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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 561 - (show annotations) (download)
Sat Oct 30 18:37:47 2010 UTC (3 years, 10 months ago) by ph10
File size: 192853 byte(s)
Added --match-limit and --recursion-limit to pcregrep; tidied some error 
messages.

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 9. Added --match-limit and --recursion-limit to pcregrep.
61
62
63 Version 8.10 25-Jun-2010
64 ------------------------
65
66 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
67 THEN.
68
69 2. (*ACCEPT) was not working when inside an atomic group.
70
71 3. Inside a character class, \B is treated as a literal by default, but
72 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
73 causes the error). The code is unchanged, but I tidied the documentation.
74
75 4. Inside a character class, PCRE always treated \R and \X as literals,
76 whereas Perl faults them if its -w option is set. I have changed PCRE so
77 that it faults them when PCRE_EXTRA is set.
78
79 5. Added support for \N, which always matches any character other than
80 newline. (It is the same as "." when PCRE_DOTALL is not set.)
81
82 6. When compiling pcregrep with newer versions of gcc which may have
83 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
84 declared with attribute warn_unused_result" were given. Just casting the
85 result to (void) does not stop the warnings; a more elaborate fudge is
86 needed. I've used a macro to implement this.
87
88 7. Minor change to pcretest.c to avoid a compiler warning.
89
90 8. Added four artifical Unicode properties to help with an option to make
91 \s etc use properties (see next item). The new properties are: Xan
92 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
93
94 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
95 use Unicode properties. (*UCP) at the start of a pattern can be used to set
96 this option. Modified pcretest to add /W to test this facility. Added
97 REG_UCP to make it available via the POSIX interface.
98
99 10. Added --line-buffered to pcregrep.
100
101 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
102 studied, and the match started with a letter with a code point greater than
103 127 whose first byte was different to the first byte of the other case of
104 the letter, the other case of this starting letter was not recognized
105 (#976).
106
107 12. If a pattern that was studied started with a repeated Unicode property
108 test, for example, \p{Nd}+, there was the theoretical possibility of
109 setting up an incorrect bitmap of starting bytes, but fortunately it could
110 not have actually happened in practice until change 8 above was made (it
111 added property types that matched character-matching opcodes).
112
113 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
114 possible starting bytes for non-anchored patterns.
115
116 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
117 \R, and also a number of cases that involve Unicode properties, both
118 explicit and implicit when PCRE_UCP is set.
119
120 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
121 input, it could crash or give wrong results if characters with values
122 greater than 0xc0 were present in the subject string. (Detail: it assumed
123 UTF-8 input when processing these items.)
124
125 16. Added a lot of (int) casts to avoid compiler warnings in systems where
126 size_t is 64-bit (#991).
127
128 17. Added a check for running out of memory when PCRE is compiled with
129 --disable-stack-for-recursion (#990).
130
131 18. If the last data line in a file for pcretest does not have a newline on
132 the end, a newline was missing in the output.
133
134 19. The default pcre_chartables.c file recognizes only ASCII characters (values
135 less than 128) in its various bitmaps. However, there is a facility for
136 generating tables according to the current locale when PCRE is compiled. It
137 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
138 characters, are recognized by isspace() and therefore were getting set in
139 these tables, and indeed these tables seem to approximate to ISO 8859. This
140 caused a problem in UTF-8 mode when pcre_study() was used to create a list
141 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
142 which of course cannot start UTF-8 characters. I have changed the code so
143 that only real ASCII characters (less than 128) and the correct starting
144 bytes for UTF-8 encodings are set for characters greater than 127 when in
145 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
146 altogether.)
147
148 20. Added the /T option to pcretest so as to be able to run tests with non-
149 standard character tables, thus making it possible to include the tests
150 used for 19 above in the standard set of tests.
151
152 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
153 reference to a subpattern the other side of a comment that contains an
154 opening parenthesis caused either an internal compiling error, or a
155 reference to the wrong subpattern.
156
157
158 Version 8.02 19-Mar-2010
159 ------------------------
160
161 1. The Unicode data tables have been updated to Unicode 5.2.0.
162
163 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
164 configured.
165
166 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
167 original author of that file, following a query about its status.
168
169 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
170 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
171
172 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
173 quantifier applied to a forward-referencing subroutine call, could compile
174 incorrect code or give the error "internal error: previously-checked
175 referenced subpattern not found".
176
177 6. Both MS Visual Studio and Symbian OS have problems with initializing
178 variables to point to external functions. For these systems, therefore,
179 pcre_malloc etc. are now initialized to local functions that call the
180 relevant global functions.
181
182 7. There were two entries missing in the vectors called coptable and poptable
183 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
184 I've fixed the data, and added a kludgy way of testing at compile time that
185 the lengths are correct (equal to the number of opcodes).
186
187 8. Following on from 7, I added a similar kludge to check the length of the
188 eint vector in pcreposix.c.
189
190 9. Error texts for pcre_compile() are held as one long string to avoid too
191 much relocation at load time. To find a text, the string is searched,
192 counting zeros. There was no check for running off the end of the string,
193 which could happen if a new error number was added without updating the
194 string.
195
196 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
197
198 11. \K was not working if it appeared in an atomic group or in a group that
199 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
200 \K is "not well defined" if used in an assertion. PCRE now accepts it if
201 the assertion is positive, but not if it is negative.
202
203 12. Change 11 fortuitously reduced the size of the stack frame used in the
204 "match()" function of pcre_exec.c by one pointer. Forthcoming
205 implementation of support for (*MARK) will need an extra pointer on the
206 stack; I have reserved it now, so that the stack frame size does not
207 decrease.
208
209 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
210 item in branch that calls a recursion is a subroutine call - as in the
211 second branch in the above example - was incorrectly given the compile-
212 time error "recursive call could loop indefinitely" because pcre_compile()
213 was not correctly checking the subroutine for matching a non-empty string.
214
215 14. The checks for overrunning compiling workspace could trigger after an
216 overrun had occurred. This is a "should never occur" error, but it can be
217 triggered by pathological patterns such as hundreds of nested parentheses.
218 The checks now trigger 100 bytes before the end of the workspace.
219
220 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
221
222
223 Version 8.01 19-Jan-2010
224 ------------------------
225
226 1. If a pattern contained a conditional subpattern with only one branch (in
227 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
228 computed the wrong minimum data length (which is of course zero for such
229 subpatterns). This could cause incorrect "no match" results.
230
231 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
232 the pattern is reset in the first branch, pcre_compile() failed with
233 "internal error: code overflow at offset...". This happened only when
234 the reset was to the original external option setting. (An optimization
235 abstracts leading options settings into an external setting, which was the
236 cause of this.)
237
238 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
239 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
240 assertion pattern did not match (meaning that the assertion was true), it
241 was incorrectly treated as false if the SKIP had been reached during the
242 matching. This also applied to assertions used as conditions.
243
244 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
245 assertion subpattern, including such a pattern used as a condition,
246 unpredictable results occurred, instead of the error return
247 PCRE_ERROR_DFA_UITEM.
248
249 5. The C++ GlobalReplace function was not working like Perl for the special
250 situation when an empty string is matched. It now does the fancy magic
251 stuff that is necessary.
252
253 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
254 removed. (These were left over from very, very early versions of PCRE.)
255
256 7. Some cosmetic changes to the code to make life easier when compiling it
257 as part of something else:
258
259 (a) Change DEBUG to PCRE_DEBUG.
260
261 (b) In pcre_compile(), rename the member of the "branch_chain" structure
262 called "current" as "current_branch", to prevent a collision with the
263 Linux macro when compiled as a kernel module.
264
265 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
266 prevent a collision with the Linux macro when compiled as a kernel
267 module.
268
269 8. In pcre_compile() there are some checks for integer overflows that used to
270 cast potentially large values to (double). This has been changed to that
271 when building, a check for int64_t is made, and if it is found, it is used
272 instead, thus avoiding the use of floating point arithmetic. (There is no
273 other use of FP in PCRE.) If int64_t is not found, the fallback is to
274 double.
275
276 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
277 2005 (difference between two addresses compared to an unsigned value).
278
279 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
280 custom one, because of the following reported problem in Windows:
281
282 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
283 under Win32.
284 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
285 therefore missing the function definition.
286 - The compiler thus generates a "C" signature for the test function.
287 - The linker fails to find the "C" function.
288 - PCRE fails to configure if asked to do so against libbz2.
289
290 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
291 messages were output:
292
293 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
294 rerunning libtoolize, to keep the correct libtool macros in-tree.
295 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
296
297 I have done both of these things.
298
299 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
300 most of the time, it *can* run out if it is given a pattern that contains a
301 runaway infinite recursion. I updated the discussion in the pcrestack man
302 page.
303
304 13. Now that we have gone to the x.xx style of version numbers, the minor
305 version may start with zero. Using 08 or 09 is a bad idea because users
306 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
307 interpreted as invalid octal numbers. I've updated the previous comment in
308 configure.ac, and also added a check that gives an error if 08 or 09 are
309 used.
310
311 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
312 causing partial matching to fail when the end of the subject matched \W
313 in a UTF-8 pattern where \W was quantified with a minimum of 3.
314
315 15. There were some discrepancies between the declarations in pcre_internal.h
316 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
317 their definitions. The declarations used "const uschar *" and the
318 definitions used USPTR. Even though USPTR is normally defined as "const
319 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
320 reported that: "This difference in casting confuses some C++ compilers, for
321 example, SunCC recognizes above declarations as different functions and
322 generates broken code for hbpcre." I have changed the declarations to use
323 USPTR.
324
325 16. GNU libtool is named differently on some systems. The autogen.sh script now
326 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
327 (FreeBSD).
328
329 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
330 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
331 comment: "Figure out how to create a longlong from a string: strtoll and
332 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
333 instance, but it only takes 2 args instead of 3!"
334
335 18. A subtle bug concerned with back references has been fixed by a change of
336 specification, with a corresponding code fix. A pattern such as
337 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
338 refers, was giving matches when it shouldn't. For example, xa=xaaa would
339 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
340 same bug. Such groups have to be quantified to be useful, or contained
341 inside another quantified group. (If there's no repetition, the reference
342 can never match.) The problem arises because, having left the group and
343 moved on to the rest of the pattern, a later failure that backtracks into
344 the group uses the captured value from the final iteration of the group
345 rather than the correct earlier one. I have fixed this in PCRE by forcing
346 any group that contains a reference to itself to be an atomic group; that
347 is, there cannot be any backtracking into it once it has completed. This is
348 similar to recursive and subroutine calls.
349
350
351 Version 8.00 19-Oct-09
352 ----------------------
353
354 1. The table for translating pcre_compile() error codes into POSIX error codes
355 was out-of-date, and there was no check on the pcre_compile() error code
356 being within the table. This could lead to an OK return being given in
357 error.
358
359 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
360 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
361 in a Windows environment.
362
363 3. The pcregrep --count option prints the count for each file even when it is
364 zero, as does GNU grep. However, pcregrep was also printing all files when
365 --files-with-matches was added. Now, when both options are given, it prints
366 counts only for those files that have at least one match. (GNU grep just
367 prints the file name in this circumstance, but including the count seems
368 more useful - otherwise, why use --count?) Also ensured that the
369 combination -clh just lists non-zero counts, with no names.
370
371 4. The long form of the pcregrep -F option was incorrectly implemented as
372 --fixed_strings instead of --fixed-strings. This is an incompatible change,
373 but it seems right to fix it, and I didn't think it was worth preserving
374 the old behaviour.
375
376 5. The command line items --regex=pattern and --regexp=pattern were not
377 recognized by pcregrep, which required --regex pattern or --regexp pattern
378 (with a space rather than an '='). The man page documented the '=' forms,
379 which are compatible with GNU grep; these now work.
380
381 6. No libpcreposix.pc file was created for pkg-config; there was just
382 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
383
384 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
385 when UCP support is not needed, by modifying the Python script that
386 generates it from Unicode data files. This should not matter if the module
387 is correctly used as a library, but I received one complaint about 50K of
388 unwanted data. My guess is that the person linked everything into his
389 program rather than using a library. Anyway, it does no harm.
390
391 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
392 was a minimum greater than 1 for a wide character in a possessive
393 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
394 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
395 character. Chaos in the form of incorrect output or a compiling loop could
396 result.
397
398 9. The restrictions on what a pattern can contain when partial matching is
399 requested for pcre_exec() have been removed. All patterns can now be
400 partially matched by this function. In addition, if there are at least two
401 slots in the offset vector, the offset of the earliest inspected character
402 for the match and the offset of the end of the subject are set in them when
403 PCRE_ERROR_PARTIAL is returned.
404
405 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
406 synonymous with PCRE_PARTIAL, for backwards compatibility, and
407 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
408 and may be more useful for multi-segment matching.
409
410 11. Partial matching with pcre_exec() is now more intuitive. A partial match
411 used to be given if ever the end of the subject was reached; now it is
412 given only if matching could not proceed because another character was
413 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
414 string "Z", which now yields "no match" instead of "partial match". In the
415 case of pcre_dfa_exec(), "no match" is given if every matching path for the
416 final character ended with (*FAIL).
417
418 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
419 if the pattern had a "must contain" character that was already found in the
420 earlier partial match, unless partial matching was again requested. For
421 example, with the pattern /dog.(body)?/, the "must contain" character is
422 "g". If the first part-match was for the string "dog", restarting with
423 "sbody" failed. This bug has been fixed.
424
425 13. The string returned by pcre_dfa_exec() after a partial match has been
426 changed so that it starts at the first inspected character rather than the
427 first character of the match. This makes a difference only if the pattern
428 starts with a lookbehind assertion or \b or \B (\K is not supported by
429 pcre_dfa_exec()). It's an incompatible change, but it makes the two
430 matching functions compatible, and I think it's the right thing to do.
431
432 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
433 so that the demonstration program is easily available in environments where
434 PCRE has not been installed from source.
435
436 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
437 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
438 library.
439
440 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
441 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
442 is not the first non-POSIX option to be added. Clearly some people find
443 these options useful.
444
445 17. If a caller to the POSIX matching function regexec() passes a non-zero
446 value for nmatch with a NULL value for pmatch, the value of
447 nmatch is forced to zero.
448
449 18. RunGrepTest did not have a test for the availability of the -u option of
450 the diff command, as RunTest does. It now checks in the same way as
451 RunTest, and also checks for the -b option.
452
453 19. If an odd number of negated classes containing just a single character
454 interposed, within parentheses, between a forward reference to a named
455 subpattern and the definition of the subpattern, compilation crashed with
456 an internal error, complaining that it could not find the referenced
457 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
458 [The bug was that it was starting one character too far in when skipping
459 over the character class, thus treating the ] as data rather than
460 terminating the class. This meant it could skip too much.]
461
462 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
463 /g option in pcretest when the pattern contains \K, which makes it possible
464 to have an empty string match not at the start, even when the pattern is
465 anchored. Updated pcretest and pcredemo to use this option.
466
467 21. If the maximum number of capturing subpatterns in a recursion was greater
468 than the maximum at the outer level, the higher number was returned, but
469 with unset values at the outer level. The correct (outer level) value is
470 now given.
471
472 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
473 PCRE did not set those parentheses (unlike Perl). I have now found a way to
474 make it do so. The string so far is captured, making this feature
475 compatible with Perl.
476
477 23. The tests have been re-organized, adding tests 11 and 12, to make it
478 possible to check the Perl 5.10 features against Perl 5.10.
479
480 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
481 pattern matches a fixed length string. PCRE did not allow this; now it
482 does. Neither allows recursion.
483
484 25. I finally figured out how to implement a request to provide the minimum
485 length of subject string that was needed in order to match a given pattern.
486 (It was back references and recursion that I had previously got hung up
487 on.) This code has now been added to pcre_study(); it finds a lower bound
488 to the length of subject needed. It is not necessarily the greatest lower
489 bound, but using it to avoid searching strings that are too short does give
490 some useful speed-ups. The value is available to calling programs via
491 pcre_fullinfo().
492
493 26. While implementing 25, I discovered to my embarrassment that pcretest had
494 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
495 study optimizations had never been tested with that matching function.
496 Oops. What is worse, even when it was passed study data, there was a bug in
497 pcre_dfa_exec() that meant it never actually used it. Double oops. There
498 were also very few tests of studied patterns with pcre_dfa_exec().
499
500 27. If (?| is used to create subpatterns with duplicate numbers, they are now
501 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
502 on the other side of the coin, they are no longer allowed to have different
503 names, because these cannot be distinguished in PCRE, and this has caused
504 confusion. (This is a difference from Perl.)
505
506 28. When duplicate subpattern names are present (necessarily with different
507 numbers, as required by 27 above), and a test is made by name in a
508 conditional pattern, either for a subpattern having been matched, or for
509 recursion in such a pattern, all the associated numbered subpatterns are
510 tested, and the overall condition is true if the condition is true for any
511 one of them. This is the way Perl works, and is also more like the way
512 testing by number works.
513
514
515 Version 7.9 11-Apr-09
516 ---------------------
517
518 1. When building with support for bzlib/zlib (pcregrep) and/or readline
519 (pcretest), all targets were linked against these libraries. This included
520 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
521 libraries. This caused unwanted dependencies to be created. This problem
522 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
523 pcretest is linked with readline.
524
525 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
526 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
527 moved outside it again, because FALSE and TRUE are already defined in AIX,
528 but BOOL is not.
529
530 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
531 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
532
533 4. The pcregrep documentation said spaces were inserted as well as colons (or
534 hyphens) following file names and line numbers when outputting matching
535 lines. This is not true; no spaces are inserted. I have also clarified the
536 wording for the --colour (or --color) option.
537
538 5. In pcregrep, when --colour was used with -o, the list of matching strings
539 was not coloured; this is different to GNU grep, so I have changed it to be
540 the same.
541
542 6. When --colo(u)r was used in pcregrep, only the first matching substring in
543 each matching line was coloured. Now it goes on to look for further matches
544 of any of the test patterns, which is the same behaviour as GNU grep.
545
546 7. A pattern that could match an empty string could cause pcregrep to loop; it
547 doesn't make sense to accept an empty string match in pcregrep, so I have
548 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
549 seems to be how GNU grep behaves.
550
551 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
552 start or after a newline", because the conditional assertion was not being
553 correctly handled. The rule now is that both the assertion and what follows
554 in the first alternative must satisfy the test.
555
556 9. If auto-callout was enabled in a pattern with a conditional group whose
557 condition was an assertion, PCRE could crash during matching, both with
558 pcre_exec() and pcre_dfa_exec().
559
560 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
561 used for matching.
562
563 11. Unicode property support in character classes was not working for
564 characters (bytes) greater than 127 when not in UTF-8 mode.
565
566 12. Added the -M command line option to pcretest.
567
568 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
569
570 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
571
572 16. Added comments and documentation about mis-use of no_arg in the C++
573 wrapper.
574
575 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
576 from Martin Jerabek that uses macro names for all relevant character and
577 string constants.
578
579 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
580 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
581 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
582 these, but not everybody uses configure.
583
584 19. A conditional group that had only one branch was not being correctly
585 recognized as an item that could match an empty string. This meant that an
586 enclosing group might also not be so recognized, causing infinite looping
587 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
588 with the subject "ab", where knowledge that the repeated group can match
589 nothing is needed in order to break the loop.
590
591 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
592 exec(), but without supplying a callout function, matching went wrong.
593
594 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
595 leak if the size of the offset vector was greater than 30. When the vector
596 is smaller, the saved offsets during recursion go onto a local stack
597 vector, but for larger vectors malloc() is used. It was failing to free
598 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
599 error, in fact).
600
601 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
602 heapframe that is used only when UTF-8 support is enabled. This caused no
603 problem, but was untidy.
604
605 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
606 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
607 included within another project.
608
609 24. Steven Van Ingelgem's patches to add more options to the CMake support,
610 slightly modified by me:
611
612 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
613 not building pcregrep.
614
615 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
616 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
617
618 25. Forward references, both numeric and by name, in patterns that made use of
619 duplicate group numbers, could behave incorrectly or give incorrect errors,
620 because when scanning forward to find the reference group, PCRE was not
621 taking into account the duplicate group numbers. A pattern such as
622 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
623
624 26. Changed a few more instances of "const unsigned char *" to USPTR, making
625 the feature of a custom pointer more persuasive (as requested by a user).
626
627 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
628 pcretest.c, inside #ifndefs, because it seems they are sometimes already
629 pre-defined.
630
631 28. Added support for (*UTF8) at the start of a pattern.
632
633 29. Arrange for flags added by the "release type" setting in CMake to be shown
634 in the configuration summary.
635
636
637 Version 7.8 05-Sep-08
638 ---------------------
639
640 1. Replaced UCP searching code with optimized version as implemented for Ad
641 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
642 stage table and inline lookup instead of a function, giving speed ups of 2
643 to 5 times on some simple patterns that I tested. Permission was given to
644 distribute the MultiStage2.py script that generates the tables (it's not in
645 the tarball, but is in the Subversion repository).
646
647 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
648 scripts.
649
650 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
651 a group with a zero qualifier. The result of the study could be incorrect,
652 or the function might crash, depending on the pattern.
653
654 4. Caseless matching was not working for non-ASCII characters in back
655 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
656 It now works when Unicode Property Support is available.
657
658 5. In pcretest, an escape such as \x{de} in the data was always generating
659 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
660 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
661 truncation.
662
663 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
664
665 7. Added two (int) casts to pcregrep when printing the difference of two
666 pointers, in case they are 64-bit values.
667
668 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
669 test 2 if it fails.
670
671 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
672 and a #define of that name to empty if it is not externally set. This is to
673 allow users of MSVC to set it if necessary.
674
675 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
676 the convenience functions in the pcre_get.c source file.
677
678 11. An option change at the start of a pattern that had top-level alternatives
679 could cause overwriting and/or a crash. This command provoked a crash in
680 some environments:
681
682 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
683
684 This potential security problem was recorded as CVE-2008-2371.
685
686 12. For a pattern where the match had to start at the beginning or immediately
687 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
688 pcre_dfa_exec() could read past the end of the passed subject if there was
689 no match. To help with detecting such bugs (e.g. with valgrind), I modified
690 pcretest so that it places the subject at the end of its malloc-ed buffer.
691
692 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
693 exec() might read past the end of the data buffer in UTF-8 mode.
694
695 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
696 the data contained the byte 0x85 as part of a UTF-8 character within its
697 first line. This applied both to normal and DFA matching.
698
699 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
700 /^[^d]*?$/8 failed to match "abc".
701
702 16. Added a missing copyright notice to pcrecpp_internal.h.
703
704 17. Make it more clear in the documentation that values returned from
705 pcre_exec() in ovector are byte offsets, not character counts.
706
707 18. Tidied a few places to stop certain compilers from issuing warnings.
708
709 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
710 supplied by Stefan Weber. I made a further small update for 7.8 because
711 there is a change of source arrangements: the pcre_searchfuncs.c module is
712 replaced by pcre_ucd.c.
713
714
715 Version 7.7 07-May-08
716 ---------------------
717
718 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
719 a string to a long long, pretend we don't even have a long long." This is
720 done by checking for the strtoq, strtoll, and _strtoi64 functions.
721
722 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
723 pre-7.6 versions, which defined a global no_arg variable instead of putting
724 it in the RE class. (See also #8 below.)
725
726 3. Remove a line of dead code, identified by coverity and reported by Nuno
727 Lopes.
728
729 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
730
731 (1) The include/exclude patterns were being applied to the whole pathnames
732 of files, instead of just to the final components.
733
734 (2) If there was more than one level of directory, the subdirectories were
735 skipped unless they satisfied the include/exclude conditions. This is
736 inconsistent with GNU grep (and could even be seen as contrary to the
737 pcregrep specification - which I improved to make it absolutely clear).
738 The action now is always to scan all levels of directory, and just
739 apply the include/exclude patterns to regular files.
740
741 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
742 --exclude_dir in the tests to avoid scanning .svn directories.
743
744 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
745 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
746 doesn't support NULs in patterns.
747
748 7. Added some missing "const"s to declarations of static tables in
749 pcre_compile.c and pcre_dfa_exec.c.
750
751 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
752 caused by fix #2 above. (Subsequently also a second patch to fix the
753 first patch. And a third patch - this was a messy problem.)
754
755 9. Applied Craig's patch to remove the use of push_back().
756
757 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
758 matching function regexec().
759
760 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
761 which, however, unlike Perl's \g{...}, are subroutine calls, not back
762 references. PCRE supports relative numbers with this syntax (I don't think
763 Oniguruma does).
764
765 12. Previously, a group with a zero repeat such as (...){0} was completely
766 omitted from the compiled regex. However, this means that if the group
767 was called as a subroutine from elsewhere in the pattern, things went wrong
768 (an internal error was given). Such groups are now left in the compiled
769 pattern, with a new opcode that causes them to be skipped at execution
770 time.
771
772 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
773 to the way PCRE behaves:
774
775 (a) A lone ] character is dis-allowed (Perl treats it as data).
776
777 (b) A back reference to an unmatched subpattern matches an empty string
778 (Perl fails the current match path).
779
780 (c) A data ] in a character class must be notated as \] because if the
781 first data character in a class is ], it defines an empty class. (In
782 Perl it is not possible to have an empty class.) The empty class []
783 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
784 The negative empty class [^] matches any one character, independently
785 of the DOTALL setting.
786
787 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
788 non-existent subpattern following a character class starting with ']' and
789 containing () gave an internal compiling error instead of "reference to
790 non-existent subpattern". Fortunately, when the pattern did exist, the
791 compiled code was correct. (When scanning forwards to check for the
792 existencd of the subpattern, it was treating the data ']' as terminating
793 the class, so got the count wrong. When actually compiling, the reference
794 was subsequently set up correctly.)
795
796 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
797 it was being rejected as not supported by pcre_dfa_exec(), even though
798 other assertions are supported. I have made pcre_dfa_exec() support
799 (*FAIL).
800
801 16. The implementation of 13c above involved the invention of a new opcode,
802 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
803 cannot be changed at match time, I realized I could make a small
804 improvement to matching performance by compiling OP_ALLANY instead of
805 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
806 on the OP_ANY path.
807
808 17. Compiling pcretest on Windows with readline support failed without the
809 following two fixes: (1) Make the unistd.h include conditional on
810 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
811
812 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
813 ncurses library to be included for pcretest when ReadLine support is
814 requested, but also to allow for it to be overridden. This patch came from
815 Daniel Bergström.
816
817 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
818 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
819 any errors with the current Unicode tables. Thanks to Peter Kankowski for
820 spotting this.
821
822
823 Version 7.6 28-Jan-08
824 ---------------------
825
826 1. A character class containing a very large number of characters with
827 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
828 overflow.
829
830 2. Patch to cut out the "long long" test in pcrecpp_unittest when
831 HAVE_LONG_LONG is not defined.
832
833 3. Applied Christian Ehrlicher's patch to update the CMake build files to
834 bring them up to date and include new features. This patch includes:
835
836 - Fixed PH's badly added libz and libbz2 support.
837 - Fixed a problem with static linking.
838 - Added pcredemo. [But later removed - see 7 below.]
839 - Fixed dftables problem and added an option.
840 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
841 HAVE_LONG_LONG.
842 - Added readline support for pcretest.
843 - Added an listing of the option settings after cmake has run.
844
845 4. A user submitted a patch to Makefile that makes it easy to create
846 "pcre.dll" under mingw when using Configure/Make. I added stuff to
847 Makefile.am that cause it to include this special target, without
848 affecting anything else. Note that the same mingw target plus all
849 the other distribution libraries and programs are now supported
850 when configuring with CMake (see 6 below) instead of with
851 Configure/Make.
852
853 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
854 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
855 exported in the Windows port". It has not yet been confirmed that the patch
856 solves the problem, but it does no harm.
857
858 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
859 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
860 with CMake, and also correct the comment about stack recursion.
861
862 7. Remove the automatic building of pcredemo from the ./configure system and
863 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
864 of a program that users should build themselves after PCRE is installed, so
865 building it automatically is not really right. What is more, it gave
866 trouble in some build environments.
867
868 8. Further tidies to CMakeLists.txt from Sheri and Christian.
869
870
871 Version 7.5 10-Jan-08
872 ---------------------
873
874 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
875 values in parens when parsing an RE using the C++ wrapper."
876
877 2. Negative specials like \S did not work in character classes in UTF-8 mode.
878 Characters greater than 255 were excluded from the class instead of being
879 included.
880
881 3. The same bug as (2) above applied to negated POSIX classes such as
882 [:^space:].
883
884 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
885 defined or documented. It seems to have been a typo for PCRE_STATIC, so
886 I have changed it.
887
888 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
889 first named subpattern) and a construct such as (?&a) would reference the
890 first named subpattern whose name started with "a" (in other words, the
891 length check was missing). Both these problems are fixed. "Subpattern name
892 expected" is now given for (?&) (a zero-length name), and this patch also
893 makes it give the same error for \k'' (previously it complained that that
894 was a reference to a non-existent subpattern).
895
896 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
897 this is right because (?- can be followed by option settings as well as by
898 digits. I have, however, made the messages clearer.
899
900 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
901 than the number used in the conditional) now cause a compile-time error.
902 This is actually not compatible with Perl, which accepts such patterns, but
903 treats the conditional as always being FALSE (as PCRE used to), but it
904 seems to me that giving a diagnostic is better.
905
906 8. Change "alphameric" to the more common word "alphanumeric" in comments
907 and messages.
908
909 9. Fix two occurrences of "backslash" in comments that should have been
910 "backspace".
911
912 10. Remove two redundant lines of code that can never be obeyed (their function
913 was moved elsewhere).
914
915 11. The program that makes PCRE's Unicode character property table had a bug
916 which caused it to generate incorrect table entries for sequences of
917 characters that have the same character type, but are in different scripts.
918 It amalgamated them into a single range, with the script of the first of
919 them. In other words, some characters were in the wrong script. There were
920 thirteen such cases, affecting characters in the following ranges:
921
922 U+002b0 - U+002c1
923 U+0060c - U+0060d
924 U+0061e - U+00612
925 U+0064b - U+0065e
926 U+0074d - U+0076d
927 U+01800 - U+01805
928 U+01d00 - U+01d77
929 U+01d9b - U+01dbf
930 U+0200b - U+0200f
931 U+030fc - U+030fe
932 U+03260 - U+0327f
933 U+0fb46 - U+0fbb1
934 U+10450 - U+1049d
935
936 12. The -o option (show only the matching part of a line) for pcregrep was not
937 compatible with GNU grep in that, if there was more than one match in a
938 line, it showed only the first of them. It now behaves in the same way as
939 GNU grep.
940
941 13. If the -o and -v options were combined for pcregrep, it printed a blank
942 line for every non-matching line. GNU grep prints nothing, and pcregrep now
943 does the same. The return code can be used to tell if there were any
944 non-matching lines.
945
946 14. Added --file-offsets and --line-offsets to pcregrep.
947
948 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
949 infinitely looping recursion. The bug was that positive lookaheads were not
950 being skipped when checking for a possible empty match (negative lookaheads
951 and both kinds of lookbehind were skipped).
952
953 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
954 inclusion of <windows.h> to before rather than after the definition of
955 INVALID_FILE_ATTRIBUTES (patch from David Byron).
956
957 17. Specifying a possessive quantifier with a specific limit for a Unicode
958 character property caused pcre_compile() to compile bad code, which led at
959 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
960 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
961 caused the error; without that there was no problem.
962
963 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
964
965 19. Added --enable-pcretest-libreadline.
966
967 20. In pcrecpp.cc, the variable 'count' was incremented twice in
968 RE::GlobalReplace(). As a result, the number of replacements returned was
969 double what it should be. I removed one of the increments, but Craig sent a
970 later patch that removed the other one (the right fix) and added unit tests
971 that check the return values (which was not done before).
972
973 21. Several CMake things:
974
975 (1) Arranged that, when cmake is used on Unix, the libraries end up with
976 the names libpcre and libpcreposix, not just pcre and pcreposix.
977
978 (2) The above change means that pcretest and pcregrep are now correctly
979 linked with the newly-built libraries, not previously installed ones.
980
981 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
982
983 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
984 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
985 UTF-8 newline character). The key issue is that the pattern starts .*;
986 this means that the match must be either at the beginning, or after a
987 newline. The bug was in the code for advancing after a failed match and
988 checking that the new position followed a newline. It was not taking
989 account of UTF-8 characters correctly.
990
991 23. PCRE was behaving differently from Perl in the way it recognized POSIX
992 character classes. PCRE was not treating the sequence [:...:] as a
993 character class unless the ... were all letters. Perl, however, seems to
994 allow any characters between [: and :], though of course it rejects as
995 unknown any "names" that contain non-letters, because all the known class
996 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
997 for example, whereas PCRE did not - it did not recognize a POSIX character
998 class. This seemed a bit dangerous, so the code has been changed to be
999 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1000 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1001 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1002 Perl does, and where it didn't before.
1003
1004 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1005 Windows environments %n is disabled by default.
1006
1007
1008 Version 7.4 21-Sep-07
1009 ---------------------
1010
1011 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1012 means that a class such as [\s] counted as "explicit reference to CR or
1013 LF". That isn't really right - the whole point of the change was to try to
1014 help when there was an actual mention of one of the two characters. So now
1015 the change happens only if \r or \n (or a literal CR or LF) character is
1016 encountered.
1017
1018 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1019 of both had grown to the point where there were only 3 bits left.
1020 Fortunately, there was spare space in the data structure, and so I have
1021 moved the internal flags into a new 16-bit field to free up more option
1022 bits.
1023
1024 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1025 but did not set the internal JCHANGED flag - either of these is enough to
1026 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1027 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1028 start sets both bits.
1029
1030 4. Added options (at build time, compile time, exec time) to change \R from
1031 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1032
1033 5. doc/pcresyntax.html was missing from the distribution.
1034
1035 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1036 compatibility, even though it is no longer used.
1037
1038 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1039 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1040 windows.h file is present (where different names are used). [This was
1041 reversed later after testing - see 16 below.]
1042
1043 8. Changed all #include <config.h> to #include "config.h". There were also
1044 some further <pcre.h> cases that I changed to "pcre.h".
1045
1046 9. When pcregrep was used with the --colour option, it missed the line ending
1047 sequence off the lines that it output.
1048
1049 10. It was pointed out to me that arrays of string pointers cause lots of
1050 relocations when a shared library is dynamically loaded. A technique of
1051 using a single long string with a table of offsets can drastically reduce
1052 these. I have refactored PCRE in four places to do this. The result is
1053 dramatic:
1054
1055 Originally: 290
1056 After changing UCP table: 187
1057 After changing error message table: 43
1058 After changing table of "verbs" 36
1059 After changing table of Posix names 22
1060
1061 Thanks to the folks working on Gregex for glib for this insight.
1062
1063 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1064 unicode-properties was also set.
1065
1066 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1067
1068 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1069 checked only for CRLF.
1070
1071 14. Added casts to pcretest.c to avoid compiler warnings.
1072
1073 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1074
1075 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1076 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1077 entirely. This removes changes made in 7 above.
1078
1079 17. The CMake files have been updated, and there is now more information about
1080 building with CMake in the NON-UNIX-USE document.
1081
1082
1083 Version 7.3 28-Aug-07
1084 ---------------------
1085
1086 1. In the rejigging of the build system that eventually resulted in 7.1, the
1087 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1088 brackets there is not right, since it causes compilers to look for an
1089 installed pcre.h, not the version that is in the source that is being
1090 compiled (which of course may be different). I have changed it back to:
1091
1092 #include "pcre.h"
1093
1094 I have a vague recollection that the change was concerned with compiling in
1095 different directories, but in the new build system, that is taken care of
1096 by the VPATH setting the Makefile.
1097
1098 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1099 when the subject happened to end in the byte 0x85 (e.g. if the last
1100 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1101 characters but of course it shouldn't be taken as a newline when it is part
1102 of another character. The bug was that, for an unlimited repeat of . in
1103 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1104 characters when looking for a newline.
1105
1106 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1107
1108 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1109 in debug output.
1110
1111 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1112 long printing in the pcrecpp unittest when running under MinGW.
1113
1114 6. ESC_K was left out of the EBCDIC table.
1115
1116 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1117 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1118 limit also applies to "virtual nesting" when a pattern is recursive, and in
1119 this case 1000 isn't so big. I have been able to remove this limit at the
1120 expense of backing off one optimization in certain circumstances. Normally,
1121 when pcre_exec() would call its internal match() function recursively and
1122 immediately return the result unconditionally, it uses a "tail recursion"
1123 feature to save stack. However, when a subpattern that can match an empty
1124 string has an unlimited repetition quantifier, it no longer makes this
1125 optimization. That gives it a stack frame in which to save the data for
1126 checking that an empty string has been matched. Previously this was taken
1127 from the 1000-entry workspace that had been reserved. So now there is no
1128 explicit limit, but more stack is used.
1129
1130 8. Applied Daniel's patches to solve problems with the import/export magic
1131 syntax that is required for Windows, and which was going wrong for the
1132 pcreposix and pcrecpp parts of the library. These were overlooked when this
1133 problem was solved for the main library.
1134
1135 9. There were some crude static tests to avoid integer overflow when computing
1136 the size of patterns that contain repeated groups with explicit upper
1137 limits. As the maximum quantifier is 65535, the maximum group length was
1138 set at 30,000 so that the product of these two numbers did not overflow a
1139 32-bit integer. However, it turns out that people want to use groups that
1140 are longer than 30,000 bytes (though not repeat them that many times).
1141 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1142 made it possible to implement the integer overflow checks in a much more
1143 dynamic way, which I have now done. The artificial limitation on group
1144 length has been removed - we now have only the limit on the total length of
1145 the compiled pattern, which depends on the LINK_SIZE setting.
1146
1147 10. Fixed a bug in the documentation for get/copy named substring when
1148 duplicate names are permitted. If none of the named substrings are set, the
1149 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1150 empty string.
1151
1152 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1153 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1154 because the ] is interpreted as the first data character and the
1155 terminating ] is not found. PCRE has been made compatible with Perl in this
1156 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1157 cause memory overwriting.
1158
1159 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1160 string has been matched (to stop an infinite loop). It was not recognizing
1161 a conditional subpattern that could match an empty string if that
1162 subpattern was within another subpattern. For example, it looped when
1163 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1164 condition was not nested. This bug has been fixed.
1165
1166 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1167 past the start of the subject in the presence of bytes with the top bit
1168 set, for example "\x8aBCD".
1169
1170 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1171 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1172
1173 14. Optimized (?!) to (*FAIL).
1174
1175 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1176 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1177 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1178 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1179 does: it's just the validity check that is more restrictive.
1180
1181 16. Inserted checks for integer overflows during escape sequence (backslash)
1182 processing, and also fixed erroneous offset values for syntax errors during
1183 backslash processing.
1184
1185 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1186 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1187
1188 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1189 caused an overrun.
1190
1191 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1192 something other than just ASCII characters) inside a group that had an
1193 unlimited repeat caused a loop at compile time (while checking to see
1194 whether the group could match an empty string).
1195
1196 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1197 [\P{Any}] did so. (Error in the code for printing property names.)
1198
1199 21. An orphan \E inside a character class could cause a crash.
1200
1201 22. A repeated capturing bracket such as (A)? could cause a wild memory
1202 reference during compilation.
1203
1204 23. There are several functions in pcre_compile() that scan along a compiled
1205 expression for various reasons (e.g. to see if it's fixed length for look
1206 behind). There were bugs in these functions when a repeated \p or \P was
1207 present in the pattern. These operators have additional parameters compared
1208 with \d, etc, and these were not being taken into account when moving along
1209 the compiled data. Specifically:
1210
1211 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1212 length.
1213
1214 (b) An item such as \pL+ within a repeated group could cause crashes or
1215 loops.
1216
1217 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1218 "reference to non-existent subpattern" error.
1219
1220 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1221
1222 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1223 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1224
1225 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1226
1227 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1228 character were causing crashes (broken optimization).
1229
1230 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1231 \p or \P) caused a compile-time loop.
1232
1233 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1234 break. For example, the unstudied pattern [\r\n]A does not match the string
1235 "\r\nA" because change 7.0/46 below moves the current point on by two
1236 characters after failing to match at the start. However, the pattern \nA
1237 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1238 the same is true. There doesn't seem any very clean way out of this, but
1239 what I have chosen to do makes the common cases work: PCRE now takes note
1240 of whether there can be an explicit match for \r or \n anywhere in the
1241 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1242 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1243 pattern has explicit CR or LF references.
1244
1245 29. Added (*CR) etc for changing newline setting at start of pattern.
1246
1247
1248 Version 7.2 19-Jun-07
1249 ---------------------
1250
1251 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1252 which is apparently normally available under Windows.
1253
1254 2. Re-jig the pcregrep tests with different newline settings in an attempt
1255 to make them independent of the local environment's newline setting.
1256
1257 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1258
1259 4. Some of the "internals" tests were previously cut out when the link size
1260 was not 2, because the output contained actual offsets. The recent new
1261 "Z" feature of pcretest means that these can be cut out, making the tests
1262 usable with all link sizes.
1263
1264 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1265 stack recursion. This gives a massive performance boost under BSD, but just
1266 a small improvement under Linux. However, it saves one field in the frame
1267 in all cases.
1268
1269 6. Added more features from the forthcoming Perl 5.10:
1270
1271 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1272 recursion call. It refers to the nth most recently opened parentheses.
1273
1274 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1275 to be opened parentheses.
1276
1277 (c) Conditions that refer to capturing parentheses can be specified
1278 relatively, for example, (?(-2)... or (?(+3)...
1279
1280 (d) \K resets the start of the current match so that everything before
1281 is not part of it.
1282
1283 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1284
1285 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1286 reference syntax.
1287
1288 (g) (?| introduces a group in which the numbering of parentheses in each
1289 alternative starts with the same number.
1290
1291 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1292
1293 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1294 PCRE_INFO_JCHANGED.
1295
1296 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1297 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1298 for detecting groups that can match an empty string.
1299
1300 9. A pattern with a very large number of alternatives (more than several
1301 hundred) was running out of internal workspace during the pre-compile
1302 phase, where pcre_compile() figures out how much memory will be needed. A
1303 bit of new cunning has reduced the workspace needed for groups with
1304 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1305 workspace instead of running out of the 4096 that are available.
1306
1307 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1308
1309 11. Applied patch from Google to remove an optimization that didn't quite work.
1310 The report of the bug said:
1311
1312 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1313 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1314 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1315
1316 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1317 it matched the wrong number of bytes.
1318
1319
1320 Version 7.1 24-Apr-07
1321 ---------------------
1322
1323 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1324 that is more "standard", making use of automake and other Autotools. There
1325 is some re-arrangement of the files and adjustment of comments consequent
1326 on this.
1327
1328 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1329 for recursive directory scanning broke on some systems because the files
1330 are not scanned in any specific order and on different systems the order
1331 was different. A call to "sort" has been inserted into RunGrepTest for the
1332 approprate test as a short-term fix. In the longer term there may be an
1333 alternative.
1334
1335 3. I had an email from Eric Raymond about problems translating some of PCRE's
1336 man pages to HTML (despite the fact that I distribute HTML pages, some
1337 people do their own conversions for various reasons). The problems
1338 concerned the use of low-level troff macros .br and .in. I have therefore
1339 removed all such uses from the man pages (some were redundant, some could
1340 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1341 HTML has been updated to handle .nf/.fi and to complain if it encounters
1342 .br or .in.
1343
1344 4. Updated comments in configure.ac that get placed in config.h.in and also
1345 arranged for config.h to be included in the distribution, with the name
1346 config.h.generic, for the benefit of those who have to compile without
1347 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1348
1349 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1350 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1351 makevp.bat for the current PCRE, using the additional files
1352 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1353
1354 6. A Windows user reported a minor discrepancy with test 2, which turned out
1355 to be caused by a trailing space on an input line that had got lost in his
1356 copy. The trailing space was an accident, so I've just removed it.
1357
1358 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1359 that is needed.
1360
1361 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1362 as "const" (a) because they are and (b) because it helps the PHP
1363 maintainers who have recently made a script to detect big data structures
1364 in the php code that should be moved to the .rodata section. I remembered
1365 to update Builducptable as well, so it won't revert if ucptable.h is ever
1366 re-created.
1367
1368 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1369 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1370 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1371 support is not required. This saves 1.5-2K of code, which is important in
1372 some applications.
1373
1374 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1375 so as not to refer to the tables, even though these functions will never be
1376 called when UTF-8 support is disabled. Otherwise there are problems with a
1377 shared library.
1378
1379 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1380
1381 (a) It was defining its arguments as char * instead of void *.
1382
1383 (b) It was assuming that all moves were upwards in memory; this was true
1384 a long time ago when I wrote it, but is no longer the case.
1385
1386 The emulated memove() is provided for those environments that have neither
1387 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1388 is clearly not the case, as these two bugs were recently reported.
1389
1390 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1391 and Detrail to create the HTML documentation, the .txt form of the man
1392 pages, and it removes trailing spaces from listed files. It also creates
1393 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1394 case, it wraps all the #defines with #ifndefs. This script should be run
1395 before "make dist".
1396
1397 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1398 with Unicode property support.
1399
1400 (a) For a maximizing quantifier, if the two different cases of the
1401 character were of different lengths in their UTF-8 codings (there are
1402 some cases like this - I found 11), and the matching function had to
1403 back up over a mixture of the two cases, it incorrectly assumed they
1404 were both the same length.
1405
1406 (b) When PCRE was configured to use the heap rather than the stack for
1407 recursion during matching, it was not correctly preserving the data for
1408 the other case of a UTF-8 character when checking ahead for a match
1409 while processing a minimizing repeat. If the check also involved
1410 matching a wide character, but failed, corruption could cause an
1411 erroneous result when trying to check for a repeat of the original
1412 character.
1413
1414 13. Some tidying changes to the testing mechanism:
1415
1416 (a) The RunTest script now detects the internal link size and whether there
1417 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1418 values substituted by "configure". (The RunGrepTest script already did
1419 this for UTF-8.) The configure.ac script no longer substitutes the
1420 relevant variables.
1421
1422 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1423 with length and offset values. This means that the output is different
1424 for different internal link sizes. Test 2 is skipped for link sizes
1425 other than 2 because of this, bypassing the problem. Unfortunately,
1426 there was also a test in test 3 (the locale tests) that used /B and
1427 failed for link sizes other than 2. Rather than cut the whole test out,
1428 I have added a new /Z option to pcretest that replaces the length and
1429 offset values with spaces. This is now used to make test 3 independent
1430 of link size. (Test 2 will be tidied up later.)
1431
1432 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1433 segmentation fault instead of returning the appropriate error message.
1434
1435 15. In multiline mode when the newline sequence was set to "any", the pattern
1436 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1437 This doesn't seem right; it now treats the CRLF combination as the line
1438 ending, and so does not match in that case. It's only a pattern such as ^$
1439 that would hit this one: something like ^ABC$ would have failed after \r
1440 and then tried again after \r\n.
1441
1442 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1443 in an attempt to make files that differ only in their line terminators
1444 compare equal. This works on Linux.
1445
1446 17. Under certain error circumstances pcregrep might try to free random memory
1447 as it exited. This is now fixed, thanks to valgrind.
1448
1449 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1450 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1451 was because its rules for how to advance for /g after matching an empty
1452 string at the end of a line did not allow for this case. They now check for
1453 it specially.
1454
1455 20. pcretest is supposed to handle patterns and data of any length, by
1456 extending its buffers when necessary. It was getting this wrong when the
1457 buffer for a data line had to be extended.
1458
1459 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1460 CRLF as a newline sequence.
1461
1462 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1463 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1464 I have nevertheless tidied it up.
1465
1466 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1467
1468 24. Added a man page for pcre-config.
1469
1470
1471 Version 7.0 19-Dec-06
1472 ---------------------
1473
1474 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1475 moving to gcc 4.1.1.
1476
1477 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1478 sys/time.h, which is documented as needed for this function. It doesn't
1479 seem to matter on Linux, but it showed up on some releases of OS X.
1480
1481 3. It seems that there are systems where bytes whose values are greater than
1482 127 match isprint() in the "C" locale. The "C" locale should be the
1483 default when a C program starts up. In most systems, only ASCII printing
1484 characters match isprint(). This difference caused the output from pcretest
1485 to vary, making some of the tests fail. I have changed pcretest so that:
1486
1487 (a) When it is outputting text in the compiled version of a pattern, bytes
1488 other than 32-126 are always shown as hex escapes.
1489
1490 (b) When it is outputting text that is a matched part of a subject string,
1491 it does the same, unless a different locale has been set for the match
1492 (using the /L modifier). In this case, it uses isprint() to decide.
1493
1494 4. Fixed a major bug that caused incorrect computation of the amount of memory
1495 required for a compiled pattern when options that changed within the
1496 pattern affected the logic of the preliminary scan that determines the
1497 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1498 that the computed length was too small. The symptoms of this bug were
1499 either the PCRE error "internal error: code overflow" from pcre_compile(),
1500 or a glibc crash with a message such as "pcretest: free(): invalid next
1501 size (fast)". Examples of patterns that provoked this bug (shown in
1502 pcretest format) are:
1503
1504 /(?-x: )/x
1505 /(?x)(?-x: \s*#\s*)/
1506 /((?i)[\x{c0}])/8
1507 /(?i:[\x{c0}])/8
1508
1509 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1510 is now done differently.
1511
1512 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1513 wrapper classes; (b) implement a new function in the C++ scanner that is
1514 more efficient than the old way of doing things because it avoids levels of
1515 recursion in the regex matching; (c) add a paragraph to the documentation
1516 for the FullMatch() function.
1517
1518 6. The escape sequence \n was being treated as whatever was defined as
1519 "newline". Not only was this contrary to the documentation, which states
1520 that \n is character 10 (hex 0A), but it also went horribly wrong when
1521 "newline" was defined as CRLF. This has been fixed.
1522
1523 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1524 was being set to -1 for the "end of line" case (supposedly a value that no
1525 character can have). Though this value is never used (the check for end of
1526 line is "zero bytes in current character"), it caused compiler complaints.
1527 I've changed it to 0xffffffff.
1528
1529 8. In pcre_version.c, the version string was being built by a sequence of
1530 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1531 string (as it is for production releases) called a macro with an empty
1532 argument. The C standard says the result of this is undefined. The gcc
1533 compiler treats it as an empty string (which was what was wanted) but it is
1534 reported that Visual C gives an error. The source has been hacked around to
1535 avoid this problem.
1536
1537 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1538 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1539 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1540 of them did).
1541
1542 10. Originally, pcretest opened its input and output without "b"; then I was
1543 told that "b" was needed in some environments, so it was added for release
1544 5.0 to both the input and output. (It makes no difference on Unix-like
1545 systems.) Later I was told that it is wrong for the input on Windows. I've
1546 now abstracted the modes into two macros, to make it easier to fiddle with
1547 them, and removed "b" from the input mode under Windows.
1548
1549 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1550
1551 12. Added -help and --help to pcretest as an official way of being reminded
1552 of the options.
1553
1554 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1555 and pcrecpp.cc because they annoy compilers at high warning levels.
1556
1557 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1558
1559 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1560 scripts are not C programs :-) and which was not noticed because it works
1561 on Linux.
1562
1563 16. pcretest is supposed to handle any length of pattern and data line (as one
1564 line or as a continued sequence of lines) by extending its input buffer if
1565 necessary. This feature was broken for very long pattern lines, leading to
1566 a string of junk being passed to pcre_compile() if the pattern was longer
1567 than about 50K.
1568
1569 17. I have done a major re-factoring of the way pcre_compile() computes the
1570 amount of memory needed for a compiled pattern. Previously, there was code
1571 that made a preliminary scan of the pattern in order to do this. That was
1572 OK when PCRE was new, but as the facilities have expanded, it has become
1573 harder and harder to keep it in step with the real compile phase, and there
1574 have been a number of bugs (see for example, 4 above). I have now found a
1575 cunning way of running the real compile function in a "fake" mode that
1576 enables it to compute how much memory it would need, while actually only
1577 ever using a few hundred bytes of working memory and without too many
1578 tests of the mode. This should make future maintenance and development
1579 easier. A side effect of this work is that the limit of 200 on the nesting
1580 depth of parentheses has been removed (though this was never a serious
1581 limitation, I suspect). However, there is a downside: pcre_compile() now
1582 runs more slowly than before (30% or more, depending on the pattern). I
1583 hope this isn't a big issue. There is no effect on runtime performance.
1584
1585 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1586 newline (only possible for the last line of a file) and it was a
1587 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1588
1589 19. Added additional timing features to pcretest. (1) The -tm option now times
1590 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1591 separate command line item, by a number that specifies the number of
1592 repeats to use when timing. The default is 50000; this gives better
1593 precision, but takes uncomfortably long for very large patterns.
1594
1595 20. Extended pcre_study() to be more clever in cases where a branch of a
1596 subpattern has no definite first character. For example, (a*|b*)[cd] would
1597 previously give no result from pcre_study(). Now it recognizes that the
1598 first character must be a, b, c, or d.
1599
1600 21. There was an incorrect error "recursive call could loop indefinitely" if
1601 a subpattern (or the entire pattern) that was being tested for matching an
1602 empty string contained only one non-empty item after a nested subpattern.
1603 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1604 incorrectly, because the \d was being skipped in the check.
1605
1606 22. The pcretest program now has a new pattern option /B and a command line
1607 option -b, which is equivalent to adding /B to every pattern. This causes
1608 it to show the compiled bytecode, without the additional information that
1609 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1610 is the same as /B/I).
1611
1612 23. A new optimization is now able automatically to treat some sequences such
1613 as a*b as a*+b. More specifically, if something simple (such as a character
1614 or a simple class like \d) has an unlimited quantifier, and is followed by
1615 something that cannot possibly match the quantified thing, the quantifier
1616 is automatically "possessified".
1617
1618 24. A recursive reference to a subpattern whose number was greater than 39
1619 went wrong under certain circumstances in UTF-8 mode. This bug could also
1620 have affected the operation of pcre_study().
1621
1622 25. Realized that a little bit of performance could be had by replacing
1623 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1624
1625 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1626
1627 27. Possessive quantifiers such as a++ were previously implemented by turning
1628 them into atomic groups such as ($>a+). Now they have their own opcodes,
1629 which improves performance. This includes the automatically created ones
1630 from 23 above.
1631
1632 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1633 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1634 the first matched character to be a colon. This applied both to named and
1635 numbered groups.
1636
1637 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1638
1639 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1640 building PCRE on Windows easier, so I have included it in the distribution.
1641
1642 31. There is now a check in pcretest against a ridiculously large number being
1643 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1644 loop, the loop is abandoned.
1645
1646 32. Forward references to subpatterns in conditions such as (?(2)...) where
1647 subpattern 2 is defined later cause pcre_compile() to search forwards in
1648 the pattern for the relevant set of parentheses. This search went wrong
1649 when there were unescaped parentheses in a character class, parentheses
1650 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1651
1652 33. "Subroutine" calls and backreferences were previously restricted to
1653 referencing subpatterns earlier in the regex. This restriction has now
1654 been removed.
1655
1656 34. Added a number of extra features that are going to be in Perl 5.10. On the
1657 whole, these are just syntactic alternatives for features that PCRE had
1658 previously implemented using the Python syntax or my own invention. The
1659 other formats are all retained for compatibility.
1660
1661 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1662 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1663 also .NET compatible.
1664
1665 (b) A recursion or subroutine call to a named group can now be defined as
1666 (?&name) as well as (?P>name).
1667
1668 (c) A backreference to a named group can now be defined as \k<name> or
1669 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1670 5.10, are also .NET compatible.
1671
1672 (d) A conditional reference to a named group can now use the syntax
1673 (?(<name>) or (?('name') as well as (?(name).
1674
1675 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1676 groups (named and numbered) that are never evaluated inline, but can be
1677 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1678 is always false. There may be only one alternative in such a group.
1679
1680 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1681 as the simple (?(R). The condition is true only if the most recent
1682 recursion is that of the given number or name. It does not search out
1683 through the entire recursion stack.
1684
1685 (g) The escape \gN or \g{N} has been added, where N is a positive or
1686 negative number, specifying an absolute or relative reference.
1687
1688 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1689 some "unreachable code" warnings.
1690
1691 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1692 things, this adds five new scripts.
1693
1694 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1695 There were also incompatibilities regarding the handling of \Q..\E inside
1696 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1697 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1698
1699 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1700 matches an empty string, and forcibly breaks the loop. There were bugs in
1701 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1702 against aaaa the result was just "a" rather than "aaaa", for example. Two
1703 separate and independent bugs (that affected different cases) have been
1704 fixed.
1705
1706 39. Refactored the code to abolish the use of different opcodes for small
1707 capturing bracket numbers. This is a tidy that I avoided doing when I
1708 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1709 The new approach is not only tidier, it makes it possible to reduce the
1710 memory needed to fix the previous bug (38).
1711
1712 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1713 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1714 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1715 mode.
1716
1717 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1718 report.
1719
1720 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1721 copy construction and assignment in the C++ wrapper.
1722
1723 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1724 couple of bugs that could have given wrong results in the "--newline=crlf"
1725 case.
1726
1727 44. Added a number of casts and did some reorganization of signed/unsigned int
1728 variables following suggestions from Dair Grant. Also renamed the variable
1729 "this" as "item" because it is a C++ keyword.
1730
1731 45. Arranged for dftables to add
1732
1733 #include "pcre_internal.h"
1734
1735 to pcre_chartables.c because without it, gcc 4.x may remove the array
1736 definition from the final binary if PCRE is built into a static library and
1737 dead code stripping is activated.
1738
1739 46. For an unanchored pattern, if a match attempt fails at the start of a
1740 newline sequence, and the newline setting is CRLF or ANY, and the next two
1741 characters are CRLF, advance by two characters instead of one.
1742
1743
1744 Version 6.7 04-Jul-06
1745 ---------------------
1746
1747 1. In order to handle tests when input lines are enormously long, pcretest has
1748 been re-factored so that it automatically extends its buffers when
1749 necessary. The code is crude, but this _is_ just a test program. The
1750 default size has been increased from 32K to 50K.
1751
1752 2. The code in pcre_study() was using the value of the re argument before
1753 testing it for NULL. (Of course, in any sensible call of the function, it
1754 won't be NULL.)
1755
1756 3. The memmove() emulation function in pcre_internal.h, which is used on
1757 systems that lack both memmove() and bcopy() - that is, hardly ever -
1758 was missing a "static" storage class specifier.
1759
1760 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1761 containing an extended class (one that cannot be represented by a bitmap
1762 because it contains high-valued characters or Unicode property items, e.g.
1763 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1764 pattern, but PCRE should not loop if you do not (it no longer does).
1765 [Detail: two cases were found: (a) a repeated subpattern containing an
1766 extended class; (b) a recursive reference to a subpattern that followed a
1767 previous extended class. It wasn't skipping over the extended class
1768 correctly when UTF-8 mode was not set.]
1769
1770 5. A negated single-character class was not being recognized as fixed-length
1771 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1772 compile error "lookbehind assertion is not fixed length".
1773
1774 6. The RunPerlTest auxiliary script was showing an unexpected difference
1775 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1776 write a Perl script that can interpret lines of an input file either as
1777 byte characters or as UTF-8, which is what "perltest" was being required to
1778 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1779 can't do is switch easily at run time between having the "use utf8;" pragma
1780 or not. In the end, I fudged it by using the RunPerlTest script to insert
1781 "use utf8;" explicitly for the UTF-8 tests.
1782
1783 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1784 the end of the subject string, contrary to the documentation and to what
1785 Perl does. This was true of both matching functions. Now it matches only at
1786 the start of the subject and immediately after *internal* newlines.
1787
1788 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1789 a pointer to an int instead of a pointer to an unsigned long int. This
1790 caused problems on 64-bit systems.
1791
1792 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1793 instance of the 'standard' template library not being so standard".
1794
1795 10. There was no check on the number of named subpatterns nor the maximum
1796 length of a subpattern name. The product of these values is used to compute
1797 the size of the memory block for a compiled pattern. By supplying a very
1798 long subpattern name and a large number of named subpatterns, the size
1799 computation could be caused to overflow. This is now prevented by limiting
1800 the length of names to 32 characters, and the number of named subpatterns
1801 to 10,000.
1802
1803 11. Subpatterns that are repeated with specific counts have to be replicated in
1804 the compiled pattern. The size of memory for this was computed from the
1805 length of the subpattern and the repeat count. The latter is limited to
1806 65535, but there was no limit on the former, meaning that integer overflow
1807 could in principle occur. The compiled length of a repeated subpattern is
1808 now limited to 30,000 bytes in order to prevent this.
1809
1810 12. Added the optional facility to have named substrings with the same name.
1811
1812 13. Added the ability to use a named substring as a condition, using the
1813 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1814 are numbers (not recommended). Forward references are permitted.
1815
1816 14. Added forward references in named backreferences (if you see what I mean).
1817
1818 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1819 pattern could run off the end of the subject. For example, the pattern
1820 "(?s)(.{1,5})"8 did this with the subject "ab".
1821
1822 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1823 PCRE_CASELESS was set when matching characters that were quantified with ?
1824 or *.
1825
1826 17. A character class other than a single negated character that had a minimum
1827 but no maximum quantifier - for example [ab]{6,} - was not handled
1828 correctly by pce_dfa_exec(). It would match only one character.
1829
1830 18. A valid (though odd) pattern that looked like a POSIX character
1831 class but used an invalid character after [ (for example [[,abc,]]) caused
1832 pcre_compile() to give the error "Failed: internal error: code overflow" or
1833 in some cases to crash with a glibc free() error. This could even happen if
1834 the pattern terminated after [[ but there just happened to be a sequence of
1835 letters, a binary zero, and a closing ] in the memory that followed.
1836
1837 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1838 over the years. Originally (before any Unicode support), just the bottom 8
1839 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1840 output from "man perlunicode" includes this:
1841
1842 The regular expression compiler produces polymorphic opcodes. That
1843 is, the pattern adapts to the data and automatically switches to
1844 the Unicode character scheme when presented with Unicode data--or
1845 instead uses a traditional byte scheme when presented with byte
1846 data.
1847
1848 Sadly, a wide octal escape does not cause a switch, and in a string with
1849 no other multibyte characters, these octal escapes are treated as before.
1850 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1851 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1852 Unicode string.
1853
1854 I have not perpetrated such confusion in PCRE. Up till now, it took just
1855 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1856 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1857 translate to the appropriate multibyte character.
1858
1859 29. Applied some refactoring to reduce the number of warnings from Microsoft
1860 and Borland compilers. This has included removing the fudge introduced
1861 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1862 a warning about an unused variable.
1863
1864 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1865 characters since release 4.0, because Perl (from release 5.004) does not.
1866 [Or at least, is documented not to: some releases seem to be in conflict
1867 with the documentation.] However, when a pattern was studied with
1868 pcre_study() and all its branches started with \s, PCRE still included VT
1869 as a possible starting character. Of course, this did no harm; it just
1870 caused an unnecessary match attempt.
1871
1872 22. Removed a now-redundant internal flag bit that recorded the fact that case
1873 dependency changed within the pattern. This was once needed for "required
1874 byte" processing, but is no longer used. This recovers a now-scarce options
1875 bit. Also moved the least significant internal flag bit to the most-
1876 significant bit of the word, which was not previously used (hangover from
1877 the days when it was an int rather than a uint) to free up another bit for
1878 the future.
1879
1880 23. Added support for CRLF line endings as well as CR and LF. As well as the
1881 default being selectable at build time, it can now be changed at runtime
1882 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1883 specify that it is scanning data with non-default line endings.
1884
1885 24. Changed the definition of CXXLINK to make it agree with the definition of
1886 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1887
1888 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1889 recursions. This makes a big different to stack usage for some patterns.
1890
1891 26. If a subpattern containing a named recursion or subroutine reference such
1892 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1893 the space required for the compiled pattern went wrong and gave too small a
1894 value. Depending on the environment, this could lead to "Failed: internal
1895 error: code overflow at offset 49" or "glibc detected double free or
1896 corruption" errors.
1897
1898 27. Applied patches from Google (a) to support the new newline modes and (b) to
1899 advance over multibyte UTF-8 characters in GlobalReplace.
1900
1901 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1902 difference for some implementation of PCRE in some Windows version.
1903
1904 29. Added some extra testing facilities to pcretest:
1905
1906 \q<number> in a data line sets the "match limit" value
1907 \Q<number> in a data line sets the "match recursion limt" value
1908 -S <number> sets the stack size, where <number> is in megabytes
1909
1910 The -S option isn't available for Windows.
1911
1912
1913 Version 6.6 06-Feb-06
1914 ---------------------
1915
1916 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1917 in pcreposix.h. I have copied the definition from pcre.h.
1918
1919 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1920 because pcre.h is no longer a built file.
1921
1922 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1923 not normally included in the compiled code.
1924
1925
1926 Version 6.5 01-Feb-06
1927 ---------------------
1928
1929 1. When using the partial match feature with pcre_dfa_exec(), it was not
1930 anchoring the second and subsequent partial matches at the new starting
1931 point. This could lead to incorrect results. For example, with the pattern
1932 /1234/, partially matching against "123" and then "a4" gave a match.
1933
1934 2. Changes to pcregrep:
1935
1936 (a) All non-match returns from pcre_exec() were being treated as failures
1937 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1938 error message is output. Some extra information is given for the
1939 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1940 probably the only errors that are likely to be caused by users (by
1941 specifying a regex that has nested indefinite repeats, for instance).
1942 If there are more than 20 of these errors, pcregrep is abandoned.
1943
1944 (b) A binary zero was treated as data while matching, but terminated the
1945 output line if it was written out. This has been fixed: binary zeroes
1946 are now no different to any other data bytes.
1947
1948 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1949 used to set a locale for matching. The --locale=xxxx long option has
1950 been added (no short equivalent) to specify a locale explicitly on the
1951 pcregrep command, overriding the environment variables.
1952
1953 (d) When -B was used with -n, some line numbers in the output were one less
1954 than they should have been.
1955
1956 (e) Added the -o (--only-matching) option.
1957
1958 (f) If -A or -C was used with -c (count only), some lines of context were
1959 accidentally printed for the final match.
1960
1961 (g) Added the -H (--with-filename) option.
1962
1963 (h) The combination of options -rh failed to suppress file names for files
1964 that were found from directory arguments.
1965
1966 (i) Added the -D (--devices) and -d (--directories) options.
1967
1968 (j) Added the -F (--fixed-strings) option.
1969
1970 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1971
1972 (l) Added the --colo(u)r option.
1973
1974 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1975 is not present by default.
1976
1977 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1978 items such as (?R) or (?1), when the recursion could match a number of
1979 alternatives. If it matched one of the alternatives, but subsequently,
1980 outside the recursion, there was a failure, the code tried to back up into
1981 the recursion. However, because of the way PCRE is implemented, this is not
1982 possible, and the result was an incorrect result from the match.
1983
1984 In order to prevent this happening, the specification of recursion has
1985 been changed so that all such subpatterns are automatically treated as
1986 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1987
1988 4. I had overlooked the fact that, in some locales, there are characters for
1989 which isalpha() is true but neither isupper() nor islower() are true. In
1990 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1991 and ordfeminine) are like this. This affected the treatment of \w and \W
1992 when they appeared in character classes, but not when they appeared outside
1993 a character class. The bit map for "word" characters is now created
1994 separately from the results of isalnum() instead of just taking it from the
1995 upper, lower, and digit maps. (Plus the underscore character, of course.)
1996
1997 5. The above bug also affected the handling of POSIX character classes such as
1998 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1999 permanent tables. Instead, the bit maps for such a class were previously
2000 created as the appropriate unions of the upper, lower, and digit bitmaps.
2001 Now they are created by subtraction from the [[:word:]] class, which has
2002 its own bitmap.
2003
2004 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2005 It is created by subtracting the vertical space characters (\x09, \x0a,
2006 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2007 subtraction was done in the overall bitmap for a character class, meaning
2008 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2009 be recognized. This bug has been fixed.
2010
2011 7. Patches from the folks at Google:
2012
2013 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2014 real life, but is still worth protecting against".
2015
2016 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2017 regular expressions".
2018
2019 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2020 have it.
2021
2022 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2023 "configure" and the latter not, in order to fix a problem somebody had
2024 with compiling the Arg class on HP-UX.
2025
2026 (e) Improve the error-handling of the C++ wrapper a little bit.
2027
2028 (f) New tests for checking recursion limiting.
2029
2030 8. The pcre_memmove() function, which is used only if the environment does not
2031 have a standard memmove() function (and is therefore rarely compiled),
2032 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2033 returning a result (though PCRE never actually uses the result).
2034
2035 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2036 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2037 returned instead of calling malloc() with an overflowing number that would
2038 most likely cause subsequent chaos.
2039
2040 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2041
2042 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2043 with this option is matched, the nmatch and pmatch options of regexec() are
2044 ignored.
2045
2046 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2047 provided in case anyone wants to the the POSIX interface with UTF-8
2048 strings.
2049
2050 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2051 C++ linking (needed for some HP-UX environments).
2052
2053 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2054 (unused parameter) and in the pcre_printint() function (omitted "default"
2055 switch label when the default is to do nothing).
2056
2057 15. Added some code to make it possible, when PCRE is compiled as a C++
2058 library, to replace subject pointers for pcre_exec() with a smart pointer
2059 class, thus making it possible to process discontinuous strings.
2060
2061 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2062 much the same function. They were added by different people who were trying
2063 to make PCRE easy to compile on non-Unix systems. It has been suggested
2064 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2065 for compiling on Windows systems. I have therefore replaced it with
2066 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2067 defaults to "extern" for C or "extern C" for C++, which works fine on
2068 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2069 SCOPE with something explicit in config.h. In addition:
2070
2071 (a) pcreposix.h still had just "extern" instead of either of these macros;
2072 I have replaced it with PCRE_DATA_SCOPE.
2073
2074 (b) Functions such as _pcre_xclass(), which are internal to the library,
2075 but external in the C sense, all had PCRE_EXPORT in their definitions.
2076 This is apparently wrong for the Windows case, so I have removed it.
2077 (It makes no difference on Unix-like systems.)
2078
2079 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2080 of recursive calls to match(). This is different to MATCH_LIMIT because
2081 that limits the total number of calls to match(), not all of which increase
2082 the depth of recursion. Limiting the recursion depth limits the amount of
2083 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2084 when PCRE is compiled, and changed at run time. A patch from Google adds
2085 this functionality to the C++ interface.
2086
2087 18. Changes to the handling of Unicode character properties:
2088
2089 (a) Updated the table to Unicode 4.1.0.
2090
2091 (b) Recognize characters that are not in the table as "Cn" (undefined).
2092
2093 (c) I revised the way the table is implemented to a much improved format
2094 which includes recognition of ranges. It now supports the ranges that
2095 are defined in UnicodeData.txt, and it also amalgamates other
2096 characters into ranges. This has reduced the number of entries in the
2097 table from around 16,000 to around 3,000, thus reducing its size
2098 considerably. I realized I did not need to use a tree structure after
2099 all - a binary chop search is just as efficient. Having reduced the
2100 number of entries, I extended their size from 6 bytes to 8 bytes to
2101 allow for more data.
2102
2103 (d) Added support for Unicode script names via properties such as \p{Han}.
2104
2105 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2106 matching that character.
2107
2108 20. When matching a repeated Unicode property with a minimum greater than zero,
2109 (for example \pL{2,}), PCRE could look past the end of the subject if it
2110 reached it while seeking the minimum number of characters. This could
2111 happen only if some of the characters were more than one byte long, because
2112 there is a check for at least the minimum number of bytes.
2113
2114 21. Refactored the implementation of \p and \P so as to be more general, to
2115 allow for more different types of property in future. This has changed the
2116 compiled form incompatibly. Anybody with saved compiled patterns that use
2117 \p or \P will have to recompile them.
2118
2119 22. Added "Any" and "L&" to the supported property types.
2120
2121 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2122 but give a compile time error if the value is greater than 0xff.
2123
2124 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2125 accidentally not being installed or uninstalled.
2126
2127 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2128 made were to insert the current release number. This seemed silly, because
2129 it made things harder for people building PCRE on systems that don't run
2130 "configure". I have turned pcre.h into a distributed file, no longer built
2131 by "configure", with the version identification directly included. There is
2132 no longer a pcre.h.in file.
2133
2134 However, this change necessitated a change to the pcre-config script as
2135 well. It is built from pcre-config.in, and one of the substitutions was the
2136 release number. I have updated configure.ac so that ./configure now finds
2137 the release number by grepping pcre.h.
2138
2139 26. Added the ability to run the tests under valgrind.
2140
2141
2142 Version 6.4 05-Sep-05
2143 ---------------------
2144
2145 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2146 "--" to be printed when multiple files were scanned, even when none of the
2147 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2148 consider it to be a bug, and have restored the previous behaviour.
2149
2150 2. A couple of code tidies to get rid of compiler warnings.
2151
2152 3. The pcretest program used to cheat by referring to symbols in the library
2153 whose names begin with _pcre_. These are internal symbols that are not
2154 really supposed to be visible externally, and in some environments it is
2155 possible to suppress them. The cheating is now confined to including
2156 certain files from the library's source, which is a bit cleaner.
2157
2158 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2159 file's purpose clearer.
2160
2161 5. Reorganized pcre_ucp_findchar().
2162
2163
2164 Version 6.3 15-Aug-05
2165 ---------------------
2166
2167 1. The file libpcre.pc.in did not have general read permission in the tarball.
2168
2169 2. There were some problems when building without C++ support:
2170
2171 (a) If C++ support was not built, "make install" and "make test" still
2172 tried to test it.
2173
2174 (b) There were problems when the value of CXX was explicitly set. Some
2175 changes have been made to try to fix these, and ...
2176
2177 (c) --disable-cpp can now be used to explicitly disable C++ support.
2178
2179 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2180 backslash in a target when C++ was disabled. This confuses some
2181 versions of "make", apparently. Using an intermediate variable solves
2182 this. (Same for CPP_LOBJ.)
2183
2184 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2185 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2186 necessary on certain architectures.
2187
2188 4. Added a setting of -export-symbols-regex to the link command to remove
2189 those symbols that are exported in the C sense, but actually are local
2190 within the library, and not documented. Their names all begin with
2191 "_pcre_". This is not a perfect job, because (a) we have to except some
2192 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2193 available (and never for static libraries). I have made a note to try to
2194 find a way round (a) in the future.
2195
2196
2197 Version 6.2 01-Aug-05
2198 ---------------------
2199
2200 1. There was no test for integer overflow of quantifier values. A construction
2201 such as {1111111111111111} would give undefined results. What is worse, if
2202 a minimum quantifier for a parenthesized subpattern overflowed and became
2203 negative, the calculation of the memory size went wrong. This could have
2204 led to memory overwriting.
2205
2206 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2207
2208 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2209 operating environments where this matters.
2210
2211 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2212 PCRE options from within the C++ wrapper.
2213
2214 5. Named capturing subpatterns were not being correctly counted when a pattern
2215 was compiled. This caused two problems: (a) If there were more than 100
2216 such subpatterns, the calculation of the memory needed for the whole
2217 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2218 back references of the form \12, where the number was greater than 9, were
2219 not recognized as back references, even though there were sufficient
2220 previous subpatterns.
2221
2222 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2223 versions of gcc, e.g. 2.95.4.
2224
2225
2226 Version 6.1 21-Jun-05
2227 ---------------------
2228
2229 1. There was one reference to the variable "posix" in pcretest.c that was not
2230 surrounded by "#if !defined NOPOSIX".
2231
2232 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2233 the cross-check on the old pcre_info() function, for the benefit of the
2234 cut-down version of PCRE that is currently imported into Exim.
2235
2236 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2237 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2238 patterns that start (?i) but I don't think that matters. The use of (?i) is
2239 just an example; this all applies to the other options as well.
2240
2241 4. Since libtool seems to echo the compile commands it is issuing, the output
2242 from "make" can be reduced a bit by putting "@" in front of each libtool
2243 compile command.
2244
2245 5. Patch from the folks at Google for configure.in to be a bit more thorough
2246 in checking for a suitable C++ installation before trying to compile the
2247 C++ stuff. This should fix a reported problem when a compiler was present,
2248 but no suitable headers.
2249
2250 6. The man pages all had just "PCRE" as their title. I have changed them to
2251 be the relevant file name. I have also arranged that these names are
2252 retained in the file doc/pcre.txt, which is a concatenation in text format
2253 of all the man pages except the little individual ones for each function.
2254
2255 7. The NON-UNIX-USE file had not been updated for the different set of source
2256 files that come with release 6. I also added a few comments about the C++
2257 wrapper.
2258
2259
2260 Version 6.0 07-Jun-05
2261 ---------------------
2262
2263 1. Some minor internal re-organization to help with my DFA experiments.
2264
2265 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2266 didn't matter for the library itself when fully configured, but did matter
2267 when compiling without UCP support, or within Exim, where the ucp files are
2268 not imported.
2269
2270 3. Refactoring of the library code to split up the various functions into
2271 different source modules. The addition of the new DFA matching code (see
2272 below) to a single monolithic source would have made it really too
2273 unwieldy, quite apart from causing all the code to be include in a
2274 statically linked application, when only some functions are used. This is
2275 relevant even without the DFA addition now that patterns can be compiled in
2276 one application and matched in another.
2277
2278 The downside of splitting up is that there have to be some external
2279 functions and data tables that are used internally in different modules of
2280 the library but which are not part of the API. These have all had their
2281 names changed to start with "_pcre_" so that they are unlikely to clash
2282 with other external names.
2283
2284 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2285 a different (DFA) algorithm. Although it is slower than the original
2286 function, it does have some advantages for certain types of matching
2287 problem.
2288
2289 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2290 including restarting after a partial match.
2291
2292 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2293 defined when compiling for Windows was sent to me. I have put it into the
2294 code, though I have no means of testing or verifying it.
2295
2296 7. Added the pcre_refcount() auxiliary function.
2297
2298 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2299 match before or at the first newline in the subject string. In pcretest,
2300 the /f option on a pattern can be used to set this.
2301
2302 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2303 would behave wrongly. This has been present in PCRE since release 4.0.
2304
2305 10. A number of changes to the pcregrep command:
2306
2307 (a) Refactored how -x works; insert ^(...)$ instead of setting
2308 PCRE_ANCHORED and checking the length, in preparation for adding
2309 something similar for -w.
2310
2311 (b) Added the -w (match as a word) option.
2312
2313 (c) Refactored the way lines are read and buffered so as to have more
2314 than one at a time available.
2315
2316 (d) Implemented a pcregrep test script.
2317
2318 (e) Added the -M (multiline match) option. This allows patterns to match
2319 over several lines of the subject. The buffering ensures that at least
2320 8K, or the rest of the document (whichever is the shorter) is available
2321 for matching (and similarly the previous 8K for lookbehind assertions).
2322
2323 (f) Changed the --help output so that it now says
2324
2325 -w, --word-regex(p)
2326
2327 instead of two lines, one with "regex" and the other with "regexp"
2328 because that confused at least one person since the short forms are the
2329 same. (This required a bit of code, as the output is generated
2330 automatically from a table. It wasn't just a text change.)
2331
2332 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2333 option but starts with a hyphen. Could be a pattern or a path name
2334 starting with a hyphen, for instance.
2335
2336 (h) "-" can be given as a file name to represent stdin.
2337
2338 (i) When file names are being printed, "(standard input)" is used for
2339 the standard input, for compatibility with GNU grep. Previously
2340 "<stdin>" was used.
2341
2342 (j) The option --label=xxx can be used to supply a name to be used for
2343 stdin when file names are being printed. There is no short form.
2344
2345 (k) Re-factored the options decoding logic because we are going to add
2346 two more options that take data. Such options can now be given in four
2347 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2348
2349 (l) Added the -A, -B, and -C options for requesting that lines of context
2350 around matches be printed.
2351
2352 (m) Added the -L option to print the names of files that do not contain
2353 any matching lines, that is, the complement of -l.
2354
2355 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2356 continue to scan other files.
2357
2358 (o) The -s option was incorrectly implemented. For compatibility with other
2359 greps, it now suppresses the error message for a non-existent or non-
2360 accessible file (but not the return code). There is a new option called
2361 -q that suppresses the output of matching lines, which was what -s was
2362 previously doing.
2363
2364 (p) Added --include and --exclude options to specify files for inclusion
2365 and exclusion when recursing.
2366
2367 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2368 Hopefully, it now does.
2369
2370 12. Missing cast in pcre_study().
2371
2372 13. Added an "uninstall" target to the makefile.
2373
2374 14. Replaced "extern" in the function prototypes in Makefile.in with
2375 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2376 world, but is set differently for Windows.
2377
2378 15. Added a second compiling function called pcre_compile2(). The only
2379 difference is that it has an extra argument, which is a pointer to an
2380 integer error code. When there is a compile-time failure, this is set
2381 non-zero, in addition to the error test pointer being set to point to an
2382 error message. The new argument may be NULL if no error number is required
2383 (but then you may as well call pcre_compile(), which is now just a
2384 wrapper). This facility is provided because some applications need a
2385 numeric error indication, but it has also enabled me to tidy up the way
2386 compile-time errors are handled in the POSIX wrapper.
2387
2388 16. Added VPATH=.libs to the makefile; this should help when building with one
2389 prefix path and installing with another. (Or so I'm told by someone who
2390 knows more about this stuff than I do.)
2391
2392 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2393 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2394 match everything, including newlines. This is not POSIX-compatible, but
2395 somebody wanted the feature. From pcretest it can be activated by using
2396 both the P and the s flags.
2397
2398 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2399
2400 19. libpcre.pc was being incorrectly installed as executable.
2401
2402 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2403 it now also looks for '\r' so that it will work unmodified on Windows.
2404
2405 21. Added Google's contributed C++ wrapper to the distribution.
2406
2407 22. Added some untidy missing memory free() calls in pcretest, to keep
2408 Electric Fence happy when testing.
2409
2410
2411
2412 Version 5.0 13-Sep-04
2413 ---------------------
2414
2415 1. Internal change: literal characters are no longer packed up into items
2416 containing multiple characters in a single byte-string. Each character
2417 is now matched using a separate opcode. However, there may be more than one
2418 byte in the character in UTF-8 mode.
2419
2420 2. The pcre_callout_block structure has two new fields: pattern_position and
2421 next_item_length. These contain the offset in the pattern to the next match
2422 item, and its length, respectively.
2423
2424 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2425 insertion of callouts before each pattern item. Added the /C option to
2426 pcretest to make use of this.
2427
2428 4. On the advice of a Windows user, the lines
2429
2430 #if defined(_WIN32) || defined(WIN32)
2431 _setmode( _fileno( stdout ), 0x8000 );
2432 #endif /* defined(_WIN32) || defined(WIN32) */
2433
2434 have been added to the source of pcretest. This apparently does useful
2435 magic in relation to line terminators.
2436
2437 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2438 for the benefit of those environments where the "b" makes a difference.
2439
2440 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2441 to know about it. I have put a hack into configure.in that adds in code
2442 to set GCC=yes if CC=icc. This seems to end up at a point in the
2443 generated configure script that is early enough to affect the setting of
2444 compiler options, which is what is needed, but I have no means of testing
2445 whether it really works. (The user who reported this had patched the
2446 generated configure script, which of course I cannot do.)
2447
2448 LATER: After change 22 below (new libtool files), the configure script
2449 seems to know about icc (and also ecc). Therefore, I have commented out
2450 this hack in configure.in.
2451
2452 7. Added support for pkg-config (2 patches were sent in).
2453
2454 8. Negated POSIX character classes that used a combination of internal tables
2455 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2456 [[:^ascii]]. Typically, they would match almost any characters. The other
2457 POSIX classes were not broken in this way.
2458
2459 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2460 to find the match, as PCRE was deluded into thinking that the match had to
2461 start at the start point or following a newline. The same bug applied to
2462 patterns with negative forward assertions or any backward assertions
2463 preceding ".*" at the start, unless the pattern required a fixed first
2464 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2465
2466 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2467 starting at the last subject character, bytes beyond the end of the subject
2468 string were read.
2469
2470 11. Renamed the variable "class" as "classbits" to make life easier for C++
2471 users. (Previously there was a macro definition, but it apparently wasn't
2472 enough.)
2473
2474 12. Added the new field "tables" to the extra data so that tables can be passed
2475 in at exec time, or the internal tables can be re-selected. This allows
2476 a compiled regex to be saved and re-used at a later time by a different
2477 program that might have everything at different addresses.
2478
2479 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2480 -R library as well as a -L library.
2481
2482 14. The debugging options of pcretest (-d on the command line or D on a
2483 pattern) showed incorrect output for anything following an extended class
2484 that contained multibyte characters and which was followed by a quantifier.
2485
2486 15. Added optional support for general category Unicode character properties
2487 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2488 support. It adds about 90K to the size of the library. The meanings of the
2489 inbuilt class escapes such as \d and \s have NOT been changed.
2490
2491 16. Updated pcredemo.c to include calls to free() to release the memory for the
2492 compiled pattern.
2493
2494 17. The generated file chartables.c was being created in the source directory
2495 instead of in the building directory. This caused the build to fail if the
2496 source directory was different from the building directory, and was
2497 read-only.
2498
2499 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2500 file. No doubt somebody will tell me if they don't make sense... Also added
2501 Dan Mooney's comments about building on OpenVMS.
2502
2503 19. Added support for partial matching via the PCRE_PARTIAL option for
2504 pcre_exec() and the \P data escape in pcretest.
2505
2506 20. Extended pcretest with 3 new pattern features:
2507
2508 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2509 write the compiled pattern to the file whose name is "rest-of-line".
2510 This is a straight binary dump of the data, with the saved pointer to
2511 the character tables forced to be NULL. The study data, if any, is
2512 written too. After writing, pcretest reads a new pattern.
2513
2514 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2515 compiled pattern from the given file. There must not be any
2516 occurrences of "<" in the file name (pretty unlikely); if there are,
2517 pcretest will instead treat the initial "<" as a pattern delimiter.
2518 After reading in the pattern, pcretest goes on to read data lines as
2519 usual.
2520
2521 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2522 and 16-bit fields in a compiled pattern, to simulate a pattern that
2523 was compiled on a host of opposite endianness.
2524
2525 21. The pcre-exec() function can now cope with patterns that were compiled on
2526 hosts of opposite endianness, with this restriction:
2527
2528 As for any compiled expression that is saved and used later, the tables
2529 pointer field cannot be preserved; the extra_data field in the arguments
2530 to pcre_exec() should be used to pass in a tables address if a value
2531 other than the default internal tables were used at compile time.
2532
2533 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2534 now diagnosed as an error. Previously, most of the time, a negative number
2535 would have been treated as zero, but if in addition "ovector" was passed as
2536 NULL, a crash could occur.
2537
2538 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2539 new versions from the libtool 1.5 distribution (the last one is a copy of
2540 a file called libtool.m4). This seems to have fixed the need to patch
2541 "configure" to support Darwin 1.3 (which I used to do). However, I still
2542 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2543 workstation).
2544
2545 24. Changed the PCRE licence to be the more standard "BSD" licence.
2546
2547
2548 Version 4.5 01-Dec-03
2549 ---------------------
2550
2551 1. There has been some re-arrangement of the code for the match() function so
2552 that it can be compiled in a version that does not call itself recursively.
2553 Instead, it keeps those local variables that need separate instances for
2554 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2555 needs to "recurse". Keeping track of where control must go is done by means
2556 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2557 hide most of the details from the main code, and operates only if
2558 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2559 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2560 operating.
2561
2562 To make it easier for callers to provide specially tailored get/free
2563 functions for this usage, two new functions, pcre_stack_malloc, and
2564 pcre_stack_free, are used. They are always called in strict stacking order,
2565 and the size of block requested is always the same.
2566
2567 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2568 PCRE has been compiled to use the stack or the heap for recursion. The
2569 -C option of pcretest uses this to show which version is compiled.
2570
2571 A new data escape \S, is added to pcretest; it causes the amounts of store
2572 obtained and freed by both kinds of malloc/free at match time to be added
2573 to the output.
2574
2575 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2576 what's available on my current Linux desktop machine.
2577
2578 3. When matching a UTF-8 string, the test for a valid string at the start has
2579 been extended. If start_offset is not zero, PCRE now checks that it points
2580 to a byte that is the start of a UTF-8 character. If not, it returns
2581 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2582 this is necessary because there may be backward assertions in the pattern.
2583 When matching the same subject several times, it may save resources to use
2584 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2585
2586 4. The code for checking the validity of UTF-8 strings has been tightened so
2587 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2588 containing "overlong sequences".
2589
2590 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2591 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2592 should have been "&", but it just so happened that all the cases this let
2593 through by mistake were picked up later in the function.
2594
2595 6. I had used a variable called "isblank" - this is a C99 function, causing
2596 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2597
2598 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2599 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2600 so the version is shown; (c) stop "make test" echoing "./RunTest".
2601
2602 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2603
2604 9. The prototype for memmove() for systems that don't have it was using
2605 size_t, but the inclusion of the header that defines size_t was later. I've
2606 moved the #includes for the C headers earlier to avoid this.
2607
2608 10. Added some adjustments to the code to make it easier to compiler on certain
2609 special systems:
2610
2611 (a) Some "const" qualifiers were missing.
2612 (b) Added the macro EXPORT before all exported functions; by default this
2613 is defined to be empty.
2614 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2615 that it reads its output file name as an argument instead of writing
2616 to the standard output and assuming this can be redirected.
2617
2618 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2619 class containing characters with values greater than 255, PCRE compilation
2620 went into a loop.
2621
2622 12. A recursive reference to a subpattern that was within another subpattern
2623 that had a minimum quantifier of zero caused PCRE to crash. For example,
2624 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2625 recursion. If the recursively-called subpattern itself had a zero repeat,
2626 that was OK.
2627
2628 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2629 buffer into which it was copied (for escape processing) was still set at
2630 1024, so long lines caused crashes.
2631
2632 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2633 "internal error: code overflow...". This applied to any character class
2634 that was followed by a possessive quantifier.
2635
2636 15. Modified the Makefile to add libpcre.la as a prerequisite for
2637 libpcreposix.la because I was told this is needed for a parallel build to
2638 work.
2639
2640 16. If a pattern that contained .* following optional items at the start was
2641 studied, the wrong optimizing data was generated, leading to matching
2642 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2643 matching string must start with a or b or c. The correct conclusion for
2644 this pattern is that a match can start with any character.
2645
2646
2647 Version 4.4 13-Aug-03
2648 ---------------------
2649
2650 1. In UTF-8 mode, a character class containing characters with values between
2651 127 and 255 was not handled correctly if the compiled pattern was studied.
2652 In fixing this, I have also improved the studying algorithm for such
2653 classes (slightly).
2654
2655 2. Three internal functions had redundant arguments passed to them. Removal
2656 might give a very teeny performance improvement.
2657
2658 3. Documentation bug: the value of the capture_top field in a callout is *one
2659 more than* the number of the hightest numbered captured substring.
2660
2661 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2662 in incorrectly linking with a previously installed version. They now link
2663 explicitly with libpcre.la.
2664
2665 5. configure.in no longer needs to recognize Cygwin specially.
2666
2667 6. A problem in pcre.in for Windows platforms is fixed.
2668
2669 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2670 pcretest, it used to include the size of the study block as part of its
2671 output. Unfortunately, the structure contains a field that has a different
2672 size on different hardware architectures. This meant that the tests that
2673 showed this size failed. As the block is currently always of a fixed size,
2674 this information isn't actually particularly useful in pcretest output, so
2675 I have just removed it.
2676
2677 8. Three pre-processor statements accidentally did not start in column 1.
2678 Sadly, there are *still* compilers around that complain, even though
2679 standard C has not required this for well over a decade. Sigh.
2680
2681 9. In pcretest, the code for checking callouts passed small integers in the
2682 callout_data field, which is a void * field. However, some picky compilers
2683 complained about the casts involved for this on 64-bit systems. Now
2684 pcretest passes the address of the small integer instead, which should get
2685 rid of the warnings.
2686
2687 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2688 both compile and run time, and gives an error if an invalid UTF-8 sequence
2689 is found. There is a option for disabling this check in cases where the
2690 string is known to be correct and/or the maximum performance is wanted.
2691
2692 11. In response to a bug report, I changed one line in Makefile.in from
2693
2694 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2695 to
2696 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2697
2698 to look similar to other lines, but I have no way of telling whether this
2699 is the right thing to do, as I do not use Windows. No doubt I'll get told
2700 if it's wrong...
2701
2702
2703 Version 4.3 21-May-03
2704 ---------------------
2705
2706 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2707 Makefile.
2708
2709 2. Some refactoring to improve the quality of the code:
2710
2711 (i) The utf8_table... variables are now declared "const".
2712
2713 (ii) The code for \cx, which used the "case flipping" table to upper case
2714 lower case letters, now just substracts 32. This is ASCII-specific,
2715 but the whole concept of \cx is ASCII-specific, so it seems
2716 reasonable.
2717
2718 (iii) PCRE was using its character types table to recognize decimal and
2719 hexadecimal digits in the pattern. This is silly, because it handles
2720 only 0-9, a-f, and A-F, but the character types table is locale-
2721 specific, which means strange things might happen. A private
2722 table is now used for this - though it costs 256 bytes, a table is
2723 much faster than multiple explicit tests. Of course, the standard
2724 character types table is still used for matching digits in subject
2725 strings against \d.
2726
2727 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2728 ending in _t are). So I've renamed it as ESC_tee.
2729
2730 3. The first argument for regexec() in the POSIX wrapper should have been
2731 defined as "const".
2732
2733 4. Changed pcretest to use malloc() for its buffers so that they can be
2734 Electric Fenced for debugging.
2735
2736 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2737 to read one or more bytes before the start of the subject string. Often this
2738 had no effect on PCRE's behaviour, but in some circumstances it could
2739 provoke a segmentation fault.
2740
2741 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2742 to try to read one or more bytes before the start of the subject string.
2743
2744 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2745 UTF-8 support could misbehave in various ways if the subject string
2746 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2747 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2748 back over UTF-8 characters.)
2749
2750
2751 Version 4.2 14-Apr-03
2752 ---------------------
2753
2754 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2755
2756 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2757 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2758 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2759 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2760 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2761 and BUILD_EXEEXT
2762 Note: automatic setting of the BUILD variables is not yet working
2763 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2764 compile-time but not at link-time
2765 [LINK]: use for linking executables only
2766 make different versions for Windows and non-Windows
2767 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2768 libraries
2769 [LINK_FOR_BUILD]: new variable
2770 [OBJEXT]: use throughout
2771 [EXEEXT]: use throughout
2772 <winshared>: new target
2773 <wininstall>: new target
2774 <dftables.o>: use native compiler
2775 <dftables>: use native linker
2776 <install>: handle Windows platform correctly
2777 <clean>: ditto
2778 <check>: ditto
2779 copy DLL to top builddir before testing
2780
2781 As part of these changes, -no-undefined was removed again. This was reported
2782 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2783 in any case.
2784
2785 3. Some tidies to get rid of compiler warnings:
2786
2787 . In the match_data structure, match_limit was an unsigned long int, whereas
2788 match_call_count was an int. I've made them both unsigned long ints.
2789
2790 . In pcretest the fact that a const uschar * doesn't automatically cast to
2791 a void * provoked a warning.
2792
2793 . Turning on some more compiler warnings threw up some "shadow" variables
2794 and a few more missing casts.
2795
2796 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2797 option, a class that contained a single character with a value between 128
2798 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2799
2800 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2801 option, a class that contained several characters, but with at least one
2802 whose value was between 128 and 255 caused PCRE to crash.
2803
2804
2805 Version 4.1 12-Mar-03
2806 ---------------------
2807
2808 1. Compiling with gcc -pedantic found a couple of places where casts were
2809 needed, and a string in dftables.c that was longer than standard compilers are
2810 required to support.
2811
2812 2. Compiling with Sun's compiler found a few more places where the code could
2813 be tidied up in order to avoid warnings.
2814
2815 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2816 first of these names is deprecated in the latest Autoconf in favour of the name
2817 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2818 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2819 analogy I have changed it to CFLAGS_FOR_BUILD.
2820
2821 4. Added -no-undefined to the linking command in the Makefile, because this is
2822 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2823 linking step for the pcreposix library.
2824
2825 5. PCRE was failing to diagnose the case of two named groups with the same
2826 name.
2827
2828 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2829 literal character that is needed in the subject for a match, and scans along to
2830 ensure that it is present before embarking on the full matching process. This
2831 saves time in cases of nested unlimited repeats that are never going to match.
2832 Problem: the scan can take a lot of time if the subject is very long (e.g.
2833 megabytes), thus penalizing straightforward matches. It is now done only if the
2834 amount of subject to be scanned is less than 1000 bytes.
2835
2836 7. A lesser problem with the same optimization is that it was recording the
2837 first character of an anchored pattern as "needed", thus provoking a search
2838 right along the subject, even when the first match of the pattern was going to
2839 fail. The "needed" character is now not set for anchored patterns, unless it
2840 follows something in the pattern that is of non-fixed length. Thus, it still
2841 fulfils its original purpose of finding quick non-matches in cases of nested
2842 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2843
2844
2845 Version 4.0 17-Feb-03
2846 ---------------------
2847
2848 1. If a comment in an extended regex that started immediately after a meta-item
2849 extended to the end of string, PCRE compiled incorrect data. This could lead to
2850 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2851
2852 2. Moved to autoconf 2.53 and libtool 1.4.2.
2853
2854 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2855 the special perltest8 script is no longer needed - all the tests can be run
2856 from a single perltest script.
2857
2858 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2859 by \s. It has now been removed in PCRE. This means it isn't recognized as
2860 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2861 class [:space:] *does* include VT, thereby creating a mess.
2862
2863 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2864 space and tab.
2865
2866 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2867 its new features into the main test script, reducing the number of scripts.
2868
2869 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2870 were backward compatible, and made the (?i) apply to the whole pattern, as if
2871 /i were given. Now it behaves more logically, and applies the option setting
2872 only to what follows. PCRE has been changed to follow suit. However, if it
2873 finds options settings right at the start of the pattern, it extracts them into
2874 the global options, as before. Thus, they show up in the info data.
2875
2876 8. Added support for the \Q...\E escape sequence. Characters in between are
2877 treated as literals. This is slightly different from Perl in that $ and @ are
2878 also handled as literals inside the quotes. In Perl, they will cause variable
2879 interpolation. Note the following examples:
2880
2881 Pattern PCRE matches Perl matches
2882
2883 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2884 \Qabc\$xyz\E abc\$xyz abc\$xyz
2885 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2886
2887 For compatibility with Perl, \Q...\E sequences are recognized inside character
2888 classes as well as outside them.
2889
2890 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2891 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2892 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2893 signed/unsigned warnings.
2894
2895 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2896 option for pcretest, so I've replaced it by a simple function that does just
2897 that job.
2898
2899 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2900 "pcregrep -".
2901
2902 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2903 Java package. This provides some syntactic sugar for simple cases of what my
2904 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2905 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2906 item, you can use this simplified notation. Note that only makes sense with
2907 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2908 greediness, whatever the setting of the PCRE_UNGREEDY option.
2909
2910 13. A change of greediness default within a pattern was not taking effect at
2911 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2912 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2913 was abstracted outside.
2914
2915 14. PCRE now supports the \G assertion. It is true when the current matching
2916 position is at the start point of the match. This differs from \A when the
2917 starting offset is non-zero. Used with the /g option of pcretest (or similar
2918 code), it works in the same way as it does for Perl's /g option. If all
2919 alternatives of a regex begin with \G, the expression is anchored to the start
2920 match position, and the "anchored" flag is set in the compiled expression.
2921
2922 15. Some bugs concerning the handling of certain option changes within patterns
2923 have been fixed. These applied to options other than (?ims). For example,
2924 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2925 been the other way round. Some of this was related to change 7 above.
2926
2927 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2928 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2929 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2930 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2931
2932 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2933 mode. Unlike ".", it always matches newline, whatever the setting of
2934 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2935 assertions. Perl allows it, but it doesn't (in general) work because it can't
2936 calculate the length of the lookbehind. At least, that's the case for Perl
2937 5.8.0 - I've been told they are going to document that it doesn't work in
2938 future.
2939
2940 18. Added an error diagnosis for escapes that PCRE does not support: these are
2941 \L, \l, \N, \P, \p, \U, \u, and \X.
2942
2943 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2944 reading past the end of the pattern in cases such as /[abcd/.
2945
2946 20. PCRE was getting more memory than necessary for patterns with classes that
2947 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2948
2949 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2950 compiling PCRE for use with Virtual Pascal.
2951
2952 22. Small fix to the Makefile to make it work properly if the build is done
2953 outside the source tree.
2954
2955 23. Added a new extension: a condition to go with recursion. If a conditional
2956 subpattern starts with (?(R) the "true" branch is used if recursion has
2957 happened, whereas the "false" branch is used only at the top level.
2958
2959 24. When there was a very long string of literal characters (over 255 bytes
2960 without UTF support, over 250 bytes with UTF support), the computation of how
2961 much memory was required could be incorrect, leading to segfaults or other
2962 strange effects.
2963
2964 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2965 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2966 there was a subsequent back reference to those brackets. This meant that, for
2967 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2968 possible to check for precisely this case. All we can do is abandon the
2969 optimization if .* occurs inside capturing brackets when there are any back
2970 references whatsoever. (See below for a better fix that came later.)
2971
2972 26. The handling of the optimization for finding the first character of a
2973 non-anchored pattern, and for finding a character that is required later in the
2974 match were failing in some cases. This didn't break the matching; it just
2975 failed to optimize when it could. The way this is done has been re-implemented.
2976
2977 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2978
2979 28. Added a new feature that provides some of the functionality that Perl
2980 provides with (?{...}). The facility is termed a "callout". The way it is done
2981 in PCRE is for the caller to provide an optional function, by setting
2982 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2983 global variable. By default it is unset, which disables all calling out. To get
2984 the function called, the regex must include (?C) at appropriate points. This
2985 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2986 This provides a means of identifying different callout points. When PCRE
2987 reaches such a point in the regex, if pcre_callout has been set, the external
2988 function is called. It is provided with data in a structure called
2989 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2990 matching continues; if it returns a non-zero value, the match at the current
2991 point fails. However, backtracking will occur if possible. [This was changed
2992 later and other features added - see item 49 below.]
2993
2994 29. pcretest is upgraded to test the callout functionality. It provides a
2995 callout function that displays information. By default, it shows the start of
2996 the match and the current position in the text. There are some new data escapes
2997 to vary what happens:
2998
2999 \C+ in addition, show current contents of captured substrings
3000 \C- do not supply a callout function
3001 \C!n return 1 when callout number n is reached
3002 \C!n!m return 1 when callout number n is reached for the mth time
3003
3004 30. If pcregrep was called with the -l option and just a single file name, it
3005 output "<stdin>" if a match was found, instead of the file name.
3006
3007 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3008 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3009 pcre_exec(). This saves a malloc/free per call. The default value of
3010 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3011 when configuring.
3012
3013 32. The default maximum size of a compiled pattern is 64K. There have been a
3014 few cases of people hitting this limit. The code now uses macros to handle the
3015 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3016 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3017 configuring. Tests 2 and 5 work only with 2-byte links because they output
3018 debugging information about compiled patterns.
3019
3020 33. Internal code re-arrangements:
3021
3022 (a) Moved the debugging function for printing out a compiled regex into
3023 its own source file (printint.c) and used #include to pull it into
3024 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3025 separate copies.
3026
3027 (b) Defined the list of op-code names for debugging as a macro in
3028 internal.h so that it is next to the definition of the opcodes.
3029
3030 (c) Defined a table of op-code lengths for simpler skipping along compiled
3031 code. This is again a macro in internal.h so that it is next to the
3032 definition of the opcodes.
3033
3034 34. Added support for recursive calls to individual subpatterns, along the
3035 lines of Robin Houston's patch (but implemented somewhat differently).
3036
3037 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3038 allow it to read and process whole directories in Win32. This code was
3039 contributed by Lionel Fourquaux; it has not been tested by me.
3040
3041 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3042 used to name a group. Names consist of alphanumerics and underscores, and must
3043 be unique. Back references use the syntax (?P=name) and recursive calls use
3044 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3045 numbers. The function pcre_fullinfo() can be used after compilation to extract
3046 a name/number map. There are three relevant calls:
3047
3048 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3049 PCRE_INFO_NAMECOUNT yields the number of entries
3050 PCRE_INFO_NAMETABLE yields a pointer to the map.
3051
3052 The map is a vector of fixed-size entries. The size of each entry depends on
3053 the length of the longest name used. The first two bytes of each entry are the
3054 group number, most significant byte first. There follows the corresponding
3055 name, zero terminated. The names are in alphabetical order.
3056
3057 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3058 case instead of 255. Making it the same both with and without UTF-8 support
3059 means that the same test output works with both.
3060
3061 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3062 calling malloc() with a zero argument.
3063
3064 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3065 optimization. I've improved things by keeping a bitmap of backreferences with
3066 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3067 fact referenced, the optimization can be applied. It is unlikely that a
3068 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3069 the match to follow \n) will appear inside brackets with a number greater than
3070 31, but if it does, any back reference > 31 suppresses the optimization.
3071
3072 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3073 of disabling numbered capturing parentheses. Any opening parenthesis that is
3074 not followed by ? behaves as if it were followed by ?: but named parentheses
3075 can still be used for capturing (and they will acquire numbers in the usual
3076 way).
3077
3078 41. Redesigned the return codes from the match() function into yes/no/error so
3079 that errors can be passed back from deep inside the nested calls. A malloc
3080 failure while inside a recursive subpattern call now causes the
3081 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3082
3083 42. It is now possible to set a limit on the number of times the match()
3084 function is called in a call to pcre_exec(). This facility makes it possible to
3085 limit the amount of recursion and backtracking, though not in a directly
3086 obvious way, because the match() function is used in a number of different
3087 circumstances. The count starts from zero for each position in the subject
3088 string (for non-anchored patterns). The default limit is, for compatibility, a
3089 large number, namely 10 000 000. You can change this in two ways:
3090
3091 (a) When configuring PCRE before making, you can use --with-match-limit=n
3092 to set a default value for the compiled library.
3093
3094 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3095 a different value is set. See 45 below.
3096
3097 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3098
3099 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3100 of things that can be changed at compile time. The first argument specifies
3101 what is wanted and the second points to where the information is to be placed.
3102 The current list of available information is:
3103
3104 PCRE_CONFIG_UTF8
3105
3106 The output is an integer that is set to one if UTF-8 support is available;
3107 otherwise it is set to zero.
3108
3109 PCRE_CONFIG_NEWLINE
3110
3111 The output is an integer that it set to the value of the code that is used for
3112 newline. It is either LF (10) or CR (13).
3113
3114 PCRE_CONFIG_LINK_SIZE
3115
3116 The output is an integer that contains the number of bytes used for internal
3117 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3118
3119 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3120
3121 The output is an integer that contains the threshold above which the POSIX
3122 interface uses malloc() for output vectors. See item 31 above.
3123
3124 PCRE_CONFIG_MATCH_LIMIT
3125
3126 The output is an unsigned integer that contains the default limit of the number
3127 of match() calls in a pcre_exec() execution. See 42 above.
3128
3129 44. pcretest has been upgraded by the addition of the -C option. This causes it
3130 to extract all the available output from the new pcre_config() function, and to
3131 output it. The program then exits immediately.
3132
3133 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3134 order to support additional features. One way would have been to define
3135 pcre_exec2() (for example) with extra arguments, but this would not have been
3136 extensible, and would also have required all calls to the original function to
3137 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3138 is used for passing in "extra" data from pcre_study().
3139
3140 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3141 contains the following fields:
3142
3143 flags a bitmap indicating which of the following fields are set
3144 study_data opaque data from pcre_study()
3145 match_limit a way of specifying a limit on match() calls for a specific
3146 call to pcre_exec()
3147 callout_data data for callouts (see 49 below)
3148
3149 The flag bits are also defined in pcre.h, and are
3150
3151 PCRE_EXTRA_STUDY_DATA
3152 PCRE_EXTRA_MATCH_LIMIT
3153 PCRE_EXTRA_CALLOUT_DATA
3154
3155 The pcre_study() function now returns one of these new pcre_extra blocks, with
3156 the actual study data pointed to by the study_data field, and the
3157 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3158 before. That is, this change is entirely upwards-compatible and requires no
3159 change to existing code.
3160
3161 If you want to pass in additional data to pcre_exec(), you can either place it
3162 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3163 block.
3164
3165 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3166 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3167 times with different match limits, until it finds the minimum value needed for
3168 pcre_exec() to complete. The value is then output. This can be instructive; for
3169 most simple matches the number is quite small, but for pathological cases it
3170 gets very large very quickly.
3171
3172 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3173 returns the size of the data block pointed to by the study_data field in a
3174 pcre_extra block, that is, the value that was passed as the argument to
3175 pcre_malloc() when PCRE was getting memory in which to place the information
3176 created by pcre_study(). The fourth argument should point to a size_t variable.
3177 pcretest has been extended so that this information is shown after a successful
3178 pcre_study() call when information about the compiled regex is being displayed.
3179
3180 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3181 because what follows is always an absolute path. (Later: it turns out that this
3182 is more than cosmetic for MinGW, because it doesn't like empty path
3183 components.)
3184
3185 49. Some changes have been made to the callout feature (see 28 above):
3186
3187 (i) A callout function now has three choices for what it returns:
3188
3189 0 => success, carry on matching
3190 > 0 => failure at this point, but backtrack if possible
3191 < 0 => serious error, return this value from pcre_exec()
3192
3193 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3194 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3195 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3196 use by callout functions. It will never be used by PCRE itself.
3197
3198 (ii) The pcre_extra structure (see 45 above) has a void * field called
3199 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3200 pcre_callout_block structure has a field of the same name. The contents of
3201 the field passed in the pcre_extra structure are passed to the callout
3202 function in the corresponding field in the callout block. This makes it
3203 easier to use the same callout-containing regex from multiple threads. For
3204 testing, the pcretest program has a new data escape
3205
3206 \C*n pass the number n (may be negative) as callout_data
3207
3208 If the callout function in pcretest receives a non-zero value as
3209 callout_data, it returns that value.
3210
3211 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3212 there were some redundant $(CFLAGS) in commands that are now specified as
3213 $(LINK), which already includes $(CFLAGS).
3214
3215 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3216 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3217 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3218 one-byte characters throughout. Note that case-insensitive matching applies
3219 only to characters whose values are less than 256. PCRE doesn't support the
3220 notion of cases for higher-valued characters.
3221
3222 (i) A character class whose characters are all within 0-255 is handled as
3223 a bit map, and the map is inverted for negative classes. Previously, a
3224 character > 255 always failed to match such a class; however it should
3225 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3226
3227 (ii) A negated character class with a single character < 255 is coded as
3228 "not this character" (OP_NOT). This wasn't working properly when the test
3229 character was multibyte, either singly or repeated.
3230
3231 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3232 mode, for example: \x{100}{2,3}.
3233
3234 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3235 singly or repeated) now correctly test multibyte characters. However,
3236 PCRE doesn't recognize any characters with values greater than 255 as
3237 digits, spaces, or word characters. Such characters always match \D, \S,
3238 and \W, and never match \d, \s, or \w.
3239
3240 (v) Classes may now contain characters and character ranges with values
3241 greater than 255. For example: [ab\x{100}-\x{400}].
3242
3243 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3244 PCRE in UTF-8 mode.
3245
3246 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3247 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3248 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3249 value.)
3250
3251 53. The single man page has become too large. I have therefore split it up into
3252 a number of separate man pages. These also give rise to individual HTML pages;
3253 these are now put in a separate directory, and there is an index.html page that
3254 lists them all. Some hyperlinking between the pages has been installed.
3255
3256 54. Added convenience functions for handling named capturing parentheses.
3257
3258 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3259 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3260 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3261 are faulted.
3262
3263 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3264 calling configure. These values are used when compiling the dftables.c program
3265 which is run to generate the source of the default character tables. They
3266 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3267 you will need to set these values.
3268
3269 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3270
3271
3272 Version 3.9 02-Jan-02
3273 ---------------------
3274
3275 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3276
3277 2. If --disable-static was given, the building process failed when trying to
3278 build pcretest and pcregrep. (For some reason it was using libtool to compile
3279 them, which is not right, as they aren't part of the library.)
3280
3281
3282 Version 3.8 18-Dec-01
3283 ---------------------
3284
3285 1. The experimental UTF-8 code was completely screwed up. It was packing the
3286 bytes in the wrong order. How dumb can you get?
3287
3288
3289 Version 3.7 29-Oct-01
3290 ---------------------
3291
3292 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3293 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3294 this didn't happen under Solaris 8, where I normally test things.
3295
3296 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3297 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3298 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3299 this fix an hour or so after the initial 3.7 release.)
3300
3301
3302 Version 3.6 23-Oct-01
3303 ---------------------
3304
3305 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3306 offsets passed as NULL with zero offset count.
3307
3308 2. The config.guess and config.sub files had not been updated when I moved to
3309 the latest autoconf.
3310
3311
3312 Version 3.5 15-Aug-01
3313 ---------------------
3314
3315 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3316 had been forgotten.
3317
3318 2. By using declared but undefined structures, we can avoid using "void"
3319 definitions in pcre.h while keeping the internal definitions of the structures
3320 private.
3321
3322 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3323 user point of view, this means that both static and shared libraries are built
3324 by default, but this can be individually controlled. More of the work of
3325 handling this static/shared cases is now inside libtool instead of PCRE's make
3326 file.
3327
3328 4. The pcretest utility is now installed along with pcregrep because it is
3329 useful for users (to test regexs) and by doing this, it automatically gets
3330 relinked by libtool. The documentation has been turned into a man page, so
3331 there are now .1, .txt, and .html versions in /doc.
3332
3333 5. Upgrades to pcregrep:
3334 (i) Added long-form option names like gnu grep.
3335 (ii) Added --help to list all options with an explanatory phrase.
3336 (iii) Added -r, --recursive to recurse into sub-directories.
3337 (iv) Added -f, --file to read patterns from a file.
3338
3339 6. pcre_exec() was referring to its "code" argument before testing that
3340 argument for NULL (and giving an error if it was NULL).
3341
3342 7. Upgraded Makefile.in to allow for compiling in a different directory from
3343 the source directory.
3344
3345 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3346 options bits, the pointer it was passed was to an int instead of to an unsigned
3347 long int. This mattered only on 64-bit systems.
3348
3349 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3350 generated) instead of pcre.in, which it its source. Also made the same change
3351 in several of the .c files.
3352
3353 10. A new release of gcc defines printf() as a macro, which broke pcretest
3354 because it had an ifdef in the middle of a string argument for printf(). Fixed
3355 by using separate calls to printf().
3356
3357 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3358 script, to force use of CR or LF instead of \n in the source. On non-Unix
3359 systems, the value can be set in config.h.
3360
3361 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3362 absolute limit. Changed the text of the error message to make this clear, and
3363 likewise updated the man page.
3364
3365 13. The limit of 99 on the number of capturing subpatterns has been removed.
3366 The new limit is 65535, which I hope will not be a "real" limit.
3367
3368
3369 Version 3.4 22-Aug-00
3370 ---------------------
3371
3372 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3373
3374 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3375
3376
3377 Version 3.3 01-Aug-00
3378 ---------------------
3379
3380 1. If an octal character was given, but the value was greater than \377, it
3381 was not getting masked to the least significant bits, as documented. This could
3382 lead to crashes in some systems.
3383
3384 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3385 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3386
3387 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3388 These just pass their arguments on to (pcre_free)(), but they are provided
3389 because some uses of PCRE bind it to non-C systems that can call its functions,
3390 but cannot call free() or pcre_free() directly.
3391
3392 4. Add "make test" as a synonym for "make check". Corrected some comments in
3393 the Makefile.
3394
3395 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3396 Makefile.
3397
3398 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3399 command called pgrep for grepping around the active processes.
3400
3401 7. Added the beginnings of support for UTF-8 character strings.
3402
3403 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3404 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3405 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3406 out for the ar command.)
3407
3408
3409 Version 3.2 12-May-00
3410 ---------------------
3411
3412 This is purely a bug fixing release.
3413
3414 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3415 of ZA. This was just one example of several cases that could provoke this bug,
3416 which was introduced by change 9 of version 2.00. The code for breaking
3417 infinite loops after an iteration that matches an empty string was't working
3418 correctly.
3419
3420 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3421 when matched against abbab (for example). After matching an empty string, it
3422 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3423 caused it to match further down the string than it should.
3424
3425 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3426 was there because it doesn't seem to be needed, and it causes trouble on some
3427 systems, as it is not a Standard C header. It has been removed.
3428
3429 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3430 were reported on the Macintosh. The changes were from
3431
3432 while ((c = *(++ptr)) != 0 && c != '\n');
3433 to
3434 while ((c = *(++ptr)) != 0 && c != '\n') ;
3435
3436 Totally extraordinary, but if that's what it takes...
3437
3438 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3439 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3440 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3441 assumes the way PCRE uses memmove() (always moving upwards).
3442
3443 6. PCRE is being used in one environment where strchr() is not available. There
3444 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3445 faster code anyway.
3446
3447
3448 Version 3.1 09-Feb-00
3449 ---------------------
3450
3451 The only change in this release is the fixing of some bugs in Makefile.in for
3452 the "install" target:
3453
3454 (1) It was failing to install pcreposix.h.
3455
3456 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3457
3458
3459 Version 3.0 01-Feb-00
3460 ---------------------
3461
3462 1. Add support for the /+ modifier to perltest (to output $` like it does in
3463 pcretest).
3464
3465 2. Add support for the /g modifier to perltest.
3466
3467 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3468 matches null strings.
3469
3470 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3471 pattern. Perl treats empty patterns specially - it reuses the most recent
3472 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3473 effect.
3474
3475 5. The POSIX interface was broken in that it was just handing over the POSIX
3476 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3477 required a bigger vector, with some working space on the end. This means that
3478 the POSIX wrapper now has to get and free some memory, and copy the results.
3479
3480 6. Added some simple autoconf support, placing the test data and the
3481 documentation in separate directories, re-organizing some of the
3482 information files, and making it build pcre-config (a GNU standard). Also added
3483 libtool support for building PCRE as a shared library, which is now the
3484 default.
3485
3486 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3487 09 are not valid octal constants. Single digits will be used for minor values
3488 less than 10.
3489
3490 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3491 existing programs that set these in the POSIX interface can use PCRE without
3492 modification.
3493
3494 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3495 return all that pcre_info() returns, plus additional data. The pcre_info()
3496 function is retained for compatibility, but is considered to be obsolete.
3497
3498 10. Added experimental recursion feature (?R) to handle one common case that
3499 Perl 5.6 will be able to do with (?p{...}).
3500
3501 11. Added support for POSIX character classes like [:alpha:], which Perl is
3502 adopting.
3503
3504
3505 Version 2.08 31-Aug-99
3506 ----------------------
3507
3508 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3509 trying to match at the startoffset position, but instead was moving forward to
3510 the next newline as if a previous match had failed.
3511
3512 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3513 and could get into a loop if a null string was matched other than at the start
3514 of the subject.
3515
3516 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3517 be distinguished at compile time, and for completeness also added PCRE_DATE.
3518
3519 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3520 in GnuWin32 environments.
3521
3522
3523 Version 2.07 29-Jul-99
3524 ----------------------
3525
3526 1. The documentation is now supplied in plain text form and HTML as well as in
3527 the form of man page sources.
3528
3529 2. C++ compilers don't like assigning (void *) values to other pointer types.
3530 In particular this affects malloc(). Although there is no problem in Standard
3531 C, I've put in casts to keep C++ compilers happy.
3532
3533 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3534 should be (const char *).
3535
3536 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3537 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3538 However, I haven't made this a standard facility. The documentation doesn't
3539 mention it, and the Makefile doesn't support it.
3540
3541 5. The Makefile now contains an "install" target, with editable destinations at
3542 the top of the file. The pcretest program is not installed.
3543
3544 6. pgrep -V now gives the PCRE version number and date.
3545
3546 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3547 causing the entire string to be ignored, instead of just the last character.
3548
3549 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3550 non-matching string, it can take a very, very long time, even for strings of
3551 quite modest length, because of the nested recursion. PCRE now does better in
3552 some of these cases. It does this by remembering the last required literal
3553 character in the pattern, and pre-searching the subject to ensure it is present
3554 before running the real match. In other words, it applies a heuristic to detect
3555 some types of certain failure quickly, and in the above example, if presented
3556 with a string that has no trailing " it gives "no match" very quickly.
3557
3558 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3559 other alternatives are tried instead.
3560
3561
3562 Version 2.06 09-Jun-99
3563 ----------------------
3564
3565 1. Change pcretest's output for amount of store used to show just the code
3566 space, because the remainder (the data block) varies in size between 32-bit and
3567 64-bit systems.
3568
3569 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3570 start matching at. This allows lookbehinds to work when searching for multiple
3571 occurrences in a string.
3572
3573 3. Added additional options to pcretest for testing multiple occurrences:
3574
3575 /+ outputs the rest of the string that follows a match
3576 /g loops for multiple occurrences, using the new startoffset argument
3577 /G loops for multiple occurrences by passing an incremented pointer
3578
3579 4. PCRE wasn't doing the "first character" optimization for patterns starting
3580 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3581 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3582 the letter 'x'. On long subject strings, this gives a significant speed-up.
3583
3584
3585 Version 2.05 21-Apr-99
3586 ----------------------
3587
3588 1. Changed the type of magic_number from int to long int so that it works
3589 properly on 16-bit systems.
3590
3591 2. Fixed a bug which caused patterns starting with .* not to work correctly
3592 when the subject string contained newline characters. PCRE was assuming
3593 anchoring for such patterns in all cases, which is not correct because .* will
3594 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3595 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3596 must be retried after every newline in the subject.
3597
3598
3599 Version 2.04 18-Feb-99
3600 ----------------------
3601
3602 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3603 computation of the store needed to hold the pattern was incorrect (too large).
3604 If such patterns were nested a few deep, this could multiply and become a real
3605 problem.
3606
3607 2. Added /M option to pcretest to show the memory requirement of a specific
3608 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3609
3610 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3611 compiled in such a way that the backtracking after subsequent failure was
3612 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3613 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3614
3615
3616 Version 2.03 02-Feb-99
3617 ----------------------
3618
3619 1. Fixed typo and small mistake in man page.
3620
3621 2. Added 4th condition (GPL supersedes if conflict) and created separate
3622 LICENCE file containing the conditions.
3623
3624 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3625 Perl, that is the internal \ allows the delimiter to be included in the
3626 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3627 the final delimiter, add \ to the end of the pattern (to test the error).
3628
3629 4. Added the convenience functions for extracting substrings after a successful
3630 match. Updated pcretest to make it able to test these functions.
3631
3632
3633 Version 2.02 14-Jan-99
3634 ----------------------
3635
3636 1. Initialized the working variables associated with each extraction so that
3637 their saving and restoring doesn't refer to uninitialized store.
3638
3639 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3640 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3641 fix the problem.
3642
3643 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3644 calls, and wasn't printing the correct value for compiling calls. Increased the
3645 default value of LOOPREPEAT, and the number of significant figures in the
3646 times.
3647
3648 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3649
3650 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3651 a building problem on Windows NT with a FAT file system.
3652
3653
3654 Version 2.01 21-Oct-98
3655 ----------------------
3656
3657 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3658 to character tables built by pcre_maketables() in the current locale. If NULL
3659 is passed, the default tables are used.
3660
3661
3662 Version 2.00 24-Sep-98
3663 ----------------------
3664
3665 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3666 it any more.
3667
3668 2. Allow quantification of (?>) groups, and make it work correctly.
3669
3670 3. The first character computation wasn't working for (?>) groups.
3671
3672 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3673 end of the subject) and add 5.005's \z, which really does match only at the
3674 very end of the subject.
3675
3676 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3677
3678 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3679 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3680 localized options. All options to pcre_study() were also removed.
3681
3682 7. Add other new features from 5.005:
3683
3684 $(?<= positive lookbehind
3685 $(?<! negative lookbehind
3686 (?imsx-imsx) added the unsetting capability
3687 such a setting is global if at outer level; local otherwise
3688 (?imsx-imsx:) non-capturing groups with option setting
3689 (?(cond)re|re) conditional pattern matching
3690
3691 A backreference to itself in a repeated group matches the previous
3692 captured string.
3693
3694 8. General tidying up of studying (both automatic and via "study")
3695 consequential on the addition of new assertions.
3696
3697 9. As in 5.005, unlimited repeated groups that could match an empty substring
3698 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3699 runtime if any iteration does actually match an empty substring.
3700
3701 10. Include the RunTest script in the distribution.
3702
3703 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3704 discrepancies, some of which were old and were also with respect to 5.004. They
3705 have now been fixed.
3706
3707
3708 Version 1.09 28-Apr-98
3709 ----------------------
3710
3711 1. A negated single character class followed by a quantifier with a minimum
3712 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3713 program crashes, or just wrong answers. This did not apply to negated classes
3714 containing more than one character, or to minima other than one.
3715
3716
3717 Version 1.08 27-Mar-98
3718 ----------------------
3719
3720 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3721
3722 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3723 latter must appear before anything that relies on it in the pattern.
3724
3725
3726 Version 1.07 16-Feb-98
3727 ----------------------
3728
3729 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3730 repeat of a potentially empty string).
3731
3732
3733 Version 1.06 23-Jan-98
3734 ----------------------
3735
3736 1. Added Markus Oberhumer's little patches for C++.
3737
3738 2. Literal strings longer than 255 characters were broken.
3739
3740
3741 Version 1.05 23-Dec-97
3742 ----------------------
3743
3744 1. Negated character classes containing more than one character were failing if
3745 PCRE_CASELESS was set at run time.
3746
3747
3748 Version 1.04 19-Dec-97
3749 ----------------------
3750
3751 1. Corrected the man page, where some "const" qualifiers had been omitted.
3752
3753 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3754 input syntax.
3755
3756 3. Fixed memory leak which occurred when a regex with back references was
3757 matched with an offsets vector that wasn't big enough. The temporary memory
3758 that is used in this case wasn't being freed if the match failed.
3759
3760 4. Tidied pcretest to ensure it frees memory that it gets.
3761
3762 5. Temporary memory was being obtained in the case where the passed offsets
3763 vector was exactly big enough.
3764
3765 6. Corrected definition of offsetof() from change 5 below.
3766
3767 7. I had screwed up change 6 below and broken the rules for the use of
3768 setjmp(). Now fixed.
3769
3770
3771 Version 1.03 18-Dec-97
3772 ----------------------
3773
3774 1. A erroneous regex with a missing opening parenthesis was correctly
3775 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3776 on some systems.
3777
3778 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3779 it was reported that one broken compiler failed on the former because "code" is
3780 also an independent variable.
3781
3782 3. The erroneous regex a[]b caused an array overrun reference.
3783
3784 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3785 fail on data ending with that character. (It was going on too far, and checking
3786 the next character, typically a binary zero.) This was specific to the
3787 optimized code for single-character negative classes.
3788
3789 5. Added a contributed patch from the TIN world which does the following:
3790
3791 + Add an undef for memmove, in case the the system defines a macro for it.
3792
3793 + Add a definition of offsetof(), in case there isn't one. (I don't know
3794 the reason behind this - offsetof() is part of the ANSI standard - but
3795 it does no harm).
3796
3797 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3798 most of the places where whitespace preceded '#'. I have given up and
3799 allowed the remaining 2 cases to be at the margin.
3800
3801 + Rename some variables in pcre to eliminate shadowing. This seems very
3802 pedantic, but does no harm, of course.
3803
3804 6. Moved the call to setjmp() into its own function, to get rid of warnings
3805 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3806
3807 7. Constructs such as \d{8,} were compiling into the equivalent of
3808 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3809 outcome, but in this particular case used more store than had been allocated,
3810 which caused the bug to be discovered because it threw up an internal error.
3811
3812 8. The debugging code in both pcre and pcretest for outputting the compiled
3813 form of a regex was going wrong in the case of back references followed by
3814 curly-bracketed repeats.
3815
3816
3817 Version 1.02 12-Dec-97
3818 ----------------------
3819
3820 1. Typos in pcre.3 and comments in the source fixed.
3821
3822 2. Applied a contributed patch to get rid of places where it used to remove
3823 'const' from variables, and fixed some signed/unsigned and uninitialized
3824 variable warnings.
3825
3826 3. Added the "runtest" target to Makefile.
3827
3828 4. Set default compiler flag to -O2 rather than just -O.
3829
3830
3831 Version 1.01 19-Nov-97
3832 ----------------------
3833
3834 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3835 like /([ab]*)*/, that is, for classes with more than one character in them.
3836
3837 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3838 as /((?>a*))*/ (a PCRE_EXTRA facility).
3839
3840
3841 Version 1.00 18-Nov-97
3842 ----------------------
3843
3844 1. Added compile-time macros to support systems such as SunOS4 which don't have
3845 memmove() or strerror() but have other things that can be used instead.
3846
3847 2. Arranged that "make clean" removes the executables.
3848
3849
3850 Version 0.99 27-Oct-97
3851 ----------------------
3852
3853 1. Fixed bug in code for optimizing classes with only one character. It was
3854 initializing a 32-byte map regardless, which could cause it to run off the end
3855 of the memory it had got.
3856
3857 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3858
3859
3860 Version 0.98 22-Oct-97
3861 ----------------------
3862
3863 1. Fixed bug in code for handling temporary memory usage when there are more
3864 back references than supplied space in the ovector. This could cause segfaults.
3865
3866
3867 Version 0.97 21-Oct-97
3868 ----------------------
3869
3870 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3871
3872 2. Optimized negated single characters not to use a bit map.
3873
3874 3. Brought error texts together as macro definitions; clarified some of them;
3875 fixed one that was wrong - it said "range out of order" when it meant "invalid
3876 escape sequence".
3877
3878 4. Changed some char * arguments to const char *.
3879
3880 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3881
3882 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3883 pcretest.
3884
3885
3886 Version 0.96 16-Oct-97
3887 ----------------------
3888
3889 1. Added a simple "pgrep" utility to the distribution.
3890
3891 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3892 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3893 where "ddd" means "one or more decimal digits".
3894
3895 3. Fixed serious bug. If a pattern had a back reference, but the call to
3896 pcre_exec() didn't supply a large enough ovector to record the related
3897 identifying subpattern, the match always failed. PCRE now remembers the number
3898 of the largest back reference, and gets some temporary memory in which to save
3899 the offsets during matching if necessary, in order to ensure that
3900 backreferences always work.
3901
3902 4. Increased the compatibility with Perl in a number of ways:
3903
3904 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3905 to request this handling. The option can be set at compile or exec time.
3906
3907 (b) $ matches before a terminating newline by default; an option
3908 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3909 mode). The option can be set at compile or exec time.
3910
3911 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3912 the same as Perl's. If the decimal number it represents is less than 10
3913 or there aren't that many previous left capturing parentheses, an octal
3914 escape is read. Inside a character class, it's always an octal escape,
3915 even if it is a single digit.
3916
3917 (d) An escaped but undefined alphabetic character is taken as a literal,
3918 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3919 escapes.
3920
3921 (e) {0} is now permitted. (The previous item is removed from the compiled
3922 pattern).
3923
3924 5. Changed all the names of code files so that the basic parts are no longer
3925 than 10 characters, and abolished the teeny "globals.c" file.
3926
3927 6. Changed the handling of character classes; they are now done with a 32-byte
3928 bit map always.
3929
3930 7. Added the -d and /D options to pcretest to make it possible to look at the
3931 internals of compilation without having to recompile pcre.
3932
3933
3934 Version 0.95 23-Sep-97
3935 ----------------------
3936
3937 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3938 \x20 at the start of a run of normal characters. These were being treated as
3939 real characters, instead of the source characters being re-checked.
3940
3941
3942 Version 0.94 18-Sep-97
3943 ----------------------
3944
3945 1. The functions are now thread-safe, with the caveat that the global variables
3946 containing pointers to malloc() and free() or alternative functions are the
3947 same for all threads.
3948
3949 2. Get pcre_study() to generate a bitmap of initial characters for non-
3950 anchored patterns when this is possible, and use it if passed to pcre_exec().
3951
3952
3953 Version 0.93 15-Sep-97
3954 ----------------------
3955
3956 1. /(b)|(:+)/ was computing an incorrect first character.
3957
3958 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3959 but not actually doing anything yet.
3960
3961 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3962 as Perl does (e.g. [-az] or [az-]).
3963
3964 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3965 all possible positions.
3966
3967 5. Split up into different modules to avoid including unneeded functions in a
3968 compiled binary. However, compile and exec are still in one module. The "study"
3969 function is split off.
3970
3971 6. The character tables are now in a separate module whose source is generated
3972 by an auxiliary program - but can then be edited by hand if required. There are
3973 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3974 toupper() in the code.
3975
3976 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3977 make them global. Abolish the function for setting them, as the caller can now
3978 set them directly.
3979
3980
3981 Version 0.92 11-Sep-97
3982 ----------------------
3983
3984 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3985 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3986
3987 2. Caseless matching was not working in character classes if the characters in
3988 the pattern were in upper case.
3989
3990 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3991
3992 4. Make PCRE_ANCHORED public and accept as a compile option.
3993
3994 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3995 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3996 pass them.
3997
3998 6. Give an error if bad option bits passed at compile or run time.
3999
4000 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4001 pcretest to cause it to pass that flag.
4002
4003 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4004 options, and the first character, if set.
4005
4006 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4007
4008
4009 Version 0.91 10-Sep-97
4010 ----------------------
4011
4012 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4013 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4014
4015 2. PCRE was looping on encountering an indefinitely repeated back reference to
4016 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4017 Perl does - treats the match as successful.
4018
4019 ****

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12