". Not only was this contrary to the documentation, which states
1518 that \n is character 10 (hex 0A), but it also went horribly wrong when
1519 "newline" was defined as CRLF. This has been fixed.
1520
1521 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1522 was being set to -1 for the "end of line" case (supposedly a value that no
1523 character can have). Though this value is never used (the check for end of
1524 line is "zero bytes in current character"), it caused compiler complaints.
1525 I've changed it to 0xffffffff.
1526
1527 8. In pcre_version.c, the version string was being built by a sequence of
1528 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1529 string (as it is for production releases) called a macro with an empty
1530 argument. The C standard says the result of this is undefined. The gcc
1531 compiler treats it as an empty string (which was what was wanted) but it is
1532 reported that Visual C gives an error. The source has been hacked around to
1533 avoid this problem.
1534
1535 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1536 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1537 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1538 of them did).
1539
1540 10. Originally, pcretest opened its input and output without "b"; then I was
1541 told that "b" was needed in some environments, so it was added for release
1542 5.0 to both the input and output. (It makes no difference on Unix-like
1543 systems.) Later I was told that it is wrong for the input on Windows. I've
1544 now abstracted the modes into two macros, to make it easier to fiddle with
1545 them, and removed "b" from the input mode under Windows.
1546
1547 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1548
1549 12. Added -help and --help to pcretest as an official way of being reminded
1550 of the options.
1551
1552 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1553 and pcrecpp.cc because they annoy compilers at high warning levels.
1554
1555 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1556
1557 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1558 scripts are not C programs :-) and which was not noticed because it works
1559 on Linux.
1560
1561 16. pcretest is supposed to handle any length of pattern and data line (as one
1562 line or as a continued sequence of lines) by extending its input buffer if
1563 necessary. This feature was broken for very long pattern lines, leading to
1564 a string of junk being passed to pcre_compile() if the pattern was longer
1565 than about 50K.
1566
1567 17. I have done a major re-factoring of the way pcre_compile() computes the
1568 amount of memory needed for a compiled pattern. Previously, there was code
1569 that made a preliminary scan of the pattern in order to do this. That was
1570 OK when PCRE was new, but as the facilities have expanded, it has become
1571 harder and harder to keep it in step with the real compile phase, and there
1572 have been a number of bugs (see for example, 4 above). I have now found a
1573 cunning way of running the real compile function in a "fake" mode that
1574 enables it to compute how much memory it would need, while actually only
1575 ever using a few hundred bytes of working memory and without too many
1576 tests of the mode. This should make future maintenance and development
1577 easier. A side effect of this work is that the limit of 200 on the nesting
1578 depth of parentheses has been removed (though this was never a serious
1579 limitation, I suspect). However, there is a downside: pcre_compile() now
1580 runs more slowly than before (30% or more, depending on the pattern). I
1581 hope this isn't a big issue. There is no effect on runtime performance.
1582
1583 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1584 newline (only possible for the last line of a file) and it was a
1585 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1586
1587 19. Added additional timing features to pcretest. (1) The -tm option now times
1588 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1589 separate command line item, by a number that specifies the number of
1590 repeats to use when timing. The default is 50000; this gives better
1591 precision, but takes uncomfortably long for very large patterns.
1592
1593 20. Extended pcre_study() to be more clever in cases where a branch of a
1594 subpattern has no definite first character. For example, (a*|b*)[cd] would
1595 previously give no result from pcre_study(). Now it recognizes that the
1596 first character must be a, b, c, or d.
1597
1598 21. There was an incorrect error "recursive call could loop indefinitely" if
1599 a subpattern (or the entire pattern) that was being tested for matching an
1600 empty string contained only one non-empty item after a nested subpattern.
1601 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1602 incorrectly, because the \d was being skipped in the check.
1603
1604 22. The pcretest program now has a new pattern option /B and a command line
1605 option -b, which is equivalent to adding /B to every pattern. This causes
1606 it to show the compiled bytecode, without the additional information that
1607 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1608 is the same as /B/I).
1609
1610 23. A new optimization is now able automatically to treat some sequences such
1611 as a*b as a*+b. More specifically, if something simple (such as a character
1612 or a simple class like \d) has an unlimited quantifier, and is followed by
1613 something that cannot possibly match the quantified thing, the quantifier
1614 is automatically "possessified".
1615
1616 24. A recursive reference to a subpattern whose number was greater than 39
1617 went wrong under certain circumstances in UTF-8 mode. This bug could also
1618 have affected the operation of pcre_study().
1619
1620 25. Realized that a little bit of performance could be had by replacing
1621 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1622
1623 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1624
1625 27. Possessive quantifiers such as a++ were previously implemented by turning
1626 them into atomic groups such as ($>a+). Now they have their own opcodes,
1627 which improves performance. This includes the automatically created ones
1628 from 23 above.
1629
1630 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1631 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1632 the first matched character to be a colon. This applied both to named and
1633 numbered groups.
1634
1635 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1636
1637 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1638 building PCRE on Windows easier, so I have included it in the distribution.
1639
1640 31. There is now a check in pcretest against a ridiculously large number being
1641 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1642 loop, the loop is abandoned.
1643
1644 32. Forward references to subpatterns in conditions such as (?(2)...) where
1645 subpattern 2 is defined later cause pcre_compile() to search forwards in
1646 the pattern for the relevant set of parentheses. This search went wrong
1647 when there were unescaped parentheses in a character class, parentheses
1648 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1649
1650 33. "Subroutine" calls and backreferences were previously restricted to
1651 referencing subpatterns earlier in the regex. This restriction has now
1652 been removed.
1653
1654 34. Added a number of extra features that are going to be in Perl 5.10. On the
1655 whole, these are just syntactic alternatives for features that PCRE had
1656 previously implemented using the Python syntax or my own invention. The
1657 other formats are all retained for compatibility.
1658
1659 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1660 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1661 also .NET compatible.
1662
1663 (b) A recursion or subroutine call to a named group can now be defined as
1664 (?&name) as well as (?P>name).
1665
1666 (c) A backreference to a named group can now be defined as \k<name> or
1667 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1668 5.10, are also .NET compatible.
1669
1670 (d) A conditional reference to a named group can now use the syntax
1671 (?(<name>) or (?('name') as well as (?(name).
1672
1673 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1674 groups (named and numbered) that are never evaluated inline, but can be
1675 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1676 is always false. There may be only one alternative in such a group.
1677
1678 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1679 as the simple (?(R). The condition is true only if the most recent
1680 recursion is that of the given number or name. It does not search out
1681 through the entire recursion stack.
1682
1683 (g) The escape \gN or \g{N} has been added, where N is a positive or
1684 negative number, specifying an absolute or relative reference.
1685
1686 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1687 some "unreachable code" warnings.
1688
1689 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1690 things, this adds five new scripts.
1691
1692 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1693 There were also incompatibilities regarding the handling of \Q..\E inside
1694 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1695 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1696
1697 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1698 matches an empty string, and forcibly breaks the loop. There were bugs in
1699 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1700 against aaaa the result was just "a" rather than "aaaa", for example. Two
1701 separate and independent bugs (that affected different cases) have been
1702 fixed.
1703
1704 39. Refactored the code to abolish the use of different opcodes for small
1705 capturing bracket numbers. This is a tidy that I avoided doing when I
1706 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1707 The new approach is not only tidier, it makes it possible to reduce the
1708 memory needed to fix the previous bug (38).
1709
1710 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1711 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1712 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1713 mode.
1714
1715 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1716 report.
1717
1718 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1719 copy construction and assignment in the C++ wrapper.
1720
1721 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1722 couple of bugs that could have given wrong results in the "--newline=crlf"
1723 case.
1724
1725 44. Added a number of casts and did some reorganization of signed/unsigned int
1726 variables following suggestions from Dair Grant. Also renamed the variable
1727 "this" as "item" because it is a C++ keyword.
1728
1729 45. Arranged for dftables to add
1730
1731 #include "pcre_internal.h"
1732
1733 to pcre_chartables.c because without it, gcc 4.x may remove the array
1734 definition from the final binary if PCRE is built into a static library and
1735 dead code stripping is activated.
1736
1737 46. For an unanchored pattern, if a match attempt fails at the start of a
1738 newline sequence, and the newline setting is CRLF or ANY, and the next two
1739 characters are CRLF, advance by two characters instead of one.
1740
1741
1742 Version 6.7 04-Jul-06
1743 ---------------------
1744
1745 1. In order to handle tests when input lines are enormously long, pcretest has
1746 been re-factored so that it automatically extends its buffers when
1747 necessary. The code is crude, but this _is_ just a test program. The
1748 default size has been increased from 32K to 50K.
1749
1750 2. The code in pcre_study() was using the value of the re argument before
1751 testing it for NULL. (Of course, in any sensible call of the function, it
1752 won't be NULL.)
1753
1754 3. The memmove() emulation function in pcre_internal.h, which is used on
1755 systems that lack both memmove() and bcopy() - that is, hardly ever -
1756 was missing a "static" storage class specifier.
1757
1758 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1759 containing an extended class (one that cannot be represented by a bitmap
1760 because it contains high-valued characters or Unicode property items, e.g.
1761 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1762 pattern, but PCRE should not loop if you do not (it no longer does).
1763 [Detail: two cases were found: (a) a repeated subpattern containing an
1764 extended class; (b) a recursive reference to a subpattern that followed a
1765 previous extended class. It wasn't skipping over the extended class
1766 correctly when UTF-8 mode was not set.]
1767
1768 5. A negated single-character class was not being recognized as fixed-length
1769 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1770 compile error "lookbehind assertion is not fixed length".
1771
1772 6. The RunPerlTest auxiliary script was showing an unexpected difference
1773 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1774 write a Perl script that can interpret lines of an input file either as
1775 byte characters or as UTF-8, which is what "perltest" was being required to
1776 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1777 can't do is switch easily at run time between having the "use utf8;" pragma
1778 or not. In the end, I fudged it by using the RunPerlTest script to insert
1779 "use utf8;" explicitly for the UTF-8 tests.
1780
1781 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1782 the end of the subject string, contrary to the documentation and to what
1783 Perl does. This was true of both matching functions. Now it matches only at
1784 the start of the subject and immediately after *internal* newlines.
1785
1786 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1787 a pointer to an int instead of a pointer to an unsigned long int. This
1788 caused problems on 64-bit systems.
1789
1790 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1791 instance of the 'standard' template library not being so standard".
1792
1793 10. There was no check on the number of named subpatterns nor the maximum
1794 length of a subpattern name. The product of these values is used to compute
1795 the size of the memory block for a compiled pattern. By supplying a very
1796 long subpattern name and a large number of named subpatterns, the size
1797 computation could be caused to overflow. This is now prevented by limiting
1798 the length of names to 32 characters, and the number of named subpatterns
1799 to 10,000.
1800
1801 11. Subpatterns that are repeated with specific counts have to be replicated in
1802 the compiled pattern. The size of memory for this was computed from the
1803 length of the subpattern and the repeat count. The latter is limited to
1804 65535, but there was no limit on the former, meaning that integer overflow
1805 could in principle occur. The compiled length of a repeated subpattern is
1806 now limited to 30,000 bytes in order to prevent this.
1807
1808 12. Added the optional facility to have named substrings with the same name.
1809
1810 13. Added the ability to use a named substring as a condition, using the
1811 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1812 are numbers (not recommended). Forward references are permitted.
1813
1814 14. Added forward references in named backreferences (if you see what I mean).
1815
1816 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1817 pattern could run off the end of the subject. For example, the pattern
1818 "(?s)(.{1,5})"8 did this with the subject "ab".
1819
1820 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1821 PCRE_CASELESS was set when matching characters that were quantified with ?
1822 or *.
1823
1824 17. A character class other than a single negated character that had a minimum
1825 but no maximum quantifier - for example [ab]{6,} - was not handled
1826 correctly by pce_dfa_exec(). It would match only one character.
1827
1828 18. A valid (though odd) pattern that looked like a POSIX character
1829 class but used an invalid character after [ (for example [[,abc,]]) caused
1830 pcre_compile() to give the error "Failed: internal error: code overflow" or
1831 in some cases to crash with a glibc free() error. This could even happen if
1832 the pattern terminated after [[ but there just happened to be a sequence of
1833 letters, a binary zero, and a closing ] in the memory that followed.
1834
1835 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1836 over the years. Originally (before any Unicode support), just the bottom 8
1837 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1838 output from "man perlunicode" includes this:
1839
1840 The regular expression compiler produces polymorphic opcodes. That
1841 is, the pattern adapts to the data and automatically switches to
1842 the Unicode character scheme when presented with Unicode data--or
1843 instead uses a traditional byte scheme when presented with byte
1844 data.
1845
1846 Sadly, a wide octal escape does not cause a switch, and in a string with
1847 no other multibyte characters, these octal escapes are treated as before.
1848 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1849 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1850 Unicode string.
1851
1852 I have not perpetrated such confusion in PCRE. Up till now, it took just
1853 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1854 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1855 translate to the appropriate multibyte character.
1856
1857 29. Applied some refactoring to reduce the number of warnings from Microsoft
1858 and Borland compilers. This has included removing the fudge introduced
1859 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1860 a warning about an unused variable.
1861
1862 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1863 characters since release 4.0, because Perl (from release 5.004) does not.
1864 [Or at least, is documented not to: some releases seem to be in conflict
1865 with the documentation.] However, when a pattern was studied with
1866 pcre_study() and all its branches started with \s, PCRE still included VT
1867 as a possible starting character. Of course, this did no harm; it just
1868 caused an unnecessary match attempt.
1869
1870 22. Removed a now-redundant internal flag bit that recorded the fact that case
1871 dependency changed within the pattern. This was once needed for "required
1872 byte" processing, but is no longer used. This recovers a now-scarce options
1873 bit. Also moved the least significant internal flag bit to the most-
1874 significant bit of the word, which was not previously used (hangover from
1875 the days when it was an int rather than a uint) to free up another bit for
1876 the future.
1877
1878 23. Added support for CRLF line endings as well as CR and LF. As well as the
1879 default being selectable at build time, it can now be changed at runtime
1880 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1881 specify that it is scanning data with non-default line endings.
1882
1883 24. Changed the definition of CXXLINK to make it agree with the definition of
1884 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1885
1886 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1887 recursions. This makes a big different to stack usage for some patterns.
1888
1889 26. If a subpattern containing a named recursion or subroutine reference such
1890 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1891 the space required for the compiled pattern went wrong and gave too small a
1892 value. Depending on the environment, this could lead to "Failed: internal
1893 error: code overflow at offset 49" or "glibc detected double free or
1894 corruption" errors.
1895
1896 27. Applied patches from Google (a) to support the new newline modes and (b) to
1897 advance over multibyte UTF-8 characters in GlobalReplace.
1898
1899 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1900 difference for some implementation of PCRE in some Windows version.
1901
1902 29. Added some extra testing facilities to pcretest:
1903
1904 \q<number> in a data line sets the "match limit" value
1905 \Q<number> in a data line sets the "match recursion limt" value
1906 -S <number> sets the stack size, where <number> is in megabytes
1907
1908 The -S option isn't available for Windows.
1909
1910
1911 Version 6.6 06-Feb-06
1912 ---------------------
1913
1914 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1915 in pcreposix.h. I have copied the definition from pcre.h.
1916
1917 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1918 because pcre.h is no longer a built file.
1919
1920 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1921 not normally included in the compiled code.
1922
1923
1924 Version 6.5 01-Feb-06
1925 ---------------------
1926
1927 1. When using the partial match feature with pcre_dfa_exec(), it was not
1928 anchoring the second and subsequent partial matches at the new starting
1929 point. This could lead to incorrect results. For example, with the pattern
1930 /1234/, partially matching against "123" and then "a4" gave a match.
1931
1932 2. Changes to pcregrep:
1933
1934 (a) All non-match returns from pcre_exec() were being treated as failures
1935 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1936 error message is output. Some extra information is given for the
1937 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1938 probably the only errors that are likely to be caused by users (by
1939 specifying a regex that has nested indefinite repeats, for instance).
1940 If there are more than 20 of these errors, pcregrep is abandoned.
1941
1942 (b) A binary zero was treated as data while matching, but terminated the
1943 output line if it was written out. This has been fixed: binary zeroes
1944 are now no different to any other data bytes.
1945
1946 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1947 used to set a locale for matching. The --locale=xxxx long option has
1948 been added (no short equivalent) to specify a locale explicitly on the
1949 pcregrep command, overriding the environment variables.
1950
1951 (d) When -B was used with -n, some line numbers in the output were one less
1952 than they should have been.
1953
1954 (e) Added the -o (--only-matching) option.
1955
1956 (f) If -A or -C was used with -c (count only), some lines of context were
1957 accidentally printed for the final match.
1958
1959 (g) Added the -H (--with-filename) option.
1960
1961 (h) The combination of options -rh failed to suppress file names for files
1962 that were found from directory arguments.
1963
1964 (i) Added the -D (--devices) and -d (--directories) options.
1965
1966 (j) Added the -F (--fixed-strings) option.
1967
1968 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1969
1970 (l) Added the --colo(u)r option.
1971
1972 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1973 is not present by default.
1974
1975 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1976 items such as (?R) or (?1), when the recursion could match a number of
1977 alternatives. If it matched one of the alternatives, but subsequently,
1978 outside the recursion, there was a failure, the code tried to back up into
1979 the recursion. However, because of the way PCRE is implemented, this is not
1980 possible, and the result was an incorrect result from the match.
1981
1982 In order to prevent this happening, the specification of recursion has
1983 been changed so that all such subpatterns are automatically treated as
1984 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1985
1986 4. I had overlooked the fact that, in some locales, there are characters for
1987 which isalpha() is true but neither isupper() nor islower() are true. In
1988 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1989 and ordfeminine) are like this. This affected the treatment of \w and \W
1990 when they appeared in character classes, but not when they appeared outside
1991 a character class. The bit map for "word" characters is now created
1992 separately from the results of isalnum() instead of just taking it from the
1993 upper, lower, and digit maps. (Plus the underscore character, of course.)
1994
1995 5. The above bug also affected the handling of POSIX character classes such as
1996 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1997 permanent tables. Instead, the bit maps for such a class were previously
1998 created as the appropriate unions of the upper, lower, and digit bitmaps.
1999 Now they are created by subtraction from the [[:word:]] class, which has
2000 its own bitmap.
2001
2002 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2003 It is created by subtracting the vertical space characters (\x09, \x0a,
2004 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2005 subtraction was done in the overall bitmap for a character class, meaning
2006 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2007 be recognized. This bug has been fixed.
2008
2009 7. Patches from the folks at Google:
2010
2011 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2012 real life, but is still worth protecting against".
2013
2014 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2015 regular expressions".
2016
2017 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2018 have it.
2019
2020 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2021 "configure" and the latter not, in order to fix a problem somebody had
2022 with compiling the Arg class on HP-UX.
2023
2024 (e) Improve the error-handling of the C++ wrapper a little bit.
2025
2026 (f) New tests for checking recursion limiting.
2027
2028 8. The pcre_memmove() function, which is used only if the environment does not
2029 have a standard memmove() function (and is therefore rarely compiled),
2030 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2031 returning a result (though PCRE never actually uses the result).
2032
2033 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2034 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2035 returned instead of calling malloc() with an overflowing number that would
2036 most likely cause subsequent chaos.
2037
2038 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2039
2040 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2041 with this option is matched, the nmatch and pmatch options of regexec() are
2042 ignored.
2043
2044 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2045 provided in case anyone wants to the the POSIX interface with UTF-8
2046 strings.
2047
2048 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2049 C++ linking (needed for some HP-UX environments).
2050
2051 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2052 (unused parameter) and in the pcre_printint() function (omitted "default"
2053 switch label when the default is to do nothing).
2054
2055 15. Added some code to make it possible, when PCRE is compiled as a C++
2056 library, to replace subject pointers for pcre_exec() with a smart pointer
2057 class, thus making it possible to process discontinuous strings.
2058
2059 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2060 much the same function. They were added by different people who were trying
2061 to make PCRE easy to compile on non-Unix systems. It has been suggested
2062 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2063 for compiling on Windows systems. I have therefore replaced it with
2064 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2065 defaults to "extern" for C or "extern C" for C++, which works fine on
2066 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2067 SCOPE with something explicit in config.h. In addition:
2068
2069 (a) pcreposix.h still had just "extern" instead of either of these macros;
2070 I have replaced it with PCRE_DATA_SCOPE.
2071
2072 (b) Functions such as _pcre_xclass(), which are internal to the library,
2073 but external in the C sense, all had PCRE_EXPORT in their definitions.
2074 This is apparently wrong for the Windows case, so I have removed it.
2075 (It makes no difference on Unix-like systems.)
2076
2077 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2078 of recursive calls to match(). This is different to MATCH_LIMIT because
2079 that limits the total number of calls to match(), not all of which increase
2080 the depth of recursion. Limiting the recursion depth limits the amount of
2081 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2082 when PCRE is compiled, and changed at run time. A patch from Google adds
2083 this functionality to the C++ interface.
2084
2085 18. Changes to the handling of Unicode character properties:
2086
2087 (a) Updated the table to Unicode 4.1.0.
2088
2089 (b) Recognize characters that are not in the table as "Cn" (undefined).
2090
2091 (c) I revised the way the table is implemented to a much improved format
2092 which includes recognition of ranges. It now supports the ranges that
2093 are defined in UnicodeData.txt, and it also amalgamates other
2094 characters into ranges. This has reduced the number of entries in the
2095 table from around 16,000 to around 3,000, thus reducing its size
2096 considerably. I realized I did not need to use a tree structure after
2097 all - a binary chop search is just as efficient. Having reduced the
2098 number of entries, I extended their size from 6 bytes to 8 bytes to
2099 allow for more data.
2100
2101 (d) Added support for Unicode script names via properties such as \p{Han}.
2102
2103 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2104 matching that character.
2105
2106 20. When matching a repeated Unicode property with a minimum greater than zero,
2107 (for example \pL{2,}), PCRE could look past the end of the subject if it
2108 reached it while seeking the minimum number of characters. This could
2109 happen only if some of the characters were more than one byte long, because
2110 there is a check for at least the minimum number of bytes.
2111
2112 21. Refactored the implementation of \p and \P so as to be more general, to
2113 allow for more different types of property in future. This has changed the
2114 compiled form incompatibly. Anybody with saved compiled patterns that use
2115 \p or \P will have to recompile them.
2116
2117 22. Added "Any" and "L&" to the supported property types.
2118
2119 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2120 but give a compile time error if the value is greater than 0xff.
2121
2122 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2123 accidentally not being installed or uninstalled.
2124
2125 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2126 made were to insert the current release number. This seemed silly, because
2127 it made things harder for people building PCRE on systems that don't run
2128 "configure". I have turned pcre.h into a distributed file, no longer built
2129 by "configure", with the version identification directly included. There is
2130 no longer a pcre.h.in file.
2131
2132 However, this change necessitated a change to the pcre-config script as
2133 well. It is built from pcre-config.in, and one of the substitutions was the
2134 release number. I have updated configure.ac so that ./configure now finds
2135 the release number by grepping pcre.h.
2136
2137 26. Added the ability to run the tests under valgrind.
2138
2139
2140 Version 6.4 05-Sep-05
2141 ---------------------
2142
2143 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2144 "--" to be printed when multiple files were scanned, even when none of the
2145 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2146 consider it to be a bug, and have restored the previous behaviour.
2147
2148 2. A couple of code tidies to get rid of compiler warnings.
2149
2150 3. The pcretest program used to cheat by referring to symbols in the library
2151 whose names begin with _pcre_. These are internal symbols that are not
2152 really supposed to be visible externally, and in some environments it is
2153 possible to suppress them. The cheating is now confined to including
2154 certain files from the library's source, which is a bit cleaner.
2155
2156 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2157 file's purpose clearer.
2158
2159 5. Reorganized pcre_ucp_findchar().
2160
2161
2162 Version 6.3 15-Aug-05
2163 ---------------------
2164
2165 1. The file libpcre.pc.in did not have general read permission in the tarball.
2166
2167 2. There were some problems when building without C++ support:
2168
2169 (a) If C++ support was not built, "make install" and "make test" still
2170 tried to test it.
2171
2172 (b) There were problems when the value of CXX was explicitly set. Some
2173 changes have been made to try to fix these, and ...
2174
2175 (c) --disable-cpp can now be used to explicitly disable C++ support.
2176
2177 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2178 backslash in a target when C++ was disabled. This confuses some
2179 versions of "make", apparently. Using an intermediate variable solves
2180 this. (Same for CPP_LOBJ.)
2181
2182 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2183 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2184 necessary on certain architectures.
2185
2186 4. Added a setting of -export-symbols-regex to the link command to remove
2187 those symbols that are exported in the C sense, but actually are local
2188 within the library, and not documented. Their names all begin with
2189 "_pcre_". This is not a perfect job, because (a) we have to except some
2190 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2191 available (and never for static libraries). I have made a note to try to
2192 find a way round (a) in the future.
2193
2194
2195 Version 6.2 01-Aug-05
2196 ---------------------
2197
2198 1. There was no test for integer overflow of quantifier values. A construction
2199 such as {1111111111111111} would give undefined results. What is worse, if
2200 a minimum quantifier for a parenthesized subpattern overflowed and became
2201 negative, the calculation of the memory size went wrong. This could have
2202 led to memory overwriting.
2203
2204 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2205
2206 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2207 operating environments where this matters.
2208
2209 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2210 PCRE options from within the C++ wrapper.
2211
2212 5. Named capturing subpatterns were not being correctly counted when a pattern
2213 was compiled. This caused two problems: (a) If there were more than 100
2214 such subpatterns, the calculation of the memory needed for the whole
2215 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2216 back references of the form \12, where the number was greater than 9, were
2217 not recognized as back references, even though there were sufficient
2218 previous subpatterns.
2219
2220 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2221 versions of gcc, e.g. 2.95.4.
2222
2223
2224 Version 6.1 21-Jun-05
2225 ---------------------
2226
2227 1. There was one reference to the variable "posix" in pcretest.c that was not
2228 surrounded by "#if !defined NOPOSIX".
2229
2230 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2231 the cross-check on the old pcre_info() function, for the benefit of the
2232 cut-down version of PCRE that is currently imported into Exim.
2233
2234 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2235 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2236 patterns that start (?i) but I don't think that matters. The use of (?i) is
2237 just an example; this all applies to the other options as well.
2238
2239 4. Since libtool seems to echo the compile commands it is issuing, the output
2240 from "make" can be reduced a bit by putting "@" in front of each libtool
2241 compile command.
2242
2243 5. Patch from the folks at Google for configure.in to be a bit more thorough
2244 in checking for a suitable C++ installation before trying to compile the
2245 C++ stuff. This should fix a reported problem when a compiler was present,
2246 but no suitable headers.
2247
2248 6. The man pages all had just "PCRE" as their title. I have changed them to
2249 be the relevant file name. I have also arranged that these names are
2250 retained in the file doc/pcre.txt, which is a concatenation in text format
2251 of all the man pages except the little individual ones for each function.
2252
2253 7. The NON-UNIX-USE file had not been updated for the different set of source
2254 files that come with release 6. I also added a few comments about the C++
2255 wrapper.
2256
2257
2258 Version 6.0 07-Jun-05
2259 ---------------------
2260
2261 1. Some minor internal re-organization to help with my DFA experiments.
2262
2263 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2264 didn't matter for the library itself when fully configured, but did matter
2265 when compiling without UCP support, or within Exim, where the ucp files are
2266 not imported.
2267
2268 3. Refactoring of the library code to split up the various functions into
2269 different source modules. The addition of the new DFA matching code (see
2270 below) to a single monolithic source would have made it really too
2271 unwieldy, quite apart from causing all the code to be include in a
2272 statically linked application, when only some functions are used. This is
2273 relevant even without the DFA addition now that patterns can be compiled in
2274 one application and matched in another.
2275
2276 The downside of splitting up is that there have to be some external
2277 functions and data tables that are used internally in different modules of
2278 the library but which are not part of the API. These have all had their
2279 names changed to start with "_pcre_" so that they are unlikely to clash
2280 with other external names.
2281
2282 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2283 a different (DFA) algorithm. Although it is slower than the original
2284 function, it does have some advantages for certain types of matching
2285 problem.
2286
2287 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2288 including restarting after a partial match.
2289
2290 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2291 defined when compiling for Windows was sent to me. I have put it into the
2292 code, though I have no means of testing or verifying it.
2293
2294 7. Added the pcre_refcount() auxiliary function.
2295
2296 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2297 match before or at the first newline in the subject string. In pcretest,
2298 the /f option on a pattern can be used to set this.
2299
2300 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2301 would behave wrongly. This has been present in PCRE since release 4.0.
2302
2303 10. A number of changes to the pcregrep command:
2304
2305 (a) Refactored how -x works; insert ^(...)$ instead of setting
2306 PCRE_ANCHORED and checking the length, in preparation for adding
2307 something similar for -w.
2308
2309 (b) Added the -w (match as a word) option.
2310
2311 (c) Refactored the way lines are read and buffered so as to have more
2312 than one at a time available.
2313
2314 (d) Implemented a pcregrep test script.
2315
2316 (e) Added the -M (multiline match) option. This allows patterns to match
2317 over several lines of the subject. The buffering ensures that at least
2318 8K, or the rest of the document (whichever is the shorter) is available
2319 for matching (and similarly the previous 8K for lookbehind assertions).
2320
2321 (f) Changed the --help output so that it now says
2322
2323 -w, --word-regex(p)
2324
2325 instead of two lines, one with "regex" and the other with "regexp"
2326 because that confused at least one person since the short forms are the
2327 same. (This required a bit of code, as the output is generated
2328 automatically from a table. It wasn't just a text change.)
2329
2330 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2331 option but starts with a hyphen. Could be a pattern or a path name
2332 starting with a hyphen, for instance.
2333
2334 (h) "-" can be given as a file name to represent stdin.
2335
2336 (i) When file names are being printed, "(standard input)" is used for
2337 the standard input, for compatibility with GNU grep. Previously
2338 "<stdin>" was used.
2339
2340 (j) The option --label=xxx can be used to supply a name to be used for
2341 stdin when file names are being printed. There is no short form.
2342
2343 (k) Re-factored the options decoding logic because we are going to add
2344 two more options that take data. Such options can now be given in four
2345 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2346
2347 (l) Added the -A, -B, and -C options for requesting that lines of context
2348 around matches be printed.
2349
2350 (m) Added the -L option to print the names of files that do not contain
2351 any matching lines, that is, the complement of -l.
2352
2353 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2354 continue to scan other files.
2355
2356 (o) The -s option was incorrectly implemented. For compatibility with other
2357 greps, it now suppresses the error message for a non-existent or non-
2358 accessible file (but not the return code). There is a new option called
2359 -q that suppresses the output of matching lines, which was what -s was
2360 previously doing.
2361
2362 (p) Added --include and --exclude options to specify files for inclusion
2363 and exclusion when recursing.
2364
2365 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2366 Hopefully, it now does.
2367
2368 12. Missing cast in pcre_study().
2369
2370 13. Added an "uninstall" target to the makefile.
2371
2372 14. Replaced "extern" in the function prototypes in Makefile.in with
2373 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2374 world, but is set differently for Windows.
2375
2376 15. Added a second compiling function called pcre_compile2(). The only
2377 difference is that it has an extra argument, which is a pointer to an
2378 integer error code. When there is a compile-time failure, this is set
2379 non-zero, in addition to the error test pointer being set to point to an
2380 error message. The new argument may be NULL if no error number is required
2381 (but then you may as well call pcre_compile(), which is now just a
2382 wrapper). This facility is provided because some applications need a
2383 numeric error indication, but it has also enabled me to tidy up the way
2384 compile-time errors are handled in the POSIX wrapper.
2385
2386 16. Added VPATH=.libs to the makefile; this should help when building with one
2387 prefix path and installing with another. (Or so I'm told by someone who
2388 knows more about this stuff than I do.)
2389
2390 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2391 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2392 match everything, including newlines. This is not POSIX-compatible, but
2393 somebody wanted the feature. From pcretest it can be activated by using
2394 both the P and the s flags.
2395
2396 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2397
2398 19. libpcre.pc was being incorrectly installed as executable.
2399
2400 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2401 it now also looks for '\r' so that it will work unmodified on Windows.
2402
2403 21. Added Google's contributed C++ wrapper to the distribution.
2404
2405 22. Added some untidy missing memory free() calls in pcretest, to keep
2406 Electric Fence happy when testing.
2407
2408
2409
2410 Version 5.0 13-Sep-04
2411 ---------------------
2412
2413 1. Internal change: literal characters are no longer packed up into items
2414 containing multiple characters in a single byte-string. Each character
2415 is now matched using a separate opcode. However, there may be more than one
2416 byte in the character in UTF-8 mode.
2417
2418 2. The pcre_callout_block structure has two new fields: pattern_position and
2419 next_item_length. These contain the offset in the pattern to the next match
2420 item, and its length, respectively.
2421
2422 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2423 insertion of callouts before each pattern item. Added the /C option to
2424 pcretest to make use of this.
2425
2426 4. On the advice of a Windows user, the lines
2427
2428 #if defined(_WIN32) || defined(WIN32)
2429 _setmode( _fileno( stdout ), 0x8000 );
2430 #endif /* defined(_WIN32) || defined(WIN32) */
2431
2432 have been added to the source of pcretest. This apparently does useful
2433 magic in relation to line terminators.
2434
2435 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2436 for the benefit of those environments where the "b" makes a difference.
2437
2438 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2439 to know about it. I have put a hack into configure.in that adds in code
2440 to set GCC=yes if CC=icc. This seems to end up at a point in the
2441 generated configure script that is early enough to affect the setting of
2442 compiler options, which is what is needed, but I have no means of testing
2443 whether it really works. (The user who reported this had patched the
2444 generated configure script, which of course I cannot do.)
2445
2446 LATER: After change 22 below (new libtool files), the configure script
2447 seems to know about icc (and also ecc). Therefore, I have commented out
2448 this hack in configure.in.
2449
2450 7. Added support for pkg-config (2 patches were sent in).
2451
2452 8. Negated POSIX character classes that used a combination of internal tables
2453 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2454 [[:^ascii]]. Typically, they would match almost any characters. The other
2455 POSIX classes were not broken in this way.
2456
2457 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2458 to find the match, as PCRE was deluded into thinking that the match had to
2459 start at the start point or following a newline. The same bug applied to
2460 patterns with negative forward assertions or any backward assertions
2461 preceding ".*" at the start, unless the pattern required a fixed first
2462 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2463
2464 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2465 starting at the last subject character, bytes beyond the end of the subject
2466 string were read.
2467
2468 11. Renamed the variable "class" as "classbits" to make life easier for C++
2469 users. (Previously there was a macro definition, but it apparently wasn't
2470 enough.)
2471
2472 12. Added the new field "tables" to the extra data so that tables can be passed
2473 in at exec time, or the internal tables can be re-selected. This allows
2474 a compiled regex to be saved and re-used at a later time by a different
2475 program that might have everything at different addresses.
2476
2477 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2478 -R library as well as a -L library.
2479
2480 14. The debugging options of pcretest (-d on the command line or D on a
2481 pattern) showed incorrect output for anything following an extended class
2482 that contained multibyte characters and which was followed by a quantifier.
2483
2484 15. Added optional support for general category Unicode character properties
2485 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2486 support. It adds about 90K to the size of the library. The meanings of the
2487 inbuilt class escapes such as \d and \s have NOT been changed.
2488
2489 16. Updated pcredemo.c to include calls to free() to release the memory for the
2490 compiled pattern.
2491
2492 17. The generated file chartables.c was being created in the source directory
2493 instead of in the building directory. This caused the build to fail if the
2494 source directory was different from the building directory, and was
2495 read-only.
2496
2497 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2498 file. No doubt somebody will tell me if they don't make sense... Also added
2499 Dan Mooney's comments about building on OpenVMS.
2500
2501 19. Added support for partial matching via the PCRE_PARTIAL option for
2502 pcre_exec() and the \P data escape in pcretest.
2503
2504 20. Extended pcretest with 3 new pattern features:
2505
2506 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2507 write the compiled pattern to the file whose name is "rest-of-line".
2508 This is a straight binary dump of the data, with the saved pointer to
2509 the character tables forced to be NULL. The study data, if any, is
2510 written too. After writing, pcretest reads a new pattern.
2511
2512 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2513 compiled pattern from the given file. There must not be any
2514 occurrences of "<" in the file name (pretty unlikely); if there are,
2515 pcretest will instead treat the initial "<" as a pattern delimiter.
2516 After reading in the pattern, pcretest goes on to read data lines as
2517 usual.
2518
2519 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2520 and 16-bit fields in a compiled pattern, to simulate a pattern that
2521 was compiled on a host of opposite endianness.
2522
2523 21. The pcre-exec() function can now cope with patterns that were compiled on
2524 hosts of opposite endianness, with this restriction:
2525
2526 As for any compiled expression that is saved and used later, the tables
2527 pointer field cannot be preserved; the extra_data field in the arguments
2528 to pcre_exec() should be used to pass in a tables address if a value
2529 other than the default internal tables were used at compile time.
2530
2531 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2532 now diagnosed as an error. Previously, most of the time, a negative number
2533 would have been treated as zero, but if in addition "ovector" was passed as
2534 NULL, a crash could occur.
2535
2536 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2537 new versions from the libtool 1.5 distribution (the last one is a copy of
2538 a file called libtool.m4). This seems to have fixed the need to patch
2539 "configure" to support Darwin 1.3 (which I used to do). However, I still
2540 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2541 workstation).
2542
2543 24. Changed the PCRE licence to be the more standard "BSD" licence.
2544
2545
2546 Version 4.5 01-Dec-03
2547 ---------------------
2548
2549 1. There has been some re-arrangement of the code for the match() function so
2550 that it can be compiled in a version that does not call itself recursively.
2551 Instead, it keeps those local variables that need separate instances for
2552 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2553 needs to "recurse". Keeping track of where control must go is done by means
2554 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2555 hide most of the details from the main code, and operates only if
2556 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2557 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2558 operating.
2559
2560 To make it easier for callers to provide specially tailored get/free
2561 functions for this usage, two new functions, pcre_stack_malloc, and
2562 pcre_stack_free, are used. They are always called in strict stacking order,
2563 and the size of block requested is always the same.
2564
2565 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2566 PCRE has been compiled to use the stack or the heap for recursion. The
2567 -C option of pcretest uses this to show which version is compiled.
2568
2569 A new data escape \S, is added to pcretest; it causes the amounts of store
2570 obtained and freed by both kinds of malloc/free at match time to be added
2571 to the output.
2572
2573 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2574 what's available on my current Linux desktop machine.
2575
2576 3. When matching a UTF-8 string, the test for a valid string at the start has
2577 been extended. If start_offset is not zero, PCRE now checks that it points
2578 to a byte that is the start of a UTF-8 character. If not, it returns
2579 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2580 this is necessary because there may be backward assertions in the pattern.
2581 When matching the same subject several times, it may save resources to use
2582 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2583
2584 4. The code for checking the validity of UTF-8 strings has been tightened so
2585 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2586 containing "overlong sequences".
2587
2588 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2589 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2590 should have been "&", but it just so happened that all the cases this let
2591 through by mistake were picked up later in the function.
2592
2593 6. I had used a variable called "isblank" - this is a C99 function, causing
2594 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2595
2596 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2597 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2598 so the version is shown; (c) stop "make test" echoing "./RunTest".
2599
2600 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2601
2602 9. The prototype for memmove() for systems that don't have it was using
2603 size_t, but the inclusion of the header that defines size_t was later. I've
2604 moved the #includes for the C headers earlier to avoid this.
2605
2606 10. Added some adjustments to the code to make it easier to compiler on certain
2607 special systems:
2608
2609 (a) Some "const" qualifiers were missing.
2610 (b) Added the macro EXPORT before all exported functions; by default this
2611 is defined to be empty.
2612 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2613 that it reads its output file name as an argument instead of writing
2614 to the standard output and assuming this can be redirected.
2615
2616 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2617 class containing characters with values greater than 255, PCRE compilation
2618 went into a loop.
2619
2620 12. A recursive reference to a subpattern that was within another subpattern
2621 that had a minimum quantifier of zero caused PCRE to crash. For example,
2622 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2623 recursion. If the recursively-called subpattern itself had a zero repeat,
2624 that was OK.
2625
2626 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2627 buffer into which it was copied (for escape processing) was still set at
2628 1024, so long lines caused crashes.
2629
2630 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2631 "internal error: code overflow...". This applied to any character class
2632 that was followed by a possessive quantifier.
2633
2634 15. Modified the Makefile to add libpcre.la as a prerequisite for
2635 libpcreposix.la because I was told this is needed for a parallel build to
2636 work.
2637
2638 16. If a pattern that contained .* following optional items at the start was
2639 studied, the wrong optimizing data was generated, leading to matching
2640 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2641 matching string must start with a or b or c. The correct conclusion for
2642 this pattern is that a match can start with any character.
2643
2644
2645 Version 4.4 13-Aug-03
2646 ---------------------
2647
2648 1. In UTF-8 mode, a character class containing characters with values between
2649 127 and 255 was not handled correctly if the compiled pattern was studied.
2650 In fixing this, I have also improved the studying algorithm for such
2651 classes (slightly).
2652
2653 2. Three internal functions had redundant arguments passed to them. Removal
2654 might give a very teeny performance improvement.
2655
2656 3. Documentation bug: the value of the capture_top field in a callout is *one
2657 more than* the number of the hightest numbered captured substring.
2658
2659 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2660 in incorrectly linking with a previously installed version. They now link
2661 explicitly with libpcre.la.
2662
2663 5. configure.in no longer needs to recognize Cygwin specially.
2664
2665 6. A problem in pcre.in for Windows platforms is fixed.
2666
2667 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2668 pcretest, it used to include the size of the study block as part of its
2669 output. Unfortunately, the structure contains a field that has a different
2670 size on different hardware architectures. This meant that the tests that
2671 showed this size failed. As the block is currently always of a fixed size,
2672 this information isn't actually particularly useful in pcretest output, so
2673 I have just removed it.
2674
2675 8. Three pre-processor statements accidentally did not start in column 1.
2676 Sadly, there are *still* compilers around that complain, even though
2677 standard C has not required this for well over a decade. Sigh.
2678
2679 9. In pcretest, the code for checking callouts passed small integers in the
2680 callout_data field, which is a void * field. However, some picky compilers
2681 complained about the casts involved for this on 64-bit systems. Now
2682 pcretest passes the address of the small integer instead, which should get
2683 rid of the warnings.
2684
2685 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2686 both compile and run time, and gives an error if an invalid UTF-8 sequence
2687 is found. There is a option for disabling this check in cases where the
2688 string is known to be correct and/or the maximum performance is wanted.
2689
2690 11. In response to a bug report, I changed one line in Makefile.in from
2691
2692 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2693 to
2694 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2695
2696 to look similar to other lines, but I have no way of telling whether this
2697 is the right thing to do, as I do not use Windows. No doubt I'll get told
2698 if it's wrong...
2699
2700
2701 Version 4.3 21-May-03
2702 ---------------------
2703
2704 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2705 Makefile.
2706
2707 2. Some refactoring to improve the quality of the code:
2708
2709 (i) The utf8_table... variables are now declared "const".
2710
2711 (ii) The code for \cx, which used the "case flipping" table to upper case
2712 lower case letters, now just substracts 32. This is ASCII-specific,
2713 but the whole concept of \cx is ASCII-specific, so it seems
2714 reasonable.
2715
2716 (iii) PCRE was using its character types table to recognize decimal and
2717 hexadecimal digits in the pattern. This is silly, because it handles
2718 only 0-9, a-f, and A-F, but the character types table is locale-
2719 specific, which means strange things might happen. A private
2720 table is now used for this - though it costs 256 bytes, a table is
2721 much faster than multiple explicit tests. Of course, the standard
2722 character types table is still used for matching digits in subject
2723 strings against \d.
2724
2725 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2726 ending in _t are). So I've renamed it as ESC_tee.
2727
2728 3. The first argument for regexec() in the POSIX wrapper should have been
2729 defined as "const".
2730
2731 4. Changed pcretest to use malloc() for its buffers so that they can be
2732 Electric Fenced for debugging.
2733
2734 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2735 to read one or more bytes before the start of the subject string. Often this
2736 had no effect on PCRE's behaviour, but in some circumstances it could
2737 provoke a segmentation fault.
2738
2739 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2740 to try to read one or more bytes before the start of the subject string.
2741
2742 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2743 UTF-8 support could misbehave in various ways if the subject string
2744 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2745 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2746 back over UTF-8 characters.)
2747
2748
2749 Version 4.2 14-Apr-03
2750 ---------------------
2751
2752 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2753
2754 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2755 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2756 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2757 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2758 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2759 and BUILD_EXEEXT
2760 Note: automatic setting of the BUILD variables is not yet working
2761 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2762 compile-time but not at link-time
2763 [LINK]: use for linking executables only
2764 make different versions for Windows and non-Windows
2765 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2766 libraries
2767 [LINK_FOR_BUILD]: new variable
2768 [OBJEXT]: use throughout
2769 [EXEEXT]: use throughout
2770 <winshared>: new target
2771 <wininstall>: new target
2772 <dftables.o>: use native compiler
2773 <dftables>: use native linker
2774 <install>: handle Windows platform correctly
2775 <clean>: ditto
2776 <check>: ditto
2777 copy DLL to top builddir before testing
2778
2779 As part of these changes, -no-undefined was removed again. This was reported
2780 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2781 in any case.
2782
2783 3. Some tidies to get rid of compiler warnings:
2784
2785 . In the match_data structure, match_limit was an unsigned long int, whereas
2786 match_call_count was an int. I've made them both unsigned long ints.
2787
2788 . In pcretest the fact that a const uschar * doesn't automatically cast to
2789 a void * provoked a warning.
2790
2791 . Turning on some more compiler warnings threw up some "shadow" variables
2792 and a few more missing casts.
2793
2794 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2795 option, a class that contained a single character with a value between 128
2796 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2797
2798 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2799 option, a class that contained several characters, but with at least one
2800 whose value was between 128 and 255 caused PCRE to crash.
2801
2802
2803 Version 4.1 12-Mar-03
2804 ---------------------
2805
2806 1. Compiling with gcc -pedantic found a couple of places where casts were
2807 needed, and a string in dftables.c that was longer than standard compilers are
2808 required to support.
2809
2810 2. Compiling with Sun's compiler found a few more places where the code could
2811 be tidied up in order to avoid warnings.
2812
2813 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2814 first of these names is deprecated in the latest Autoconf in favour of the name
2815 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2816 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2817 analogy I have changed it to CFLAGS_FOR_BUILD.
2818
2819 4. Added -no-undefined to the linking command in the Makefile, because this is
2820 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2821 linking step for the pcreposix library.
2822
2823 5. PCRE was failing to diagnose the case of two named groups with the same
2824 name.
2825
2826 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2827 literal character that is needed in the subject for a match, and scans along to
2828 ensure that it is present before embarking on the full matching process. This
2829 saves time in cases of nested unlimited repeats that are never going to match.
2830 Problem: the scan can take a lot of time if the subject is very long (e.g.
2831 megabytes), thus penalizing straightforward matches. It is now done only if the
2832 amount of subject to be scanned is less than 1000 bytes.
2833
2834 7. A lesser problem with the same optimization is that it was recording the
2835 first character of an anchored pattern as "needed", thus provoking a search
2836 right along the subject, even when the first match of the pattern was going to
2837 fail. The "needed" character is now not set for anchored patterns, unless it
2838 follows something in the pattern that is of non-fixed length. Thus, it still
2839 fulfils its original purpose of finding quick non-matches in cases of nested
2840 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2841
2842
2843 Version 4.0 17-Feb-03
2844 ---------------------
2845
2846 1. If a comment in an extended regex that started immediately after a meta-item
2847 extended to the end of string, PCRE compiled incorrect data. This could lead to
2848 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2849
2850 2. Moved to autoconf 2.53 and libtool 1.4.2.
2851
2852 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2853 the special perltest8 script is no longer needed - all the tests can be run
2854 from a single perltest script.
2855
2856 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2857 by \s. It has now been removed in PCRE. This means it isn't recognized as
2858 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2859 class [:space:] *does* include VT, thereby creating a mess.
2860
2861 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2862 space and tab.
2863
2864 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2865 its new features into the main test script, reducing the number of scripts.
2866
2867 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2868 were backward compatible, and made the (?i) apply to the whole pattern, as if
2869 /i were given. Now it behaves more logically, and applies the option setting
2870 only to what follows. PCRE has been changed to follow suit. However, if it
2871 finds options settings right at the start of the pattern, it extracts them into
2872 the global options, as before. Thus, they show up in the info data.
2873
2874 8. Added support for the \Q...\E escape sequence. Characters in between are
2875 treated as literals. This is slightly different from Perl in that $ and @ are
2876 also handled as literals inside the quotes. In Perl, they will cause variable
2877 interpolation. Note the following examples:
2878
2879 Pattern PCRE matches Perl matches
2880
2881 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2882 \Qabc\$xyz\E abc\$xyz abc\$xyz
2883 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2884
2885 For compatibility with Perl, \Q...\E sequences are recognized inside character
2886 classes as well as outside them.
2887
2888 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2889 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2890 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2891 signed/unsigned warnings.
2892
2893 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2894 option for pcretest, so I've replaced it by a simple function that does just
2895 that job.
2896
2897 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2898 "pcregrep -".
2899
2900 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2901 Java package. This provides some syntactic sugar for simple cases of what my
2902 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2903 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2904 item, you can use this simplified notation. Note that only makes sense with
2905 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2906 greediness, whatever the setting of the PCRE_UNGREEDY option.
2907
2908 13. A change of greediness default within a pattern was not taking effect at
2909 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2910 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2911 was abstracted outside.
2912
2913 14. PCRE now supports the \G assertion. It is true when the current matching
2914 position is at the start point of the match. This differs from \A when the
2915 starting offset is non-zero. Used with the /g option of pcretest (or similar
2916 code), it works in the same way as it does for Perl's /g option. If all
2917 alternatives of a regex begin with \G, the expression is anchored to the start
2918 match position, and the "anchored" flag is set in the compiled expression.
2919
2920 15. Some bugs concerning the handling of certain option changes within patterns
2921 have been fixed. These applied to options other than (?ims). For example,
2922 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2923 been the other way round. Some of this was related to change 7 above.
2924
2925 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2926 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2927 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2928 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2929
2930 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2931 mode. Unlike ".", it always matches newline, whatever the setting of
2932 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2933 assertions. Perl allows it, but it doesn't (in general) work because it can't
2934 calculate the length of the lookbehind. At least, that's the case for Perl
2935 5.8.0 - I've been told they are going to document that it doesn't work in
2936 future.
2937
2938 18. Added an error diagnosis for escapes that PCRE does not support: these are
2939 \L, \l, \N, \P, \p, \U, \u, and \X.
2940
2941 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2942 reading past the end of the pattern in cases such as /[abcd/.
2943
2944 20. PCRE was getting more memory than necessary for patterns with classes that
2945 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2946
2947 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2948 compiling PCRE for use with Virtual Pascal.
2949
2950 22. Small fix to the Makefile to make it work properly if the build is done
2951 outside the source tree.
2952
2953 23. Added a new extension: a condition to go with recursion. If a conditional
2954 subpattern starts with (?(R) the "true" branch is used if recursion has
2955 happened, whereas the "false" branch is used only at the top level.
2956
2957 24. When there was a very long string of literal characters (over 255 bytes
2958 without UTF support, over 250 bytes with UTF support), the computation of how
2959 much memory was required could be incorrect, leading to segfaults or other
2960 strange effects.
2961
2962 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2963 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2964 there was a subsequent back reference to those brackets. This meant that, for
2965 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2966 possible to check for precisely this case. All we can do is abandon the
2967 optimization if .* occurs inside capturing brackets when there are any back
2968 references whatsoever. (See below for a better fix that came later.)
2969
2970 26. The handling of the optimization for finding the first character of a
2971 non-anchored pattern, and for finding a character that is required later in the
2972 match were failing in some cases. This didn't break the matching; it just
2973 failed to optimize when it could. The way this is done has been re-implemented.
2974
2975 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2976
2977 28. Added a new feature that provides some of the functionality that Perl
2978 provides with (?{...}). The facility is termed a "callout". The way it is done
2979 in PCRE is for the caller to provide an optional function, by setting
2980 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2981 global variable. By default it is unset, which disables all calling out. To get
2982 the function called, the regex must include (?C) at appropriate points. This
2983 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2984 This provides a means of identifying different callout points. When PCRE
2985 reaches such a point in the regex, if pcre_callout has been set, the external
2986 function is called. It is provided with data in a structure called
2987 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2988 matching continues; if it returns a non-zero value, the match at the current
2989 point fails. However, backtracking will occur if possible. [This was changed
2990 later and other features added - see item 49 below.]
2991
2992 29. pcretest is upgraded to test the callout functionality. It provides a
2993 callout function that displays information. By default, it shows the start of
2994 the match and the current position in the text. There are some new data escapes
2995 to vary what happens:
2996
2997 \C+ in addition, show current contents of captured substrings
2998 \C- do not supply a callout function
2999 \C!n return 1 when callout number n is reached
3000 \C!n!m return 1 when callout number n is reached for the mth time
3001
3002 30. If pcregrep was called with the -l option and just a single file name, it
3003 output "<stdin>" if a match was found, instead of the file name.
3004
3005 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3006 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3007 pcre_exec(). This saves a malloc/free per call. The default value of
3008 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3009 when configuring.
3010
3011 32. The default maximum size of a compiled pattern is 64K. There have been a
3012 few cases of people hitting this limit. The code now uses macros to handle the
3013 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3014 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3015 configuring. Tests 2 and 5 work only with 2-byte links because they output
3016 debugging information about compiled patterns.
3017
3018 33. Internal code re-arrangements:
3019
3020 (a) Moved the debugging function for printing out a compiled regex into
3021 its own source file (printint.c) and used #include to pull it into
3022 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3023 separate copies.
3024
3025 (b) Defined the list of op-code names for debugging as a macro in
3026 internal.h so that it is next to the definition of the opcodes.
3027
3028 (c) Defined a table of op-code lengths for simpler skipping along compiled
3029 code. This is again a macro in internal.h so that it is next to the
3030 definition of the opcodes.
3031
3032 34. Added support for recursive calls to individual subpatterns, along the
3033 lines of Robin Houston's patch (but implemented somewhat differently).
3034
3035 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3036 allow it to read and process whole directories in Win32. This code was
3037 contributed by Lionel Fourquaux; it has not been tested by me.
3038
3039 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3040 used to name a group. Names consist of alphanumerics and underscores, and must
3041 be unique. Back references use the syntax (?P=name) and recursive calls use
3042 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3043 numbers. The function pcre_fullinfo() can be used after compilation to extract
3044 a name/number map. There are three relevant calls:
3045
3046 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3047 PCRE_INFO_NAMECOUNT yields the number of entries
3048 PCRE_INFO_NAMETABLE yields a pointer to the map.
3049
3050 The map is a vector of fixed-size entries. The size of each entry depends on
3051 the length of the longest name used. The first two bytes of each entry are the
3052 group number, most significant byte first. There follows the corresponding
3053 name, zero terminated. The names are in alphabetical order.
3054
3055 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3056 case instead of 255. Making it the same both with and without UTF-8 support
3057 means that the same test output works with both.
3058
3059 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3060 calling malloc() with a zero argument.
3061
3062 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3063 optimization. I've improved things by keeping a bitmap of backreferences with
3064 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3065 fact referenced, the optimization can be applied. It is unlikely that a
3066 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3067 the match to follow \n) will appear inside brackets with a number greater than
3068 31, but if it does, any back reference > 31 suppresses the optimization.
3069
3070 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3071 of disabling numbered capturing parentheses. Any opening parenthesis that is
3072 not followed by ? behaves as if it were followed by ?: but named parentheses
3073 can still be used for capturing (and they will acquire numbers in the usual
3074 way).
3075
3076 41. Redesigned the return codes from the match() function into yes/no/error so
3077 that errors can be passed back from deep inside the nested calls. A malloc
3078 failure while inside a recursive subpattern call now causes the
3079 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3080
3081 42. It is now possible to set a limit on the number of times the match()
3082 function is called in a call to pcre_exec(). This facility makes it possible to
3083 limit the amount of recursion and backtracking, though not in a directly
3084 obvious way, because the match() function is used in a number of different
3085 circumstances. The count starts from zero for each position in the subject
3086 string (for non-anchored patterns). The default limit is, for compatibility, a
3087 large number, namely 10 000 000. You can change this in two ways:
3088
3089 (a) When configuring PCRE before making, you can use --with-match-limit=n
3090 to set a default value for the compiled library.
3091
3092 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3093 a different value is set. See 45 below.
3094
3095 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3096
3097 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3098 of things that can be changed at compile time. The first argument specifies
3099 what is wanted and the second points to where the information is to be placed.
3100 The current list of available information is:
3101
3102 PCRE_CONFIG_UTF8
3103
3104 The output is an integer that is set to one if UTF-8 support is available;
3105 otherwise it is set to zero.
3106
3107 PCRE_CONFIG_NEWLINE
3108
3109 The output is an integer that it set to the value of the code that is used for
3110 newline. It is either LF (10) or CR (13).
3111
3112 PCRE_CONFIG_LINK_SIZE
3113
3114 The output is an integer that contains the number of bytes used for internal
3115 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3116
3117 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3118
3119 The output is an integer that contains the threshold above which the POSIX
3120 interface uses malloc() for output vectors. See item 31 above.
3121
3122 PCRE_CONFIG_MATCH_LIMIT
3123
3124 The output is an unsigned integer that contains the default limit of the number
3125 of match() calls in a pcre_exec() execution. See 42 above.
3126
3127 44. pcretest has been upgraded by the addition of the -C option. This causes it
3128 to extract all the available output from the new pcre_config() function, and to
3129 output it. The program then exits immediately.
3130
3131 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3132 order to support additional features. One way would have been to define
3133 pcre_exec2() (for example) with extra arguments, but this would not have been
3134 extensible, and would also have required all calls to the original function to
3135 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3136 is used for passing in "extra" data from pcre_study().
3137
3138 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3139 contains the following fields:
3140
3141 flags a bitmap indicating which of the following fields are set
3142 study_data opaque data from pcre_study()
3143 match_limit a way of specifying a limit on match() calls for a specific
3144 call to pcre_exec()
3145 callout_data data for callouts (see 49 below)
3146
3147 The flag bits are also defined in pcre.h, and are
3148
3149 PCRE_EXTRA_STUDY_DATA
3150 PCRE_EXTRA_MATCH_LIMIT
3151 PCRE_EXTRA_CALLOUT_DATA
3152
3153 The pcre_study() function now returns one of these new pcre_extra blocks, with
3154 the actual study data pointed to by the study_data field, and the
3155 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3156 before. That is, this change is entirely upwards-compatible and requires no
3157 change to existing code.
3158
3159 If you want to pass in additional data to pcre_exec(), you can either place it
3160 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3161 block.
3162
3163 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3164 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3165 times with different match limits, until it finds the minimum value needed for
3166 pcre_exec() to complete. The value is then output. This can be instructive; for
3167 most simple matches the number is quite small, but for pathological cases it
3168 gets very large very quickly.
3169
3170 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3171 returns the size of the data block pointed to by the study_data field in a
3172 pcre_extra block, that is, the value that was passed as the argument to
3173 pcre_malloc() when PCRE was getting memory in which to place the information
3174 created by pcre_study(). The fourth argument should point to a size_t variable.
3175 pcretest has been extended so that this information is shown after a successful
3176 pcre_study() call when information about the compiled regex is being displayed.
3177
3178 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3179 because what follows is always an absolute path. (Later: it turns out that this
3180 is more than cosmetic for MinGW, because it doesn't like empty path
3181 components.)
3182
3183 49. Some changes have been made to the callout feature (see 28 above):
3184
3185 (i) A callout function now has three choices for what it returns:
3186
3187 0 => success, carry on matching
3188 > 0 => failure at this point, but backtrack if possible
3189 < 0 => serious error, return this value from pcre_exec()
3190
3191 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3192 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3193 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3194 use by callout functions. It will never be used by PCRE itself.
3195
3196 (ii) The pcre_extra structure (see 45 above) has a void * field called
3197 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3198 pcre_callout_block structure has a field of the same name. The contents of
3199 the field passed in the pcre_extra structure are passed to the callout
3200 function in the corresponding field in the callout block. This makes it
3201 easier to use the same callout-containing regex from multiple threads. For
3202 testing, the pcretest program has a new data escape
3203
3204 \C*n pass the number n (may be negative) as callout_data
3205
3206 If the callout function in pcretest receives a non-zero value as
3207 callout_data, it returns that value.
3208
3209 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3210 there were some redundant $(CFLAGS) in commands that are now specified as
3211 $(LINK), which already includes $(CFLAGS).
3212
3213 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3214 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3215 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3216 one-byte characters throughout. Note that case-insensitive matching applies
3217 only to characters whose values are less than 256. PCRE doesn't support the
3218 notion of cases for higher-valued characters.
3219
3220 (i) A character class whose characters are all within 0-255 is handled as
3221 a bit map, and the map is inverted for negative classes. Previously, a
3222 character > 255 always failed to match such a class; however it should
3223 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3224
3225 (ii) A negated character class with a single character < 255 is coded as
3226 "not this character" (OP_NOT). This wasn't working properly when the test
3227 character was multibyte, either singly or repeated.
3228
3229 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3230 mode, for example: \x{100}{2,3}.
3231
3232 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3233 singly or repeated) now correctly test multibyte characters. However,
3234 PCRE doesn't recognize any characters with values greater than 255 as
3235 digits, spaces, or word characters. Such characters always match \D, \S,
3236 and \W, and never match \d, \s, or \w.
3237
3238 (v) Classes may now contain characters and character ranges with values
3239 greater than 255. For example: [ab\x{100}-\x{400}].
3240
3241 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3242 PCRE in UTF-8 mode.
3243
3244 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3245 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3246 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3247 value.)
3248
3249 53. The single man page has become too large. I have therefore split it up into
3250 a number of separate man pages. These also give rise to individual HTML pages;
3251 these are now put in a separate directory, and there is an index.html page that
3252 lists them all. Some hyperlinking between the pages has been installed.
3253
3254 54. Added convenience functions for handling named capturing parentheses.
3255
3256 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3257 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3258 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3259 are faulted.
3260
3261 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3262 calling configure. These values are used when compiling the dftables.c program
3263 which is run to generate the source of the default character tables. They
3264 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3265 you will need to set these values.
3266
3267 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3268
3269
3270 Version 3.9 02-Jan-02
3271 ---------------------
3272
3273 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3274
3275 2. If --disable-static was given, the building process failed when trying to
3276 build pcretest and pcregrep. (For some reason it was using libtool to compile
3277 them, which is not right, as they aren't part of the library.)
3278
3279
3280 Version 3.8 18-Dec-01
3281 ---------------------
3282
3283 1. The experimental UTF-8 code was completely screwed up. It was packing the
3284 bytes in the wrong order. How dumb can you get?
3285
3286
3287 Version 3.7 29-Oct-01
3288 ---------------------
3289
3290 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3291 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3292 this didn't happen under Solaris 8, where I normally test things.
3293
3294 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3295 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3296 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3297 this fix an hour or so after the initial 3.7 release.)
3298
3299
3300 Version 3.6 23-Oct-01
3301 ---------------------
3302
3303 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3304 offsets passed as NULL with zero offset count.
3305
3306 2. The config.guess and config.sub files had not been updated when I moved to
3307 the latest autoconf.
3308
3309
3310 Version 3.5 15-Aug-01
3311 ---------------------
3312
3313 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3314 had been forgotten.
3315
3316 2. By using declared but undefined structures, we can avoid using "void"
3317 definitions in pcre.h while keeping the internal definitions of the structures
3318 private.
3319
3320 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3321 user point of view, this means that both static and shared libraries are built
3322 by default, but this can be individually controlled. More of the work of
3323 handling this static/shared cases is now inside libtool instead of PCRE's make
3324 file.
3325
3326 4. The pcretest utility is now installed along with pcregrep because it is
3327 useful for users (to test regexs) and by doing this, it automatically gets
3328 relinked by libtool. The documentation has been turned into a man page, so
3329 there are now .1, .txt, and .html versions in /doc.
3330
3331 5. Upgrades to pcregrep:
3332 (i) Added long-form option names like gnu grep.
3333 (ii) Added --help to list all options with an explanatory phrase.
3334 (iii) Added -r, --recursive to recurse into sub-directories.
3335 (iv) Added -f, --file to read patterns from a file.
3336
3337 6. pcre_exec() was referring to its "code" argument before testing that
3338 argument for NULL (and giving an error if it was NULL).
3339
3340 7. Upgraded Makefile.in to allow for compiling in a different directory from
3341 the source directory.
3342
3343 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3344 options bits, the pointer it was passed was to an int instead of to an unsigned
3345 long int. This mattered only on 64-bit systems.
3346
3347 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3348 generated) instead of pcre.in, which it its source. Also made the same change
3349 in several of the .c files.
3350
3351 10. A new release of gcc defines printf() as a macro, which broke pcretest
3352 because it had an ifdef in the middle of a string argument for printf(). Fixed
3353 by using separate calls to printf().
3354
3355 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3356 script, to force use of CR or LF instead of \n in the source. On non-Unix
3357 systems, the value can be set in config.h.
3358
3359 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3360 absolute limit. Changed the text of the error message to make this clear, and
3361 likewise updated the man page.
3362
3363 13. The limit of 99 on the number of capturing subpatterns has been removed.
3364 The new limit is 65535, which I hope will not be a "real" limit.
3365
3366
3367 Version 3.4 22-Aug-00
3368 ---------------------
3369
3370 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3371
3372 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3373
3374
3375 Version 3.3 01-Aug-00
3376 ---------------------
3377
3378 1. If an octal character was given, but the value was greater than \377, it
3379 was not getting masked to the least significant bits, as documented. This could
3380 lead to crashes in some systems.
3381
3382 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3383 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3384
3385 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3386 These just pass their arguments on to (pcre_free)(), but they are provided
3387 because some uses of PCRE bind it to non-C systems that can call its functions,
3388 but cannot call free() or pcre_free() directly.
3389
3390 4. Add "make test" as a synonym for "make check". Corrected some comments in
3391 the Makefile.
3392
3393 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3394 Makefile.
3395
3396 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3397 command called pgrep for grepping around the active processes.
3398
3399 7. Added the beginnings of support for UTF-8 character strings.
3400
3401 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3402 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3403 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3404 out for the ar command.)
3405
3406
3407 Version 3.2 12-May-00
3408 ---------------------
3409
3410 This is purely a bug fixing release.
3411
3412 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3413 of ZA. This was just one example of several cases that could provoke this bug,
3414 which was introduced by change 9 of version 2.00. The code for breaking
3415 infinite loops after an iteration that matches an empty string was't working
3416 correctly.
3417
3418 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3419 when matched against abbab (for example). After matching an empty string, it
3420 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3421 caused it to match further down the string than it should.
3422
3423 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3424 was there because it doesn't seem to be needed, and it causes trouble on some
3425 systems, as it is not a Standard C header. It has been removed.
3426
3427 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3428 were reported on the Macintosh. The changes were from
3429
3430 while ((c = *(++ptr)) != 0 && c != '\n');
3431 to
3432 while ((c = *(++ptr)) != 0 && c != '\n') ;
3433
3434 Totally extraordinary, but if that's what it takes...
3435
3436 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3437 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3438 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3439 assumes the way PCRE uses memmove() (always moving upwards).
3440
3441 6. PCRE is being used in one environment where strchr() is not available. There
3442 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3443 faster code anyway.
3444
3445
3446 Version 3.1 09-Feb-00
3447 ---------------------
3448
3449 The only change in this release is the fixing of some bugs in Makefile.in for
3450 the "install" target:
3451
3452 (1) It was failing to install pcreposix.h.
3453
3454 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3455
3456
3457 Version 3.0 01-Feb-00
3458 ---------------------
3459
3460 1. Add support for the /+ modifier to perltest (to output $` like it does in
3461 pcretest).
3462
3463 2. Add support for the /g modifier to perltest.
3464
3465 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3466 matches null strings.
3467
3468 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3469 pattern. Perl treats empty patterns specially - it reuses the most recent
3470 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3471 effect.
3472
3473 5. The POSIX interface was broken in that it was just handing over the POSIX
3474 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3475 required a bigger vector, with some working space on the end. This means that
3476 the POSIX wrapper now has to get and free some memory, and copy the results.
3477
3478 6. Added some simple autoconf support, placing the test data and the
3479 documentation in separate directories, re-organizing some of the
3480 information files, and making it build pcre-config (a GNU standard). Also added
3481 libtool support for building PCRE as a shared library, which is now the
3482 default.
3483
3484 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3485 09 are not valid octal constants. Single digits will be used for minor values
3486 less than 10.
3487
3488 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3489 existing programs that set these in the POSIX interface can use PCRE without
3490 modification.
3491
3492 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3493 return all that pcre_info() returns, plus additional data. The pcre_info()
3494 function is retained for compatibility, but is considered to be obsolete.
3495
3496 10. Added experimental recursion feature (?R) to handle one common case that
3497 Perl 5.6 will be able to do with (?p{...}).
3498
3499 11. Added support for POSIX character classes like [:alpha:], which Perl is
3500 adopting.
3501
3502
3503 Version 2.08 31-Aug-99
3504 ----------------------
3505
3506 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3507 trying to match at the startoffset position, but instead was moving forward to
3508 the next newline as if a previous match had failed.
3509
3510 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3511 and could get into a loop if a null string was matched other than at the start
3512 of the subject.
3513
3514 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3515 be distinguished at compile time, and for completeness also added PCRE_DATE.
3516
3517 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3518 in GnuWin32 environments.
3519
3520
3521 Version 2.07 29-Jul-99
3522 ----------------------
3523
3524 1. The documentation is now supplied in plain text form and HTML as well as in
3525 the form of man page sources.
3526
3527 2. C++ compilers don't like assigning (void *) values to other pointer types.
3528 In particular this affects malloc(). Although there is no problem in Standard
3529 C, I've put in casts to keep C++ compilers happy.
3530
3531 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3532 should be (const char *).
3533
3534 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3535 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3536 However, I haven't made this a standard facility. The documentation doesn't
3537 mention it, and the Makefile doesn't support it.
3538
3539 5. The Makefile now contains an "install" target, with editable destinations at
3540 the top of the file. The pcretest program is not installed.
3541
3542 6. pgrep -V now gives the PCRE version number and date.
3543
3544 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3545 causing the entire string to be ignored, instead of just the last character.
3546
3547 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3548 non-matching string, it can take a very, very long time, even for strings of
3549 quite modest length, because of the nested recursion. PCRE now does better in
3550 some of these cases. It does this by remembering the last required literal
3551 character in the pattern, and pre-searching the subject to ensure it is present
3552 before running the real match. In other words, it applies a heuristic to detect
3553 some types of certain failure quickly, and in the above example, if presented
3554 with a string that has no trailing " it gives "no match" very quickly.
3555
3556 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3557 other alternatives are tried instead.
3558
3559
3560 Version 2.06 09-Jun-99
3561 ----------------------
3562
3563 1. Change pcretest's output for amount of store used to show just the code
3564 space, because the remainder (the data block) varies in size between 32-bit and
3565 64-bit systems.
3566
3567 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3568 start matching at. This allows lookbehinds to work when searching for multiple
3569 occurrences in a string.
3570
3571 3. Added additional options to pcretest for testing multiple occurrences:
3572
3573 /+ outputs the rest of the string that follows a match
3574 /g loops for multiple occurrences, using the new startoffset argument
3575 /G loops for multiple occurrences by passing an incremented pointer
3576
3577 4. PCRE wasn't doing the "first character" optimization for patterns starting
3578 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3579 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3580 the letter 'x'. On long subject strings, this gives a significant speed-up.
3581
3582
3583 Version 2.05 21-Apr-99
3584 ----------------------
3585
3586 1. Changed the type of magic_number from int to long int so that it works
3587 properly on 16-bit systems.
3588
3589 2. Fixed a bug which caused patterns starting with .* not to work correctly
3590 when the subject string contained newline characters. PCRE was assuming
3591 anchoring for such patterns in all cases, which is not correct because .* will
3592 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3593 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3594 must be retried after every newline in the subject.
3595
3596
3597 Version 2.04 18-Feb-99
3598 ----------------------
3599
3600 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3601 computation of the store needed to hold the pattern was incorrect (too large).
3602 If such patterns were nested a few deep, this could multiply and become a real
3603 problem.
3604
3605 2. Added /M option to pcretest to show the memory requirement of a specific
3606 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3607
3608 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3609 compiled in such a way that the backtracking after subsequent failure was
3610 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3611 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3612
3613
3614 Version 2.03 02-Feb-99
3615 ----------------------
3616
3617 1. Fixed typo and small mistake in man page.
3618
3619 2. Added 4th condition (GPL supersedes if conflict) and created separate
3620 LICENCE file containing the conditions.
3621
3622 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3623 Perl, that is the internal \ allows the delimiter to be included in the
3624 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3625 the final delimiter, add \ to the end of the pattern (to test the error).
3626
3627 4. Added the convenience functions for extracting substrings after a successful
3628 match. Updated pcretest to make it able to test these functions.
3629
3630
3631 Version 2.02 14-Jan-99
3632 ----------------------
3633
3634 1. Initialized the working variables associated with each extraction so that
3635 their saving and restoring doesn't refer to uninitialized store.
3636
3637 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3638 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3639 fix the problem.
3640
3641 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3642 calls, and wasn't printing the correct value for compiling calls. Increased the
3643 default value of LOOPREPEAT, and the number of significant figures in the
3644 times.
3645
3646 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3647
3648 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3649 a building problem on Windows NT with a FAT file system.
3650
3651
3652 Version 2.01 21-Oct-98
3653 ----------------------
3654
3655 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3656 to character tables built by pcre_maketables() in the current locale. If NULL
3657 is passed, the default tables are used.
3658
3659
3660 Version 2.00 24-Sep-98
3661 ----------------------
3662
3663 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3664 it any more.
3665
3666 2. Allow quantification of (?>) groups, and make it work correctly.
3667
3668 3. The first character computation wasn't working for (?>) groups.
3669
3670 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3671 end of the subject) and add 5.005's \z, which really does match only at the
3672 very end of the subject.
3673
3674 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3675
3676 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3677 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3678 localized options. All options to pcre_study() were also removed.
3679
3680 7. Add other new features from 5.005:
3681
3682 $(?<= positive lookbehind
3683 $(?<! negative lookbehind
3684 (?imsx-imsx) added the unsetting capability
3685 such a setting is global if at outer level; local otherwise
3686 (?imsx-imsx:) non-capturing groups with option setting
3687 (?(cond)re|re) conditional pattern matching
3688
3689 A backreference to itself in a repeated group matches the previous
3690 captured string.
3691
3692 8. General tidying up of studying (both automatic and via "study")
3693 consequential on the addition of new assertions.
3694
3695 9. As in 5.005, unlimited repeated groups that could match an empty substring
3696 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3697 runtime if any iteration does actually match an empty substring.
3698
3699 10. Include the RunTest script in the distribution.
3700
3701 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3702 discrepancies, some of which were old and were also with respect to 5.004. They
3703 have now been fixed.
3704
3705
3706 Version 1.09 28-Apr-98
3707 ----------------------
3708
3709 1. A negated single character class followed by a quantifier with a minimum
3710 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3711 program crashes, or just wrong answers. This did not apply to negated classes
3712 containing more than one character, or to minima other than one.
3713
3714
3715 Version 1.08 27-Mar-98
3716 ----------------------
3717
3718 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3719
3720 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3721 latter must appear before anything that relies on it in the pattern.
3722
3723
3724 Version 1.07 16-Feb-98
3725 ----------------------
3726
3727 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3728 repeat of a potentially empty string).
3729
3730
3731 Version 1.06 23-Jan-98
3732 ----------------------
3733
3734 1. Added Markus Oberhumer's little patches for C++.
3735
3736 2. Literal strings longer than 255 characters were broken.
3737
3738
3739 Version 1.05 23-Dec-97
3740 ----------------------
3741
3742 1. Negated character classes containing more than one character were failing if
3743 PCRE_CASELESS was set at run time.
3744
3745
3746 Version 1.04 19-Dec-97
3747 ----------------------
3748
3749 1. Corrected the man page, where some "const" qualifiers had been omitted.
3750
3751 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3752 input syntax.
3753
3754 3. Fixed memory leak which occurred when a regex with back references was
3755 matched with an offsets vector that wasn't big enough. The temporary memory
3756 that is used in this case wasn't being freed if the match failed.
3757
3758 4. Tidied pcretest to ensure it frees memory that it gets.
3759
3760 5. Temporary memory was being obtained in the case where the passed offsets
3761 vector was exactly big enough.
3762
3763 6. Corrected definition of offsetof() from change 5 below.
3764
3765 7. I had screwed up change 6 below and broken the rules for the use of
3766 setjmp(). Now fixed.
3767
3768
3769 Version 1.03 18-Dec-97
3770 ----------------------
3771
3772 1. A erroneous regex with a missing opening parenthesis was correctly
3773 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3774 on some systems.
3775
3776 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3777 it was reported that one broken compiler failed on the former because "code" is
3778 also an independent variable.
3779
3780 3. The erroneous regex a[]b caused an array overrun reference.
3781
3782 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3783 fail on data ending with that character. (It was going on too far, and checking
3784 the next character, typically a binary zero.) This was specific to the
3785 optimized code for single-character negative classes.
3786
3787 5. Added a contributed patch from the TIN world which does the following:
3788
3789 + Add an undef for memmove, in case the the system defines a macro for it.
3790
3791 + Add a definition of offsetof(), in case there isn't one. (I don't know
3792 the reason behind this - offsetof() is part of the ANSI standard - but
3793 it does no harm).
3794
3795 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3796 most of the places where whitespace preceded '#'. I have given up and
3797 allowed the remaining 2 cases to be at the margin.
3798
3799 + Rename some variables in pcre to eliminate shadowing. This seems very
3800 pedantic, but does no harm, of course.
3801
3802 6. Moved the call to setjmp() into its own function, to get rid of warnings
3803 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3804
3805 7. Constructs such as \d{8,} were compiling into the equivalent of
3806 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3807 outcome, but in this particular case used more store than had been allocated,
3808 which caused the bug to be discovered because it threw up an internal error.
3809
3810 8. The debugging code in both pcre and pcretest for outputting the compiled
3811 form of a regex was going wrong in the case of back references followed by
3812 curly-bracketed repeats.
3813
3814
3815 Version 1.02 12-Dec-97
3816 ----------------------
3817
3818 1. Typos in pcre.3 and comments in the source fixed.
3819
3820 2. Applied a contributed patch to get rid of places where it used to remove
3821 'const' from variables, and fixed some signed/unsigned and uninitialized
3822 variable warnings.
3823
3824 3. Added the "runtest" target to Makefile.
3825
3826 4. Set default compiler flag to -O2 rather than just -O.
3827
3828
3829 Version 1.01 19-Nov-97
3830 ----------------------
3831
3832 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3833 like /([ab]*)*/, that is, for classes with more than one character in them.
3834
3835 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3836 as /((?>a*))*/ (a PCRE_EXTRA facility).
3837
3838
3839 Version 1.00 18-Nov-97
3840 ----------------------
3841
3842 1. Added compile-time macros to support systems such as SunOS4 which don't have
3843 memmove() or strerror() but have other things that can be used instead.
3844
3845 2. Arranged that "make clean" removes the executables.
3846
3847
3848 Version 0.99 27-Oct-97
3849 ----------------------
3850
3851 1. Fixed bug in code for optimizing classes with only one character. It was
3852 initializing a 32-byte map regardless, which could cause it to run off the end
3853 of the memory it had got.
3854
3855 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3856
3857
3858 Version 0.98 22-Oct-97
3859 ----------------------
3860
3861 1. Fixed bug in code for handling temporary memory usage when there are more
3862 back references than supplied space in the ovector. This could cause segfaults.
3863
3864
3865 Version 0.97 21-Oct-97
3866 ----------------------
3867
3868 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3869
3870 2. Optimized negated single characters not to use a bit map.
3871
3872 3. Brought error texts together as macro definitions; clarified some of them;
3873 fixed one that was wrong - it said "range out of order" when it meant "invalid
3874 escape sequence".
3875
3876 4. Changed some char * arguments to const char *.
3877
3878 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3879
3880 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3881 pcretest.
3882
3883
3884 Version 0.96 16-Oct-97
3885 ----------------------
3886
3887 1. Added a simple "pgrep" utility to the distribution.
3888
3889 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3890 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3891 where "ddd" means "one or more decimal digits".
3892
3893 3. Fixed serious bug. If a pattern had a back reference, but the call to
3894 pcre_exec() didn't supply a large enough ovector to record the related
3895 identifying subpattern, the match always failed. PCRE now remembers the number
3896 of the largest back reference, and gets some temporary memory in which to save
3897 the offsets during matching if necessary, in order to ensure that
3898 backreferences always work.
3899
3900 4. Increased the compatibility with Perl in a number of ways:
3901
3902 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3903 to request this handling. The option can be set at compile or exec time.
3904
3905 (b) $ matches before a terminating newline by default; an option
3906 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3907 mode). The option can be set at compile or exec time.
3908
3909 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3910 the same as Perl's. If the decimal number it represents is less than 10
3911 or there aren't that many previous left capturing parentheses, an octal
3912 escape is read. Inside a character class, it's always an octal escape,
3913 even if it is a single digit.
3914
3915 (d) An escaped but undefined alphabetic character is taken as a literal,
3916 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3917 escapes.
3918
3919 (e) {0} is now permitted. (The previous item is removed from the compiled
3920 pattern).
3921
3922 5. Changed all the names of code files so that the basic parts are no longer
3923 than 10 characters, and abolished the teeny "globals.c" file.
3924
3925 6. Changed the handling of character classes; they are now done with a 32-byte
3926 bit map always.
3927
3928 7. Added the -d and /D options to pcretest to make it possible to look at the
3929 internals of compilation without having to recompile pcre.
3930
3931
3932 Version 0.95 23-Sep-97
3933 ----------------------
3934
3935 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3936 \x20 at the start of a run of normal characters. These were being treated as
3937 real characters, instead of the source characters being re-checked.
3938
3939
3940 Version 0.94 18-Sep-97
3941 ----------------------
3942
3943 1. The functions are now thread-safe, with the caveat that the global variables
3944 containing pointers to malloc() and free() or alternative functions are the
3945 same for all threads.
3946
3947 2. Get pcre_study() to generate a bitmap of initial characters for non-
3948 anchored patterns when this is possible, and use it if passed to pcre_exec().
3949
3950
3951 Version 0.93 15-Sep-97
3952 ----------------------
3953
3954 1. /(b)|(:+)/ was computing an incorrect first character.
3955
3956 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3957 but not actually doing anything yet.
3958
3959 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3960 as Perl does (e.g. [-az] or [az-]).
3961
3962 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3963 all possible positions.
3964
3965 5. Split up into different modules to avoid including unneeded functions in a
3966 compiled binary. However, compile and exec are still in one module. The "study"
3967 function is split off.
3968
3969 6. The character tables are now in a separate module whose source is generated
3970 by an auxiliary program - but can then be edited by hand if required. There are
3971 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3972 toupper() in the code.
3973
3974 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3975 make them global. Abolish the function for setting them, as the caller can now
3976 set them directly.
3977
3978
3979 Version 0.92 11-Sep-97
3980 ----------------------
3981
3982 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3983 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3984
3985 2. Caseless matching was not working in character classes if the characters in
3986 the pattern were in upper case.
3987
3988 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3989
3990 4. Make PCRE_ANCHORED public and accept as a compile option.
3991
3992 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3993 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3994 pass them.
3995
3996 6. Give an error if bad option bits passed at compile or run time.
3997
3998 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3999 pcretest to cause it to pass that flag.
4000
4001 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4002 options, and the first character, if set.
4003
4004 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4005
4006
4007 Version 0.91 10-Sep-97
4008 ----------------------
4009
4010 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4011 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4012
4013 2. PCRE was looping on encountering an indefinitely repeated back reference to
4014 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4015 Perl does - treats the match as successful.
4016
4017 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5