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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 564 - (show annotations) (download)
Sun Oct 31 16:07:24 2010 UTC (3 years, 8 months ago) by ph10
File size: 193086 byte(s)
Fix bad return code from pcregrep when -o is used (yielded 1, not 0, after a 
match).

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 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
63
64 11. When the -o option was used, pcregrep was setting a return code of 1, even
65 when matches were found, and --line-buffered was not being honoured.
66
67
68 Version 8.10 25-Jun-2010
69 ------------------------
70
71 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
72 THEN.
73
74 2. (*ACCEPT) was not working when inside an atomic group.
75
76 3. Inside a character class, \B is treated as a literal by default, but
77 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
78 causes the error). The code is unchanged, but I tidied the documentation.
79
80 4. Inside a character class, PCRE always treated \R and \X as literals,
81 whereas Perl faults them if its -w option is set. I have changed PCRE so
82 that it faults them when PCRE_EXTRA is set.
83
84 5. Added support for \N, which always matches any character other than
85 newline. (It is the same as "." when PCRE_DOTALL is not set.)
86
87 6. When compiling pcregrep with newer versions of gcc which may have
88 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
89 declared with attribute warn_unused_result" were given. Just casting the
90 result to (void) does not stop the warnings; a more elaborate fudge is
91 needed. I've used a macro to implement this.
92
93 7. Minor change to pcretest.c to avoid a compiler warning.
94
95 8. Added four artifical Unicode properties to help with an option to make
96 \s etc use properties (see next item). The new properties are: Xan
97 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
98
99 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
100 use Unicode properties. (*UCP) at the start of a pattern can be used to set
101 this option. Modified pcretest to add /W to test this facility. Added
102 REG_UCP to make it available via the POSIX interface.
103
104 10. Added --line-buffered to pcregrep.
105
106 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
107 studied, and the match started with a letter with a code point greater than
108 127 whose first byte was different to the first byte of the other case of
109 the letter, the other case of this starting letter was not recognized
110 (#976).
111
112 12. If a pattern that was studied started with a repeated Unicode property
113 test, for example, \p{Nd}+, there was the theoretical possibility of
114 setting up an incorrect bitmap of starting bytes, but fortunately it could
115 not have actually happened in practice until change 8 above was made (it
116 added property types that matched character-matching opcodes).
117
118 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
119 possible starting bytes for non-anchored patterns.
120
121 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
122 \R, and also a number of cases that involve Unicode properties, both
123 explicit and implicit when PCRE_UCP is set.
124
125 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
126 input, it could crash or give wrong results if characters with values
127 greater than 0xc0 were present in the subject string. (Detail: it assumed
128 UTF-8 input when processing these items.)
129
130 16. Added a lot of (int) casts to avoid compiler warnings in systems where
131 size_t is 64-bit (#991).
132
133 17. Added a check for running out of memory when PCRE is compiled with
134 --disable-stack-for-recursion (#990).
135
136 18. If the last data line in a file for pcretest does not have a newline on
137 the end, a newline was missing in the output.
138
139 19. The default pcre_chartables.c file recognizes only ASCII characters (values
140 less than 128) in its various bitmaps. However, there is a facility for
141 generating tables according to the current locale when PCRE is compiled. It
142 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
143 characters, are recognized by isspace() and therefore were getting set in
144 these tables, and indeed these tables seem to approximate to ISO 8859. This
145 caused a problem in UTF-8 mode when pcre_study() was used to create a list
146 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
147 which of course cannot start UTF-8 characters. I have changed the code so
148 that only real ASCII characters (less than 128) and the correct starting
149 bytes for UTF-8 encodings are set for characters greater than 127 when in
150 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
151 altogether.)
152
153 20. Added the /T option to pcretest so as to be able to run tests with non-
154 standard character tables, thus making it possible to include the tests
155 used for 19 above in the standard set of tests.
156
157 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
158 reference to a subpattern the other side of a comment that contains an
159 opening parenthesis caused either an internal compiling error, or a
160 reference to the wrong subpattern.
161
162
163 Version 8.02 19-Mar-2010
164 ------------------------
165
166 1. The Unicode data tables have been updated to Unicode 5.2.0.
167
168 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
169 configured.
170
171 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
172 original author of that file, following a query about its status.
173
174 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
175 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
176
177 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
178 quantifier applied to a forward-referencing subroutine call, could compile
179 incorrect code or give the error "internal error: previously-checked
180 referenced subpattern not found".
181
182 6. Both MS Visual Studio and Symbian OS have problems with initializing
183 variables to point to external functions. For these systems, therefore,
184 pcre_malloc etc. are now initialized to local functions that call the
185 relevant global functions.
186
187 7. There were two entries missing in the vectors called coptable and poptable
188 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
189 I've fixed the data, and added a kludgy way of testing at compile time that
190 the lengths are correct (equal to the number of opcodes).
191
192 8. Following on from 7, I added a similar kludge to check the length of the
193 eint vector in pcreposix.c.
194
195 9. Error texts for pcre_compile() are held as one long string to avoid too
196 much relocation at load time. To find a text, the string is searched,
197 counting zeros. There was no check for running off the end of the string,
198 which could happen if a new error number was added without updating the
199 string.
200
201 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
202
203 11. \K was not working if it appeared in an atomic group or in a group that
204 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
205 \K is "not well defined" if used in an assertion. PCRE now accepts it if
206 the assertion is positive, but not if it is negative.
207
208 12. Change 11 fortuitously reduced the size of the stack frame used in the
209 "match()" function of pcre_exec.c by one pointer. Forthcoming
210 implementation of support for (*MARK) will need an extra pointer on the
211 stack; I have reserved it now, so that the stack frame size does not
212 decrease.
213
214 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
215 item in branch that calls a recursion is a subroutine call - as in the
216 second branch in the above example - was incorrectly given the compile-
217 time error "recursive call could loop indefinitely" because pcre_compile()
218 was not correctly checking the subroutine for matching a non-empty string.
219
220 14. The checks for overrunning compiling workspace could trigger after an
221 overrun had occurred. This is a "should never occur" error, but it can be
222 triggered by pathological patterns such as hundreds of nested parentheses.
223 The checks now trigger 100 bytes before the end of the workspace.
224
225 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
226
227
228 Version 8.01 19-Jan-2010
229 ------------------------
230
231 1. If a pattern contained a conditional subpattern with only one branch (in
232 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
233 computed the wrong minimum data length (which is of course zero for such
234 subpatterns). This could cause incorrect "no match" results.
235
236 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
237 the pattern is reset in the first branch, pcre_compile() failed with
238 "internal error: code overflow at offset...". This happened only when
239 the reset was to the original external option setting. (An optimization
240 abstracts leading options settings into an external setting, which was the
241 cause of this.)
242
243 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
244 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
245 assertion pattern did not match (meaning that the assertion was true), it
246 was incorrectly treated as false if the SKIP had been reached during the
247 matching. This also applied to assertions used as conditions.
248
249 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
250 assertion subpattern, including such a pattern used as a condition,
251 unpredictable results occurred, instead of the error return
252 PCRE_ERROR_DFA_UITEM.
253
254 5. The C++ GlobalReplace function was not working like Perl for the special
255 situation when an empty string is matched. It now does the fancy magic
256 stuff that is necessary.
257
258 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
259 removed. (These were left over from very, very early versions of PCRE.)
260
261 7. Some cosmetic changes to the code to make life easier when compiling it
262 as part of something else:
263
264 (a) Change DEBUG to PCRE_DEBUG.
265
266 (b) In pcre_compile(), rename the member of the "branch_chain" structure
267 called "current" as "current_branch", to prevent a collision with the
268 Linux macro when compiled as a kernel module.
269
270 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
271 prevent a collision with the Linux macro when compiled as a kernel
272 module.
273
274 8. In pcre_compile() there are some checks for integer overflows that used to
275 cast potentially large values to (double). This has been changed to that
276 when building, a check for int64_t is made, and if it is found, it is used
277 instead, thus avoiding the use of floating point arithmetic. (There is no
278 other use of FP in PCRE.) If int64_t is not found, the fallback is to
279 double.
280
281 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
282 2005 (difference between two addresses compared to an unsigned value).
283
284 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
285 custom one, because of the following reported problem in Windows:
286
287 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
288 under Win32.
289 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
290 therefore missing the function definition.
291 - The compiler thus generates a "C" signature for the test function.
292 - The linker fails to find the "C" function.
293 - PCRE fails to configure if asked to do so against libbz2.
294
295 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
296 messages were output:
297
298 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
299 rerunning libtoolize, to keep the correct libtool macros in-tree.
300 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
301
302 I have done both of these things.
303
304 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
305 most of the time, it *can* run out if it is given a pattern that contains a
306 runaway infinite recursion. I updated the discussion in the pcrestack man
307 page.
308
309 13. Now that we have gone to the x.xx style of version numbers, the minor
310 version may start with zero. Using 08 or 09 is a bad idea because users
311 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
312 interpreted as invalid octal numbers. I've updated the previous comment in
313 configure.ac, and also added a check that gives an error if 08 or 09 are
314 used.
315
316 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
317 causing partial matching to fail when the end of the subject matched \W
318 in a UTF-8 pattern where \W was quantified with a minimum of 3.
319
320 15. There were some discrepancies between the declarations in pcre_internal.h
321 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
322 their definitions. The declarations used "const uschar *" and the
323 definitions used USPTR. Even though USPTR is normally defined as "const
324 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
325 reported that: "This difference in casting confuses some C++ compilers, for
326 example, SunCC recognizes above declarations as different functions and
327 generates broken code for hbpcre." I have changed the declarations to use
328 USPTR.
329
330 16. GNU libtool is named differently on some systems. The autogen.sh script now
331 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
332 (FreeBSD).
333
334 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
335 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
336 comment: "Figure out how to create a longlong from a string: strtoll and
337 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
338 instance, but it only takes 2 args instead of 3!"
339
340 18. A subtle bug concerned with back references has been fixed by a change of
341 specification, with a corresponding code fix. A pattern such as
342 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
343 refers, was giving matches when it shouldn't. For example, xa=xaaa would
344 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
345 same bug. Such groups have to be quantified to be useful, or contained
346 inside another quantified group. (If there's no repetition, the reference
347 can never match.) The problem arises because, having left the group and
348 moved on to the rest of the pattern, a later failure that backtracks into
349 the group uses the captured value from the final iteration of the group
350 rather than the correct earlier one. I have fixed this in PCRE by forcing
351 any group that contains a reference to itself to be an atomic group; that
352 is, there cannot be any backtracking into it once it has completed. This is
353 similar to recursive and subroutine calls.
354
355
356 Version 8.00 19-Oct-09
357 ----------------------
358
359 1. The table for translating pcre_compile() error codes into POSIX error codes
360 was out-of-date, and there was no check on the pcre_compile() error code
361 being within the table. This could lead to an OK return being given in
362 error.
363
364 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
365 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
366 in a Windows environment.
367
368 3. The pcregrep --count option prints the count for each file even when it is
369 zero, as does GNU grep. However, pcregrep was also printing all files when
370 --files-with-matches was added. Now, when both options are given, it prints
371 counts only for those files that have at least one match. (GNU grep just
372 prints the file name in this circumstance, but including the count seems
373 more useful - otherwise, why use --count?) Also ensured that the
374 combination -clh just lists non-zero counts, with no names.
375
376 4. The long form of the pcregrep -F option was incorrectly implemented as
377 --fixed_strings instead of --fixed-strings. This is an incompatible change,
378 but it seems right to fix it, and I didn't think it was worth preserving
379 the old behaviour.
380
381 5. The command line items --regex=pattern and --regexp=pattern were not
382 recognized by pcregrep, which required --regex pattern or --regexp pattern
383 (with a space rather than an '='). The man page documented the '=' forms,
384 which are compatible with GNU grep; these now work.
385
386 6. No libpcreposix.pc file was created for pkg-config; there was just
387 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
388
389 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
390 when UCP support is not needed, by modifying the Python script that
391 generates it from Unicode data files. This should not matter if the module
392 is correctly used as a library, but I received one complaint about 50K of
393 unwanted data. My guess is that the person linked everything into his
394 program rather than using a library. Anyway, it does no harm.
395
396 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
397 was a minimum greater than 1 for a wide character in a possessive
398 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
399 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
400 character. Chaos in the form of incorrect output or a compiling loop could
401 result.
402
403 9. The restrictions on what a pattern can contain when partial matching is
404 requested for pcre_exec() have been removed. All patterns can now be
405 partially matched by this function. In addition, if there are at least two
406 slots in the offset vector, the offset of the earliest inspected character
407 for the match and the offset of the end of the subject are set in them when
408 PCRE_ERROR_PARTIAL is returned.
409
410 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
411 synonymous with PCRE_PARTIAL, for backwards compatibility, and
412 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
413 and may be more useful for multi-segment matching.
414
415 11. Partial matching with pcre_exec() is now more intuitive. A partial match
416 used to be given if ever the end of the subject was reached; now it is
417 given only if matching could not proceed because another character was
418 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
419 string "Z", which now yields "no match" instead of "partial match". In the
420 case of pcre_dfa_exec(), "no match" is given if every matching path for the
421 final character ended with (*FAIL).
422
423 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
424 if the pattern had a "must contain" character that was already found in the
425 earlier partial match, unless partial matching was again requested. For
426 example, with the pattern /dog.(body)?/, the "must contain" character is
427 "g". If the first part-match was for the string "dog", restarting with
428 "sbody" failed. This bug has been fixed.
429
430 13. The string returned by pcre_dfa_exec() after a partial match has been
431 changed so that it starts at the first inspected character rather than the
432 first character of the match. This makes a difference only if the pattern
433 starts with a lookbehind assertion or \b or \B (\K is not supported by
434 pcre_dfa_exec()). It's an incompatible change, but it makes the two
435 matching functions compatible, and I think it's the right thing to do.
436
437 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
438 so that the demonstration program is easily available in environments where
439 PCRE has not been installed from source.
440
441 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
442 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
443 library.
444
445 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
446 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
447 is not the first non-POSIX option to be added. Clearly some people find
448 these options useful.
449
450 17. If a caller to the POSIX matching function regexec() passes a non-zero
451 value for nmatch with a NULL value for pmatch, the value of
452 nmatch is forced to zero.
453
454 18. RunGrepTest did not have a test for the availability of the -u option of
455 the diff command, as RunTest does. It now checks in the same way as
456 RunTest, and also checks for the -b option.
457
458 19. If an odd number of negated classes containing just a single character
459 interposed, within parentheses, between a forward reference to a named
460 subpattern and the definition of the subpattern, compilation crashed with
461 an internal error, complaining that it could not find the referenced
462 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
463 [The bug was that it was starting one character too far in when skipping
464 over the character class, thus treating the ] as data rather than
465 terminating the class. This meant it could skip too much.]
466
467 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
468 /g option in pcretest when the pattern contains \K, which makes it possible
469 to have an empty string match not at the start, even when the pattern is
470 anchored. Updated pcretest and pcredemo to use this option.
471
472 21. If the maximum number of capturing subpatterns in a recursion was greater
473 than the maximum at the outer level, the higher number was returned, but
474 with unset values at the outer level. The correct (outer level) value is
475 now given.
476
477 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
478 PCRE did not set those parentheses (unlike Perl). I have now found a way to
479 make it do so. The string so far is captured, making this feature
480 compatible with Perl.
481
482 23. The tests have been re-organized, adding tests 11 and 12, to make it
483 possible to check the Perl 5.10 features against Perl 5.10.
484
485 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
486 pattern matches a fixed length string. PCRE did not allow this; now it
487 does. Neither allows recursion.
488
489 25. I finally figured out how to implement a request to provide the minimum
490 length of subject string that was needed in order to match a given pattern.
491 (It was back references and recursion that I had previously got hung up
492 on.) This code has now been added to pcre_study(); it finds a lower bound
493 to the length of subject needed. It is not necessarily the greatest lower
494 bound, but using it to avoid searching strings that are too short does give
495 some useful speed-ups. The value is available to calling programs via
496 pcre_fullinfo().
497
498 26. While implementing 25, I discovered to my embarrassment that pcretest had
499 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
500 study optimizations had never been tested with that matching function.
501 Oops. What is worse, even when it was passed study data, there was a bug in
502 pcre_dfa_exec() that meant it never actually used it. Double oops. There
503 were also very few tests of studied patterns with pcre_dfa_exec().
504
505 27. If (?| is used to create subpatterns with duplicate numbers, they are now
506 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
507 on the other side of the coin, they are no longer allowed to have different
508 names, because these cannot be distinguished in PCRE, and this has caused
509 confusion. (This is a difference from Perl.)
510
511 28. When duplicate subpattern names are present (necessarily with different
512 numbers, as required by 27 above), and a test is made by name in a
513 conditional pattern, either for a subpattern having been matched, or for
514 recursion in such a pattern, all the associated numbered subpatterns are
515 tested, and the overall condition is true if the condition is true for any
516 one of them. This is the way Perl works, and is also more like the way
517 testing by number works.
518
519
520 Version 7.9 11-Apr-09
521 ---------------------
522
523 1. When building with support for bzlib/zlib (pcregrep) and/or readline
524 (pcretest), all targets were linked against these libraries. This included
525 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
526 libraries. This caused unwanted dependencies to be created. This problem
527 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
528 pcretest is linked with readline.
529
530 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
531 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
532 moved outside it again, because FALSE and TRUE are already defined in AIX,
533 but BOOL is not.
534
535 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
536 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
537
538 4. The pcregrep documentation said spaces were inserted as well as colons (or
539 hyphens) following file names and line numbers when outputting matching
540 lines. This is not true; no spaces are inserted. I have also clarified the
541 wording for the --colour (or --color) option.
542
543 5. In pcregrep, when --colour was used with -o, the list of matching strings
544 was not coloured; this is different to GNU grep, so I have changed it to be
545 the same.
546
547 6. When --colo(u)r was used in pcregrep, only the first matching substring in
548 each matching line was coloured. Now it goes on to look for further matches
549 of any of the test patterns, which is the same behaviour as GNU grep.
550
551 7. A pattern that could match an empty string could cause pcregrep to loop; it
552 doesn't make sense to accept an empty string match in pcregrep, so I have
553 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
554 seems to be how GNU grep behaves.
555
556 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
557 start or after a newline", because the conditional assertion was not being
558 correctly handled. The rule now is that both the assertion and what follows
559 in the first alternative must satisfy the test.
560
561 9. If auto-callout was enabled in a pattern with a conditional group whose
562 condition was an assertion, PCRE could crash during matching, both with
563 pcre_exec() and pcre_dfa_exec().
564
565 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
566 used for matching.
567
568 11. Unicode property support in character classes was not working for
569 characters (bytes) greater than 127 when not in UTF-8 mode.
570
571 12. Added the -M command line option to pcretest.
572
573 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
574
575 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
576
577 16. Added comments and documentation about mis-use of no_arg in the C++
578 wrapper.
579
580 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
581 from Martin Jerabek that uses macro names for all relevant character and
582 string constants.
583
584 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
585 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
586 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
587 these, but not everybody uses configure.
588
589 19. A conditional group that had only one branch was not being correctly
590 recognized as an item that could match an empty string. This meant that an
591 enclosing group might also not be so recognized, causing infinite looping
592 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
593 with the subject "ab", where knowledge that the repeated group can match
594 nothing is needed in order to break the loop.
595
596 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
597 exec(), but without supplying a callout function, matching went wrong.
598
599 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
600 leak if the size of the offset vector was greater than 30. When the vector
601 is smaller, the saved offsets during recursion go onto a local stack
602 vector, but for larger vectors malloc() is used. It was failing to free
603 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
604 error, in fact).
605
606 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
607 heapframe that is used only when UTF-8 support is enabled. This caused no
608 problem, but was untidy.
609
610 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
611 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
612 included within another project.
613
614 24. Steven Van Ingelgem's patches to add more options to the CMake support,
615 slightly modified by me:
616
617 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
618 not building pcregrep.
619
620 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
621 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
622
623 25. Forward references, both numeric and by name, in patterns that made use of
624 duplicate group numbers, could behave incorrectly or give incorrect errors,
625 because when scanning forward to find the reference group, PCRE was not
626 taking into account the duplicate group numbers. A pattern such as
627 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
628
629 26. Changed a few more instances of "const unsigned char *" to USPTR, making
630 the feature of a custom pointer more persuasive (as requested by a user).
631
632 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
633 pcretest.c, inside #ifndefs, because it seems they are sometimes already
634 pre-defined.
635
636 28. Added support for (*UTF8) at the start of a pattern.
637
638 29. Arrange for flags added by the "release type" setting in CMake to be shown
639 in the configuration summary.
640
641
642 Version 7.8 05-Sep-08
643 ---------------------
644
645 1. Replaced UCP searching code with optimized version as implemented for Ad
646 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
647 stage table and inline lookup instead of a function, giving speed ups of 2
648 to 5 times on some simple patterns that I tested. Permission was given to
649 distribute the MultiStage2.py script that generates the tables (it's not in
650 the tarball, but is in the Subversion repository).
651
652 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
653 scripts.
654
655 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
656 a group with a zero qualifier. The result of the study could be incorrect,
657 or the function might crash, depending on the pattern.
658
659 4. Caseless matching was not working for non-ASCII characters in back
660 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
661 It now works when Unicode Property Support is available.
662
663 5. In pcretest, an escape such as \x{de} in the data was always generating
664 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
665 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
666 truncation.
667
668 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
669
670 7. Added two (int) casts to pcregrep when printing the difference of two
671 pointers, in case they are 64-bit values.
672
673 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
674 test 2 if it fails.
675
676 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
677 and a #define of that name to empty if it is not externally set. This is to
678 allow users of MSVC to set it if necessary.
679
680 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
681 the convenience functions in the pcre_get.c source file.
682
683 11. An option change at the start of a pattern that had top-level alternatives
684 could cause overwriting and/or a crash. This command provoked a crash in
685 some environments:
686
687 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
688
689 This potential security problem was recorded as CVE-2008-2371.
690
691 12. For a pattern where the match had to start at the beginning or immediately
692 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
693 pcre_dfa_exec() could read past the end of the passed subject if there was
694 no match. To help with detecting such bugs (e.g. with valgrind), I modified
695 pcretest so that it places the subject at the end of its malloc-ed buffer.
696
697 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
698 exec() might read past the end of the data buffer in UTF-8 mode.
699
700 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
701 the data contained the byte 0x85 as part of a UTF-8 character within its
702 first line. This applied both to normal and DFA matching.
703
704 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
705 /^[^d]*?$/8 failed to match "abc".
706
707 16. Added a missing copyright notice to pcrecpp_internal.h.
708
709 17. Make it more clear in the documentation that values returned from
710 pcre_exec() in ovector are byte offsets, not character counts.
711
712 18. Tidied a few places to stop certain compilers from issuing warnings.
713
714 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
715 supplied by Stefan Weber. I made a further small update for 7.8 because
716 there is a change of source arrangements: the pcre_searchfuncs.c module is
717 replaced by pcre_ucd.c.
718
719
720 Version 7.7 07-May-08
721 ---------------------
722
723 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
724 a string to a long long, pretend we don't even have a long long." This is
725 done by checking for the strtoq, strtoll, and _strtoi64 functions.
726
727 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
728 pre-7.6 versions, which defined a global no_arg variable instead of putting
729 it in the RE class. (See also #8 below.)
730
731 3. Remove a line of dead code, identified by coverity and reported by Nuno
732 Lopes.
733
734 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
735
736 (1) The include/exclude patterns were being applied to the whole pathnames
737 of files, instead of just to the final components.
738
739 (2) If there was more than one level of directory, the subdirectories were
740 skipped unless they satisfied the include/exclude conditions. This is
741 inconsistent with GNU grep (and could even be seen as contrary to the
742 pcregrep specification - which I improved to make it absolutely clear).
743 The action now is always to scan all levels of directory, and just
744 apply the include/exclude patterns to regular files.
745
746 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
747 --exclude_dir in the tests to avoid scanning .svn directories.
748
749 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
750 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
751 doesn't support NULs in patterns.
752
753 7. Added some missing "const"s to declarations of static tables in
754 pcre_compile.c and pcre_dfa_exec.c.
755
756 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
757 caused by fix #2 above. (Subsequently also a second patch to fix the
758 first patch. And a third patch - this was a messy problem.)
759
760 9. Applied Craig's patch to remove the use of push_back().
761
762 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
763 matching function regexec().
764
765 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
766 which, however, unlike Perl's \g{...}, are subroutine calls, not back
767 references. PCRE supports relative numbers with this syntax (I don't think
768 Oniguruma does).
769
770 12. Previously, a group with a zero repeat such as (...){0} was completely
771 omitted from the compiled regex. However, this means that if the group
772 was called as a subroutine from elsewhere in the pattern, things went wrong
773 (an internal error was given). Such groups are now left in the compiled
774 pattern, with a new opcode that causes them to be skipped at execution
775 time.
776
777 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
778 to the way PCRE behaves:
779
780 (a) A lone ] character is dis-allowed (Perl treats it as data).
781
782 (b) A back reference to an unmatched subpattern matches an empty string
783 (Perl fails the current match path).
784
785 (c) A data ] in a character class must be notated as \] because if the
786 first data character in a class is ], it defines an empty class. (In
787 Perl it is not possible to have an empty class.) The empty class []
788 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
789 The negative empty class [^] matches any one character, independently
790 of the DOTALL setting.
791
792 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
793 non-existent subpattern following a character class starting with ']' and
794 containing () gave an internal compiling error instead of "reference to
795 non-existent subpattern". Fortunately, when the pattern did exist, the
796 compiled code was correct. (When scanning forwards to check for the
797 existencd of the subpattern, it was treating the data ']' as terminating
798 the class, so got the count wrong. When actually compiling, the reference
799 was subsequently set up correctly.)
800
801 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
802 it was being rejected as not supported by pcre_dfa_exec(), even though
803 other assertions are supported. I have made pcre_dfa_exec() support
804 (*FAIL).
805
806 16. The implementation of 13c above involved the invention of a new opcode,
807 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
808 cannot be changed at match time, I realized I could make a small
809 improvement to matching performance by compiling OP_ALLANY instead of
810 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
811 on the OP_ANY path.
812
813 17. Compiling pcretest on Windows with readline support failed without the
814 following two fixes: (1) Make the unistd.h include conditional on
815 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
816
817 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
818 ncurses library to be included for pcretest when ReadLine support is
819 requested, but also to allow for it to be overridden. This patch came from
820 Daniel Bergström.
821
822 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
823 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
824 any errors with the current Unicode tables. Thanks to Peter Kankowski for
825 spotting this.
826
827
828 Version 7.6 28-Jan-08
829 ---------------------
830
831 1. A character class containing a very large number of characters with
832 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
833 overflow.
834
835 2. Patch to cut out the "long long" test in pcrecpp_unittest when
836 HAVE_LONG_LONG is not defined.
837
838 3. Applied Christian Ehrlicher's patch to update the CMake build files to
839 bring them up to date and include new features. This patch includes:
840
841 - Fixed PH's badly added libz and libbz2 support.
842 - Fixed a problem with static linking.
843 - Added pcredemo. [But later removed - see 7 below.]
844 - Fixed dftables problem and added an option.
845 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
846 HAVE_LONG_LONG.
847 - Added readline support for pcretest.
848 - Added an listing of the option settings after cmake has run.
849
850 4. A user submitted a patch to Makefile that makes it easy to create
851 "pcre.dll" under mingw when using Configure/Make. I added stuff to
852 Makefile.am that cause it to include this special target, without
853 affecting anything else. Note that the same mingw target plus all
854 the other distribution libraries and programs are now supported
855 when configuring with CMake (see 6 below) instead of with
856 Configure/Make.
857
858 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
859 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
860 exported in the Windows port". It has not yet been confirmed that the patch
861 solves the problem, but it does no harm.
862
863 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
864 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
865 with CMake, and also correct the comment about stack recursion.
866
867 7. Remove the automatic building of pcredemo from the ./configure system and
868 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
869 of a program that users should build themselves after PCRE is installed, so
870 building it automatically is not really right. What is more, it gave
871 trouble in some build environments.
872
873 8. Further tidies to CMakeLists.txt from Sheri and Christian.
874
875
876 Version 7.5 10-Jan-08
877 ---------------------
878
879 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
880 values in parens when parsing an RE using the C++ wrapper."
881
882 2. Negative specials like \S did not work in character classes in UTF-8 mode.
883 Characters greater than 255 were excluded from the class instead of being
884 included.
885
886 3. The same bug as (2) above applied to negated POSIX classes such as
887 [:^space:].
888
889 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
890 defined or documented. It seems to have been a typo for PCRE_STATIC, so
891 I have changed it.
892
893 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
894 first named subpattern) and a construct such as (?&a) would reference the
895 first named subpattern whose name started with "a" (in other words, the
896 length check was missing). Both these problems are fixed. "Subpattern name
897 expected" is now given for (?&) (a zero-length name), and this patch also
898 makes it give the same error for \k'' (previously it complained that that
899 was a reference to a non-existent subpattern).
900
901 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
902 this is right because (?- can be followed by option settings as well as by
903 digits. I have, however, made the messages clearer.
904
905 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
906 than the number used in the conditional) now cause a compile-time error.
907 This is actually not compatible with Perl, which accepts such patterns, but
908 treats the conditional as always being FALSE (as PCRE used to), but it
909 seems to me that giving a diagnostic is better.
910
911 8. Change "alphameric" to the more common word "alphanumeric" in comments
912 and messages.
913
914 9. Fix two occurrences of "backslash" in comments that should have been
915 "backspace".
916
917 10. Remove two redundant lines of code that can never be obeyed (their function
918 was moved elsewhere).
919
920 11. The program that makes PCRE's Unicode character property table had a bug
921 which caused it to generate incorrect table entries for sequences of
922 characters that have the same character type, but are in different scripts.
923 It amalgamated them into a single range, with the script of the first of
924 them. In other words, some characters were in the wrong script. There were
925 thirteen such cases, affecting characters in the following ranges:
926
927 U+002b0 - U+002c1
928 U+0060c - U+0060d
929 U+0061e - U+00612
930 U+0064b - U+0065e
931 U+0074d - U+0076d
932 U+01800 - U+01805
933 U+01d00 - U+01d77
934 U+01d9b - U+01dbf
935 U+0200b - U+0200f
936 U+030fc - U+030fe
937 U+03260 - U+0327f
938 U+0fb46 - U+0fbb1
939 U+10450 - U+1049d
940
941 12. The -o option (show only the matching part of a line) for pcregrep was not
942 compatible with GNU grep in that, if there was more than one match in a
943 line, it showed only the first of them. It now behaves in the same way as
944 GNU grep.
945
946 13. If the -o and -v options were combined for pcregrep, it printed a blank
947 line for every non-matching line. GNU grep prints nothing, and pcregrep now
948 does the same. The return code can be used to tell if there were any
949 non-matching lines.
950
951 14. Added --file-offsets and --line-offsets to pcregrep.
952
953 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
954 infinitely looping recursion. The bug was that positive lookaheads were not
955 being skipped when checking for a possible empty match (negative lookaheads
956 and both kinds of lookbehind were skipped).
957
958 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
959 inclusion of <windows.h> to before rather than after the definition of
960 INVALID_FILE_ATTRIBUTES (patch from David Byron).
961
962 17. Specifying a possessive quantifier with a specific limit for a Unicode
963 character property caused pcre_compile() to compile bad code, which led at
964 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
965 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
966 caused the error; without that there was no problem.
967
968 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
969
970 19. Added --enable-pcretest-libreadline.
971
972 20. In pcrecpp.cc, the variable 'count' was incremented twice in
973 RE::GlobalReplace(). As a result, the number of replacements returned was
974 double what it should be. I removed one of the increments, but Craig sent a
975 later patch that removed the other one (the right fix) and added unit tests
976 that check the return values (which was not done before).
977
978 21. Several CMake things:
979
980 (1) Arranged that, when cmake is used on Unix, the libraries end up with
981 the names libpcre and libpcreposix, not just pcre and pcreposix.
982
983 (2) The above change means that pcretest and pcregrep are now correctly
984 linked with the newly-built libraries, not previously installed ones.
985
986 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
987
988 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
989 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
990 UTF-8 newline character). The key issue is that the pattern starts .*;
991 this means that the match must be either at the beginning, or after a
992 newline. The bug was in the code for advancing after a failed match and
993 checking that the new position followed a newline. It was not taking
994 account of UTF-8 characters correctly.
995
996 23. PCRE was behaving differently from Perl in the way it recognized POSIX
997 character classes. PCRE was not treating the sequence [:...:] as a
998 character class unless the ... were all letters. Perl, however, seems to
999 allow any characters between [: and :], though of course it rejects as
1000 unknown any "names" that contain non-letters, because all the known class
1001 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1002 for example, whereas PCRE did not - it did not recognize a POSIX character
1003 class. This seemed a bit dangerous, so the code has been changed to be
1004 closer to Perl. The behaviour is not identical to Perl, because PCRE will
1005 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1006 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1007 Perl does, and where it didn't before.
1008
1009 24. Rewrite so as to remove the single use of %n from pcregrep because in some
1010 Windows environments %n is disabled by default.
1011
1012
1013 Version 7.4 21-Sep-07
1014 ---------------------
1015
1016 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
1017 means that a class such as [\s] counted as "explicit reference to CR or
1018 LF". That isn't really right - the whole point of the change was to try to
1019 help when there was an actual mention of one of the two characters. So now
1020 the change happens only if \r or \n (or a literal CR or LF) character is
1021 encountered.
1022
1023 2. The 32-bit options word was also used for 6 internal flags, but the numbers
1024 of both had grown to the point where there were only 3 bits left.
1025 Fortunately, there was spare space in the data structure, and so I have
1026 moved the internal flags into a new 16-bit field to free up more option
1027 bits.
1028
1029 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1030 but did not set the internal JCHANGED flag - either of these is enough to
1031 control the way the "get" function works - but the PCRE_INFO_JCHANGED
1032 facility is supposed to tell if (?J) was ever used, so now (?J) at the
1033 start sets both bits.
1034
1035 4. Added options (at build time, compile time, exec time) to change \R from
1036 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1037
1038 5. doc/pcresyntax.html was missing from the distribution.
1039
1040 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1041 compatibility, even though it is no longer used.
1042
1043 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1044 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1045 windows.h file is present (where different names are used). [This was
1046 reversed later after testing - see 16 below.]
1047
1048 8. Changed all #include <config.h> to #include "config.h". There were also
1049 some further <pcre.h> cases that I changed to "pcre.h".
1050
1051 9. When pcregrep was used with the --colour option, it missed the line ending
1052 sequence off the lines that it output.
1053
1054 10. It was pointed out to me that arrays of string pointers cause lots of
1055 relocations when a shared library is dynamically loaded. A technique of
1056 using a single long string with a table of offsets can drastically reduce
1057 these. I have refactored PCRE in four places to do this. The result is
1058 dramatic:
1059
1060 Originally: 290
1061 After changing UCP table: 187
1062 After changing error message table: 43
1063 After changing table of "verbs" 36
1064 After changing table of Posix names 22
1065
1066 Thanks to the folks working on Gregex for glib for this insight.
1067
1068 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1069 unicode-properties was also set.
1070
1071 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1072
1073 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1074 checked only for CRLF.
1075
1076 14. Added casts to pcretest.c to avoid compiler warnings.
1077
1078 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1079
1080 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1081 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1082 entirely. This removes changes made in 7 above.
1083
1084 17. The CMake files have been updated, and there is now more information about
1085 building with CMake in the NON-UNIX-USE document.
1086
1087
1088 Version 7.3 28-Aug-07
1089 ---------------------
1090
1091 1. In the rejigging of the build system that eventually resulted in 7.1, the
1092 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1093 brackets there is not right, since it causes compilers to look for an
1094 installed pcre.h, not the version that is in the source that is being
1095 compiled (which of course may be different). I have changed it back to:
1096
1097 #include "pcre.h"
1098
1099 I have a vague recollection that the change was concerned with compiling in
1100 different directories, but in the new build system, that is taken care of
1101 by the VPATH setting the Makefile.
1102
1103 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1104 when the subject happened to end in the byte 0x85 (e.g. if the last
1105 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1106 characters but of course it shouldn't be taken as a newline when it is part
1107 of another character. The bug was that, for an unlimited repeat of . in
1108 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1109 characters when looking for a newline.
1110
1111 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1112
1113 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1114 in debug output.
1115
1116 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1117 long printing in the pcrecpp unittest when running under MinGW.
1118
1119 6. ESC_K was left out of the EBCDIC table.
1120
1121 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1122 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1123 limit also applies to "virtual nesting" when a pattern is recursive, and in
1124 this case 1000 isn't so big. I have been able to remove this limit at the
1125 expense of backing off one optimization in certain circumstances. Normally,
1126 when pcre_exec() would call its internal match() function recursively and
1127 immediately return the result unconditionally, it uses a "tail recursion"
1128 feature to save stack. However, when a subpattern that can match an empty
1129 string has an unlimited repetition quantifier, it no longer makes this
1130 optimization. That gives it a stack frame in which to save the data for
1131 checking that an empty string has been matched. Previously this was taken
1132 from the 1000-entry workspace that had been reserved. So now there is no
1133 explicit limit, but more stack is used.
1134
1135 8. Applied Daniel's patches to solve problems with the import/export magic
1136 syntax that is required for Windows, and which was going wrong for the
1137 pcreposix and pcrecpp parts of the library. These were overlooked when this
1138 problem was solved for the main library.
1139
1140 9. There were some crude static tests to avoid integer overflow when computing
1141 the size of patterns that contain repeated groups with explicit upper
1142 limits. As the maximum quantifier is 65535, the maximum group length was
1143 set at 30,000 so that the product of these two numbers did not overflow a
1144 32-bit integer. However, it turns out that people want to use groups that
1145 are longer than 30,000 bytes (though not repeat them that many times).
1146 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1147 made it possible to implement the integer overflow checks in a much more
1148 dynamic way, which I have now done. The artificial limitation on group
1149 length has been removed - we now have only the limit on the total length of
1150 the compiled pattern, which depends on the LINK_SIZE setting.
1151
1152 10. Fixed a bug in the documentation for get/copy named substring when
1153 duplicate names are permitted. If none of the named substrings are set, the
1154 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1155 empty string.
1156
1157 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1158 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1159 because the ] is interpreted as the first data character and the
1160 terminating ] is not found. PCRE has been made compatible with Perl in this
1161 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1162 cause memory overwriting.
1163
1164 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1165 string has been matched (to stop an infinite loop). It was not recognizing
1166 a conditional subpattern that could match an empty string if that
1167 subpattern was within another subpattern. For example, it looped when
1168 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1169 condition was not nested. This bug has been fixed.
1170
1171 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1172 past the start of the subject in the presence of bytes with the top bit
1173 set, for example "\x8aBCD".
1174
1175 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1176 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1177
1178 14. Optimized (?!) to (*FAIL).
1179
1180 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1181 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1182 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1183 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1184 does: it's just the validity check that is more restrictive.
1185
1186 16. Inserted checks for integer overflows during escape sequence (backslash)
1187 processing, and also fixed erroneous offset values for syntax errors during
1188 backslash processing.
1189
1190 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1191 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1192
1193 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1194 caused an overrun.
1195
1196 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1197 something other than just ASCII characters) inside a group that had an
1198 unlimited repeat caused a loop at compile time (while checking to see
1199 whether the group could match an empty string).
1200
1201 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1202 [\P{Any}] did so. (Error in the code for printing property names.)
1203
1204 21. An orphan \E inside a character class could cause a crash.
1205
1206 22. A repeated capturing bracket such as (A)? could cause a wild memory
1207 reference during compilation.
1208
1209 23. There are several functions in pcre_compile() that scan along a compiled
1210 expression for various reasons (e.g. to see if it's fixed length for look
1211 behind). There were bugs in these functions when a repeated \p or \P was
1212 present in the pattern. These operators have additional parameters compared
1213 with \d, etc, and these were not being taken into account when moving along
1214 the compiled data. Specifically:
1215
1216 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1217 length.
1218
1219 (b) An item such as \pL+ within a repeated group could cause crashes or
1220 loops.
1221
1222 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1223 "reference to non-existent subpattern" error.
1224
1225 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1226
1227 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1228 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1229
1230 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1231
1232 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1233 character were causing crashes (broken optimization).
1234
1235 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1236 \p or \P) caused a compile-time loop.
1237
1238 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1239 break. For example, the unstudied pattern [\r\n]A does not match the string
1240 "\r\nA" because change 7.0/46 below moves the current point on by two
1241 characters after failing to match at the start. However, the pattern \nA
1242 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1243 the same is true. There doesn't seem any very clean way out of this, but
1244 what I have chosen to do makes the common cases work: PCRE now takes note
1245 of whether there can be an explicit match for \r or \n anywhere in the
1246 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1247 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1248 pattern has explicit CR or LF references.
1249
1250 29. Added (*CR) etc for changing newline setting at start of pattern.
1251
1252
1253 Version 7.2 19-Jun-07
1254 ---------------------
1255
1256 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1257 which is apparently normally available under Windows.
1258
1259 2. Re-jig the pcregrep tests with different newline settings in an attempt
1260 to make them independent of the local environment's newline setting.
1261
1262 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1263
1264 4. Some of the "internals" tests were previously cut out when the link size
1265 was not 2, because the output contained actual offsets. The recent new
1266 "Z" feature of pcretest means that these can be cut out, making the tests
1267 usable with all link sizes.
1268
1269 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1270 stack recursion. This gives a massive performance boost under BSD, but just
1271 a small improvement under Linux. However, it saves one field in the frame
1272 in all cases.
1273
1274 6. Added more features from the forthcoming Perl 5.10:
1275
1276 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1277 recursion call. It refers to the nth most recently opened parentheses.
1278
1279 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1280 to be opened parentheses.
1281
1282 (c) Conditions that refer to capturing parentheses can be specified
1283 relatively, for example, (?(-2)... or (?(+3)...
1284
1285 (d) \K resets the start of the current match so that everything before
1286 is not part of it.
1287
1288 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1289
1290 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1291 reference syntax.
1292
1293 (g) (?| introduces a group in which the numbering of parentheses in each
1294 alternative starts with the same number.
1295
1296 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1297
1298 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1299 PCRE_INFO_JCHANGED.
1300
1301 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1302 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1303 for detecting groups that can match an empty string.
1304
1305 9. A pattern with a very large number of alternatives (more than several
1306 hundred) was running out of internal workspace during the pre-compile
1307 phase, where pcre_compile() figures out how much memory will be needed. A
1308 bit of new cunning has reduced the workspace needed for groups with
1309 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1310 workspace instead of running out of the 4096 that are available.
1311
1312 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1313
1314 11. Applied patch from Google to remove an optimization that didn't quite work.
1315 The report of the bug said:
1316
1317 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1318 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1319 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1320
1321 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1322 it matched the wrong number of bytes.
1323
1324
1325 Version 7.1 24-Apr-07
1326 ---------------------
1327
1328 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1329 that is more "standard", making use of automake and other Autotools. There
1330 is some re-arrangement of the files and adjustment of comments consequent
1331 on this.
1332
1333 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1334 for recursive directory scanning broke on some systems because the files
1335 are not scanned in any specific order and on different systems the order
1336 was different. A call to "sort" has been inserted into RunGrepTest for the
1337 approprate test as a short-term fix. In the longer term there may be an
1338 alternative.
1339
1340 3. I had an email from Eric Raymond about problems translating some of PCRE's
1341 man pages to HTML (despite the fact that I distribute HTML pages, some
1342 people do their own conversions for various reasons). The problems
1343 concerned the use of low-level troff macros .br and .in. I have therefore
1344 removed all such uses from the man pages (some were redundant, some could
1345 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1346 HTML has been updated to handle .nf/.fi and to complain if it encounters
1347 .br or .in.
1348
1349 4. Updated comments in configure.ac that get placed in config.h.in and also
1350 arranged for config.h to be included in the distribution, with the name
1351 config.h.generic, for the benefit of those who have to compile without
1352 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1353
1354 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1355 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1356 makevp.bat for the current PCRE, using the additional files
1357 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1358
1359 6. A Windows user reported a minor discrepancy with test 2, which turned out
1360 to be caused by a trailing space on an input line that had got lost in his
1361 copy. The trailing space was an accident, so I've just removed it.
1362
1363 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1364 that is needed.
1365
1366 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1367 as "const" (a) because they are and (b) because it helps the PHP
1368 maintainers who have recently made a script to detect big data structures
1369 in the php code that should be moved to the .rodata section. I remembered
1370 to update Builducptable as well, so it won't revert if ucptable.h is ever
1371 re-created.
1372
1373 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1374 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1375 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1376 support is not required. This saves 1.5-2K of code, which is important in
1377 some applications.
1378
1379 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1380 so as not to refer to the tables, even though these functions will never be
1381 called when UTF-8 support is disabled. Otherwise there are problems with a
1382 shared library.
1383
1384 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1385
1386 (a) It was defining its arguments as char * instead of void *.
1387
1388 (b) It was assuming that all moves were upwards in memory; this was true
1389 a long time ago when I wrote it, but is no longer the case.
1390
1391 The emulated memove() is provided for those environments that have neither
1392 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1393 is clearly not the case, as these two bugs were recently reported.
1394
1395 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1396 and Detrail to create the HTML documentation, the .txt form of the man
1397 pages, and it removes trailing spaces from listed files. It also creates
1398 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1399 case, it wraps all the #defines with #ifndefs. This script should be run
1400 before "make dist".
1401
1402 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1403 with Unicode property support.
1404
1405 (a) For a maximizing quantifier, if the two different cases of the
1406 character were of different lengths in their UTF-8 codings (there are
1407 some cases like this - I found 11), and the matching function had to
1408 back up over a mixture of the two cases, it incorrectly assumed they
1409 were both the same length.
1410
1411 (b) When PCRE was configured to use the heap rather than the stack for
1412 recursion during matching, it was not correctly preserving the data for
1413 the other case of a UTF-8 character when checking ahead for a match
1414 while processing a minimizing repeat. If the check also involved
1415 matching a wide character, but failed, corruption could cause an
1416 erroneous result when trying to check for a repeat of the original
1417 character.
1418
1419 13. Some tidying changes to the testing mechanism:
1420
1421 (a) The RunTest script now detects the internal link size and whether there
1422 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1423 values substituted by "configure". (The RunGrepTest script already did
1424 this for UTF-8.) The configure.ac script no longer substitutes the
1425 relevant variables.
1426
1427 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1428 with length and offset values. This means that the output is different
1429 for different internal link sizes. Test 2 is skipped for link sizes
1430 other than 2 because of this, bypassing the problem. Unfortunately,
1431 there was also a test in test 3 (the locale tests) that used /B and
1432 failed for link sizes other than 2. Rather than cut the whole test out,
1433 I have added a new /Z option to pcretest that replaces the length and
1434 offset values with spaces. This is now used to make test 3 independent
1435 of link size. (Test 2 will be tidied up later.)
1436
1437 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1438 segmentation fault instead of returning the appropriate error message.
1439
1440 15. In multiline mode when the newline sequence was set to "any", the pattern
1441 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1442 This doesn't seem right; it now treats the CRLF combination as the line
1443 ending, and so does not match in that case. It's only a pattern such as ^$
1444 that would hit this one: something like ^ABC$ would have failed after \r
1445 and then tried again after \r\n.
1446
1447 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1448 in an attempt to make files that differ only in their line terminators
1449 compare equal. This works on Linux.
1450
1451 17. Under certain error circumstances pcregrep might try to free random memory
1452 as it exited. This is now fixed, thanks to valgrind.
1453
1454 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1455 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1456 was because its rules for how to advance for /g after matching an empty
1457 string at the end of a line did not allow for this case. They now check for
1458 it specially.
1459
1460 20. pcretest is supposed to handle patterns and data of any length, by
1461 extending its buffers when necessary. It was getting this wrong when the
1462 buffer for a data line had to be extended.
1463
1464 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1465 CRLF as a newline sequence.
1466
1467 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1468 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1469 I have nevertheless tidied it up.
1470
1471 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1472
1473 24. Added a man page for pcre-config.
1474
1475
1476 Version 7.0 19-Dec-06
1477 ---------------------
1478
1479 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1480 moving to gcc 4.1.1.
1481
1482 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1483 sys/time.h, which is documented as needed for this function. It doesn't
1484 seem to matter on Linux, but it showed up on some releases of OS X.
1485
1486 3. It seems that there are systems where bytes whose values are greater than
1487 127 match isprint() in the "C" locale. The "C" locale should be the
1488 default when a C program starts up. In most systems, only ASCII printing
1489 characters match isprint(). This difference caused the output from pcretest
1490 to vary, making some of the tests fail. I have changed pcretest so that:
1491
1492 (a) When it is outputting text in the compiled version of a pattern, bytes
1493 other than 32-126 are always shown as hex escapes.
1494
1495 (b) When it is outputting text that is a matched part of a subject string,
1496 it does the same, unless a different locale has been set for the match
1497 (using the /L modifier). In this case, it uses isprint() to decide.
1498
1499 4. Fixed a major bug that caused incorrect computation of the amount of memory
1500 required for a compiled pattern when options that changed within the
1501 pattern affected the logic of the preliminary scan that determines the
1502 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1503 that the computed length was too small. The symptoms of this bug were
1504 either the PCRE error "internal error: code overflow" from pcre_compile(),
1505 or a glibc crash with a message such as "pcretest: free(): invalid next
1506 size (fast)". Examples of patterns that provoked this bug (shown in
1507 pcretest format) are:
1508
1509 /(?-x: )/x
1510 /(?x)(?-x: \s*#\s*)/
1511 /((?i)[\x{c0}])/8
1512 /(?i:[\x{c0}])/8
1513
1514 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1515 is now done differently.
1516
1517 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1518 wrapper classes; (b) implement a new function in the C++ scanner that is
1519 more efficient than the old way of doing things because it avoids levels of
1520 recursion in the regex matching; (c) add a paragraph to the documentation
1521 for the FullMatch() function.
1522
1523 6. The escape sequence \n was being treated as whatever was defined as
1524 "newline". Not only was this contrary to the documentation, which states
1525 that \n is character 10 (hex 0A), but it also went horribly wrong when
1526 "newline" was defined as CRLF. This has been fixed.
1527
1528 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1529 was being set to -1 for the "end of line" case (supposedly a value that no
1530 character can have). Though this value is never used (the check for end of
1531 line is "zero bytes in current character"), it caused compiler complaints.
1532 I've changed it to 0xffffffff.
1533
1534 8. In pcre_version.c, the version string was being built by a sequence of
1535 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1536 string (as it is for production releases) called a macro with an empty
1537 argument. The C standard says the result of this is undefined. The gcc
1538 compiler treats it as an empty string (which was what was wanted) but it is
1539 reported that Visual C gives an error. The source has been hacked around to
1540 avoid this problem.
1541
1542 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1543 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1544 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1545 of them did).
1546
1547 10. Originally, pcretest opened its input and output without "b"; then I was
1548 told that "b" was needed in some environments, so it was added for release
1549 5.0 to both the input and output. (It makes no difference on Unix-like
1550 systems.) Later I was told that it is wrong for the input on Windows. I've
1551 now abstracted the modes into two macros, to make it easier to fiddle with
1552 them, and removed "b" from the input mode under Windows.
1553
1554 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1555
1556 12. Added -help and --help to pcretest as an official way of being reminded
1557 of the options.
1558
1559 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1560 and pcrecpp.cc because they annoy compilers at high warning levels.
1561
1562 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1563
1564 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1565 scripts are not C programs :-) and which was not noticed because it works
1566 on Linux.
1567
1568 16. pcretest is supposed to handle any length of pattern and data line (as one
1569 line or as a continued sequence of lines) by extending its input buffer if
1570 necessary. This feature was broken for very long pattern lines, leading to
1571 a string of junk being passed to pcre_compile() if the pattern was longer
1572 than about 50K.
1573
1574 17. I have done a major re-factoring of the way pcre_compile() computes the
1575 amount of memory needed for a compiled pattern. Previously, there was code
1576 that made a preliminary scan of the pattern in order to do this. That was
1577 OK when PCRE was new, but as the facilities have expanded, it has become
1578 harder and harder to keep it in step with the real compile phase, and there
1579 have been a number of bugs (see for example, 4 above). I have now found a
1580 cunning way of running the real compile function in a "fake" mode that
1581 enables it to compute how much memory it would need, while actually only
1582 ever using a few hundred bytes of working memory and without too many
1583 tests of the mode. This should make future maintenance and development
1584 easier. A side effect of this work is that the limit of 200 on the nesting
1585 depth of parentheses has been removed (though this was never a serious
1586 limitation, I suspect). However, there is a downside: pcre_compile() now
1587 runs more slowly than before (30% or more, depending on the pattern). I
1588 hope this isn't a big issue. There is no effect on runtime performance.
1589
1590 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1591 newline (only possible for the last line of a file) and it was a
1592 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1593
1594 19. Added additional timing features to pcretest. (1) The -tm option now times
1595 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1596 separate command line item, by a number that specifies the number of
1597 repeats to use when timing. The default is 50000; this gives better
1598 precision, but takes uncomfortably long for very large patterns.
1599
1600 20. Extended pcre_study() to be more clever in cases where a branch of a
1601 subpattern has no definite first character. For example, (a*|b*)[cd] would
1602 previously give no result from pcre_study(). Now it recognizes that the
1603 first character must be a, b, c, or d.
1604
1605 21. There was an incorrect error "recursive call could loop indefinitely" if
1606 a subpattern (or the entire pattern) that was being tested for matching an
1607 empty string contained only one non-empty item after a nested subpattern.
1608 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1609 incorrectly, because the \d was being skipped in the check.
1610
1611 22. The pcretest program now has a new pattern option /B and a command line
1612 option -b, which is equivalent to adding /B to every pattern. This causes
1613 it to show the compiled bytecode, without the additional information that
1614 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1615 is the same as /B/I).
1616
1617 23. A new optimization is now able automatically to treat some sequences such
1618 as a*b as a*+b. More specifically, if something simple (such as a character
1619 or a simple class like \d) has an unlimited quantifier, and is followed by
1620 something that cannot possibly match the quantified thing, the quantifier
1621 is automatically "possessified".
1622
1623 24. A recursive reference to a subpattern whose number was greater than 39
1624 went wrong under certain circumstances in UTF-8 mode. This bug could also
1625 have affected the operation of pcre_study().
1626
1627 25. Realized that a little bit of performance could be had by replacing
1628 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1629
1630 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1631
1632 27. Possessive quantifiers such as a++ were previously implemented by turning
1633 them into atomic groups such as ($>a+). Now they have their own opcodes,
1634 which improves performance. This includes the automatically created ones
1635 from 23 above.
1636
1637 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1638 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1639 the first matched character to be a colon. This applied both to named and
1640 numbered groups.
1641
1642 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1643
1644 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1645 building PCRE on Windows easier, so I have included it in the distribution.
1646
1647 31. There is now a check in pcretest against a ridiculously large number being
1648 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1649 loop, the loop is abandoned.
1650
1651 32. Forward references to subpatterns in conditions such as (?(2)...) where
1652 subpattern 2 is defined later cause pcre_compile() to search forwards in
1653 the pattern for the relevant set of parentheses. This search went wrong
1654 when there were unescaped parentheses in a character class, parentheses
1655 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1656
1657 33. "Subroutine" calls and backreferences were previously restricted to
1658 referencing subpatterns earlier in the regex. This restriction has now
1659 been removed.
1660
1661 34. Added a number of extra features that are going to be in Perl 5.10. On the
1662 whole, these are just syntactic alternatives for features that PCRE had
1663 previously implemented using the Python syntax or my own invention. The
1664 other formats are all retained for compatibility.
1665
1666 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1667 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1668 also .NET compatible.
1669
1670 (b) A recursion or subroutine call to a named group can now be defined as
1671 (?&name) as well as (?P>name).
1672
1673 (c) A backreference to a named group can now be defined as \k<name> or
1674 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1675 5.10, are also .NET compatible.
1676
1677 (d) A conditional reference to a named group can now use the syntax
1678 (?(<name>) or (?('name') as well as (?(name).
1679
1680 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1681 groups (named and numbered) that are never evaluated inline, but can be
1682 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1683 is always false. There may be only one alternative in such a group.
1684
1685 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1686 as the simple (?(R). The condition is true only if the most recent
1687 recursion is that of the given number or name. It does not search out
1688 through the entire recursion stack.
1689
1690 (g) The escape \gN or \g{N} has been added, where N is a positive or
1691 negative number, specifying an absolute or relative reference.
1692
1693 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1694 some "unreachable code" warnings.
1695
1696 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1697 things, this adds five new scripts.
1698
1699 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1700 There were also incompatibilities regarding the handling of \Q..\E inside
1701 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1702 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1703
1704 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1705 matches an empty string, and forcibly breaks the loop. There were bugs in
1706 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1707 against aaaa the result was just "a" rather than "aaaa", for example. Two
1708 separate and independent bugs (that affected different cases) have been
1709 fixed.
1710
1711 39. Refactored the code to abolish the use of different opcodes for small
1712 capturing bracket numbers. This is a tidy that I avoided doing when I
1713 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1714 The new approach is not only tidier, it makes it possible to reduce the
1715 memory needed to fix the previous bug (38).
1716
1717 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1718 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1719 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1720 mode.
1721
1722 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1723 report.
1724
1725 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1726 copy construction and assignment in the C++ wrapper.
1727
1728 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1729 couple of bugs that could have given wrong results in the "--newline=crlf"
1730 case.
1731
1732 44. Added a number of casts and did some reorganization of signed/unsigned int
1733 variables following suggestions from Dair Grant. Also renamed the variable
1734 "this" as "item" because it is a C++ keyword.
1735
1736 45. Arranged for dftables to add
1737
1738 #include "pcre_internal.h"
1739
1740 to pcre_chartables.c because without it, gcc 4.x may remove the array
1741 definition from the final binary if PCRE is built into a static library and
1742 dead code stripping is activated.
1743
1744 46. For an unanchored pattern, if a match attempt fails at the start of a
1745 newline sequence, and the newline setting is CRLF or ANY, and the next two
1746 characters are CRLF, advance by two characters instead of one.
1747
1748
1749 Version 6.7 04-Jul-06
1750 ---------------------
1751
1752 1. In order to handle tests when input lines are enormously long, pcretest has
1753 been re-factored so that it automatically extends its buffers when
1754 necessary. The code is crude, but this _is_ just a test program. The
1755 default size has been increased from 32K to 50K.
1756
1757 2. The code in pcre_study() was using the value of the re argument before
1758 testing it for NULL. (Of course, in any sensible call of the function, it
1759 won't be NULL.)
1760
1761 3. The memmove() emulation function in pcre_internal.h, which is used on
1762 systems that lack both memmove() and bcopy() - that is, hardly ever -
1763 was missing a "static" storage class specifier.
1764
1765 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1766 containing an extended class (one that cannot be represented by a bitmap
1767 because it contains high-valued characters or Unicode property items, e.g.
1768 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1769 pattern, but PCRE should not loop if you do not (it no longer does).
1770 [Detail: two cases were found: (a) a repeated subpattern containing an
1771 extended class; (b) a recursive reference to a subpattern that followed a
1772 previous extended class. It wasn't skipping over the extended class
1773 correctly when UTF-8 mode was not set.]
1774
1775 5. A negated single-character class was not being recognized as fixed-length
1776 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1777 compile error "lookbehind assertion is not fixed length".
1778
1779 6. The RunPerlTest auxiliary script was showing an unexpected difference
1780 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1781 write a Perl script that can interpret lines of an input file either as
1782 byte characters or as UTF-8, which is what "perltest" was being required to
1783 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1784 can't do is switch easily at run time between having the "use utf8;" pragma
1785 or not. In the end, I fudged it by using the RunPerlTest script to insert
1786 "use utf8;" explicitly for the UTF-8 tests.
1787
1788 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1789 the end of the subject string, contrary to the documentation and to what
1790 Perl does. This was true of both matching functions. Now it matches only at
1791 the start of the subject and immediately after *internal* newlines.
1792
1793 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1794 a pointer to an int instead of a pointer to an unsigned long int. This
1795 caused problems on 64-bit systems.
1796
1797 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1798 instance of the 'standard' template library not being so standard".
1799
1800 10. There was no check on the number of named subpatterns nor the maximum
1801 length of a subpattern name. The product of these values is used to compute
1802 the size of the memory block for a compiled pattern. By supplying a very
1803 long subpattern name and a large number of named subpatterns, the size
1804 computation could be caused to overflow. This is now prevented by limiting
1805 the length of names to 32 characters, and the number of named subpatterns
1806 to 10,000.
1807
1808 11. Subpatterns that are repeated with specific counts have to be replicated in
1809 the compiled pattern. The size of memory for this was computed from the
1810 length of the subpattern and the repeat count. The latter is limited to
1811 65535, but there was no limit on the former, meaning that integer overflow
1812 could in principle occur. The compiled length of a repeated subpattern is
1813 now limited to 30,000 bytes in order to prevent this.
1814
1815 12. Added the optional facility to have named substrings with the same name.
1816
1817 13. Added the ability to use a named substring as a condition, using the
1818 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1819 are numbers (not recommended). Forward references are permitted.
1820
1821 14. Added forward references in named backreferences (if you see what I mean).
1822
1823 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1824 pattern could run off the end of the subject. For example, the pattern
1825 "(?s)(.{1,5})"8 did this with the subject "ab".
1826
1827 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1828 PCRE_CASELESS was set when matching characters that were quantified with ?
1829 or *.
1830
1831 17. A character class other than a single negated character that had a minimum
1832 but no maximum quantifier - for example [ab]{6,} - was not handled
1833 correctly by pce_dfa_exec(). It would match only one character.
1834
1835 18. A valid (though odd) pattern that looked like a POSIX character
1836 class but used an invalid character after [ (for example [[,abc,]]) caused
1837 pcre_compile() to give the error "Failed: internal error: code overflow" or
1838 in some cases to crash with a glibc free() error. This could even happen if
1839 the pattern terminated after [[ but there just happened to be a sequence of
1840 letters, a binary zero, and a closing ] in the memory that followed.
1841
1842 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1843 over the years. Originally (before any Unicode support), just the bottom 8
1844 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1845 output from "man perlunicode" includes this:
1846
1847 The regular expression compiler produces polymorphic opcodes. That
1848 is, the pattern adapts to the data and automatically switches to
1849 the Unicode character scheme when presented with Unicode data--or
1850 instead uses a traditional byte scheme when presented with byte
1851 data.
1852
1853 Sadly, a wide octal escape does not cause a switch, and in a string with
1854 no other multibyte characters, these octal escapes are treated as before.
1855 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1856 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1857 Unicode string.
1858
1859 I have not perpetrated such confusion in PCRE. Up till now, it took just
1860 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1861 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1862 translate to the appropriate multibyte character.
1863
1864 29. Applied some refactoring to reduce the number of warnings from Microsoft
1865 and Borland compilers. This has included removing the fudge introduced
1866 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1867 a warning about an unused variable.
1868
1869 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1870 characters since release 4.0, because Perl (from release 5.004) does not.
1871 [Or at least, is documented not to: some releases seem to be in conflict
1872 with the documentation.] However, when a pattern was studied with
1873 pcre_study() and all its branches started with \s, PCRE still included VT
1874 as a possible starting character. Of course, this did no harm; it just
1875 caused an unnecessary match attempt.
1876
1877 22. Removed a now-redundant internal flag bit that recorded the fact that case
1878 dependency changed within the pattern. This was once needed for "required
1879 byte" processing, but is no longer used. This recovers a now-scarce options
1880 bit. Also moved the least significant internal flag bit to the most-
1881 significant bit of the word, which was not previously used (hangover from
1882 the days when it was an int rather than a uint) to free up another bit for
1883 the future.
1884
1885 23. Added support for CRLF line endings as well as CR and LF. As well as the
1886 default being selectable at build time, it can now be changed at runtime
1887 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1888 specify that it is scanning data with non-default line endings.
1889
1890 24. Changed the definition of CXXLINK to make it agree with the definition of
1891 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1892
1893 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1894 recursions. This makes a big different to stack usage for some patterns.
1895
1896 26. If a subpattern containing a named recursion or subroutine reference such
1897 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1898 the space required for the compiled pattern went wrong and gave too small a
1899 value. Depending on the environment, this could lead to "Failed: internal
1900 error: code overflow at offset 49" or "glibc detected double free or
1901 corruption" errors.
1902
1903 27. Applied patches from Google (a) to support the new newline modes and (b) to
1904 advance over multibyte UTF-8 characters in GlobalReplace.
1905
1906 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1907 difference for some implementation of PCRE in some Windows version.
1908
1909 29. Added some extra testing facilities to pcretest:
1910
1911 \q<number> in a data line sets the "match limit" value
1912 \Q<number> in a data line sets the "match recursion limt" value
1913 -S <number> sets the stack size, where <number> is in megabytes
1914
1915 The -S option isn't available for Windows.
1916
1917
1918 Version 6.6 06-Feb-06
1919 ---------------------
1920
1921 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1922 in pcreposix.h. I have copied the definition from pcre.h.
1923
1924 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1925 because pcre.h is no longer a built file.
1926
1927 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1928 not normally included in the compiled code.
1929
1930
1931 Version 6.5 01-Feb-06
1932 ---------------------
1933
1934 1. When using the partial match feature with pcre_dfa_exec(), it was not
1935 anchoring the second and subsequent partial matches at the new starting
1936 point. This could lead to incorrect results. For example, with the pattern
1937 /1234/, partially matching against "123" and then "a4" gave a match.
1938
1939 2. Changes to pcregrep:
1940
1941 (a) All non-match returns from pcre_exec() were being treated as failures
1942 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1943 error message is output. Some extra information is given for the
1944 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1945 probably the only errors that are likely to be caused by users (by
1946 specifying a regex that has nested indefinite repeats, for instance).
1947 If there are more than 20 of these errors, pcregrep is abandoned.
1948
1949 (b) A binary zero was treated as data while matching, but terminated the
1950 output line if it was written out. This has been fixed: binary zeroes
1951 are now no different to any other data bytes.
1952
1953 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1954 used to set a locale for matching. The --locale=xxxx long option has
1955 been added (no short equivalent) to specify a locale explicitly on the
1956 pcregrep command, overriding the environment variables.
1957
1958 (d) When -B was used with -n, some line numbers in the output were one less
1959 than they should have been.
1960
1961 (e) Added the -o (--only-matching) option.
1962
1963 (f) If -A or -C was used with -c (count only), some lines of context were
1964 accidentally printed for the final match.
1965
1966 (g) Added the -H (--with-filename) option.
1967
1968 (h) The combination of options -rh failed to suppress file names for files
1969 that were found from directory arguments.
1970
1971 (i) Added the -D (--devices) and -d (--directories) options.
1972
1973 (j) Added the -F (--fixed-strings) option.
1974
1975 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1976
1977 (l) Added the --colo(u)r option.
1978
1979 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1980 is not present by default.
1981
1982 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1983 items such as (?R) or (?1), when the recursion could match a number of
1984 alternatives. If it matched one of the alternatives, but subsequently,
1985 outside the recursion, there was a failure, the code tried to back up into
1986 the recursion. However, because of the way PCRE is implemented, this is not
1987 possible, and the result was an incorrect result from the match.
1988
1989 In order to prevent this happening, the specification of recursion has
1990 been changed so that all such subpatterns are automatically treated as
1991 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1992
1993 4. I had overlooked the fact that, in some locales, there are characters for
1994 which isalpha() is true but neither isupper() nor islower() are true. In
1995 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1996 and ordfeminine) are like this. This affected the treatment of \w and \W
1997 when they appeared in character classes, but not when they appeared outside
1998 a character class. The bit map for "word" characters is now created
1999 separately from the results of isalnum() instead of just taking it from the
2000 upper, lower, and digit maps. (Plus the underscore character, of course.)
2001
2002 5. The above bug also affected the handling of POSIX character classes such as
2003 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2004 permanent tables. Instead, the bit maps for such a class were previously
2005 created as the appropriate unions of the upper, lower, and digit bitmaps.
2006 Now they are created by subtraction from the [[:word:]] class, which has
2007 its own bitmap.
2008
2009 6. The [[:blank:]] character class matches horizontal, but not vertical space.
2010 It is created by subtracting the vertical space characters (\x09, \x0a,
2011 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2012 subtraction was done in the overall bitmap for a character class, meaning
2013 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2014 be recognized. This bug has been fixed.
2015
2016 7. Patches from the folks at Google:
2017
2018 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2019 real life, but is still worth protecting against".
2020
2021 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2022 regular expressions".
2023
2024 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2025 have it.
2026
2027 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2028 "configure" and the latter not, in order to fix a problem somebody had
2029 with compiling the Arg class on HP-UX.
2030
2031 (e) Improve the error-handling of the C++ wrapper a little bit.
2032
2033 (f) New tests for checking recursion limiting.
2034
2035 8. The pcre_memmove() function, which is used only if the environment does not
2036 have a standard memmove() function (and is therefore rarely compiled),
2037 contained two bugs: (a) use of int instead of size_t, and (b) it was not
2038 returning a result (though PCRE never actually uses the result).
2039
2040 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2041 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2042 returned instead of calling malloc() with an overflowing number that would
2043 most likely cause subsequent chaos.
2044
2045 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2046
2047 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2048 with this option is matched, the nmatch and pmatch options of regexec() are
2049 ignored.
2050
2051 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2052 provided in case anyone wants to the the POSIX interface with UTF-8
2053 strings.
2054
2055 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2056 C++ linking (needed for some HP-UX environments).
2057
2058 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2059 (unused parameter) and in the pcre_printint() function (omitted "default"
2060 switch label when the default is to do nothing).
2061
2062 15. Added some code to make it possible, when PCRE is compiled as a C++
2063 library, to replace subject pointers for pcre_exec() with a smart pointer
2064 class, thus making it possible to process discontinuous strings.
2065
2066 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2067 much the same function. They were added by different people who were trying
2068 to make PCRE easy to compile on non-Unix systems. It has been suggested
2069 that PCRE_EXPORT be abolished now that there is more automatic apparatus
2070 for compiling on Windows systems. I have therefore replaced it with
2071 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2072 defaults to "extern" for C or "extern C" for C++, which works fine on
2073 Unix-like systems. It is now possible to override the value of PCRE_DATA_
2074 SCOPE with something explicit in config.h. In addition:
2075
2076 (a) pcreposix.h still had just "extern" instead of either of these macros;
2077 I have replaced it with PCRE_DATA_SCOPE.
2078
2079 (b) Functions such as _pcre_xclass(), which are internal to the library,
2080 but external in the C sense, all had PCRE_EXPORT in their definitions.
2081 This is apparently wrong for the Windows case, so I have removed it.
2082 (It makes no difference on Unix-like systems.)
2083
2084 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2085 of recursive calls to match(). This is different to MATCH_LIMIT because
2086 that limits the total number of calls to match(), not all of which increase
2087 the depth of recursion. Limiting the recursion depth limits the amount of
2088 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2089 when PCRE is compiled, and changed at run time. A patch from Google adds
2090 this functionality to the C++ interface.
2091
2092 18. Changes to the handling of Unicode character properties:
2093
2094 (a) Updated the table to Unicode 4.1.0.
2095
2096 (b) Recognize characters that are not in the table as "Cn" (undefined).
2097
2098 (c) I revised the way the table is implemented to a much improved format
2099 which includes recognition of ranges. It now supports the ranges that
2100 are defined in UnicodeData.txt, and it also amalgamates other
2101 characters into ranges. This has reduced the number of entries in the
2102 table from around 16,000 to around 3,000, thus reducing its size
2103 considerably. I realized I did not need to use a tree structure after
2104 all - a binary chop search is just as efficient. Having reduced the
2105 number of entries, I extended their size from 6 bytes to 8 bytes to
2106 allow for more data.
2107
2108 (d) Added support for Unicode script names via properties such as \p{Han}.
2109
2110 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2111 matching that character.
2112
2113 20. When matching a repeated Unicode property with a minimum greater than zero,
2114 (for example \pL{2,}), PCRE could look past the end of the subject if it
2115 reached it while seeking the minimum number of characters. This could
2116 happen only if some of the characters were more than one byte long, because
2117 there is a check for at least the minimum number of bytes.
2118
2119 21. Refactored the implementation of \p and \P so as to be more general, to
2120 allow for more different types of property in future. This has changed the
2121 compiled form incompatibly. Anybody with saved compiled patterns that use
2122 \p or \P will have to recompile them.
2123
2124 22. Added "Any" and "L&" to the supported property types.
2125
2126 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2127 but give a compile time error if the value is greater than 0xff.
2128
2129 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2130 accidentally not being installed or uninstalled.
2131
2132 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2133 made were to insert the current release number. This seemed silly, because
2134 it made things harder for people building PCRE on systems that don't run
2135 "configure". I have turned pcre.h into a distributed file, no longer built
2136 by "configure", with the version identification directly included. There is
2137 no longer a pcre.h.in file.
2138
2139 However, this change necessitated a change to the pcre-config script as
2140 well. It is built from pcre-config.in, and one of the substitutions was the
2141 release number. I have updated configure.ac so that ./configure now finds
2142 the release number by grepping pcre.h.
2143
2144 26. Added the ability to run the tests under valgrind.
2145
2146
2147 Version 6.4 05-Sep-05
2148 ---------------------
2149
2150 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2151 "--" to be printed when multiple files were scanned, even when none of the
2152 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2153 consider it to be a bug, and have restored the previous behaviour.
2154
2155 2. A couple of code tidies to get rid of compiler warnings.
2156
2157 3. The pcretest program used to cheat by referring to symbols in the library
2158 whose names begin with _pcre_. These are internal symbols that are not
2159 really supposed to be visible externally, and in some environments it is
2160 possible to suppress them. The cheating is now confined to including
2161 certain files from the library's source, which is a bit cleaner.
2162
2163 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2164 file's purpose clearer.
2165
2166 5. Reorganized pcre_ucp_findchar().
2167
2168
2169 Version 6.3 15-Aug-05
2170 ---------------------
2171
2172 1. The file libpcre.pc.in did not have general read permission in the tarball.
2173
2174 2. There were some problems when building without C++ support:
2175
2176 (a) If C++ support was not built, "make install" and "make test" still
2177 tried to test it.
2178
2179 (b) There were problems when the value of CXX was explicitly set. Some
2180 changes have been made to try to fix these, and ...
2181
2182 (c) --disable-cpp can now be used to explicitly disable C++ support.
2183
2184 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2185 backslash in a target when C++ was disabled. This confuses some
2186 versions of "make", apparently. Using an intermediate variable solves
2187 this. (Same for CPP_LOBJ.)
2188
2189 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2190 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2191 necessary on certain architectures.
2192
2193 4. Added a setting of -export-symbols-regex to the link command to remove
2194 those symbols that are exported in the C sense, but actually are local
2195 within the library, and not documented. Their names all begin with
2196 "_pcre_". This is not a perfect job, because (a) we have to except some
2197 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2198 available (and never for static libraries). I have made a note to try to
2199 find a way round (a) in the future.
2200
2201
2202 Version 6.2 01-Aug-05
2203 ---------------------
2204
2205 1. There was no test for integer overflow of quantifier values. A construction
2206 such as {1111111111111111} would give undefined results. What is worse, if
2207 a minimum quantifier for a parenthesized subpattern overflowed and became
2208 negative, the calculation of the memory size went wrong. This could have
2209 led to memory overwriting.
2210
2211 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2212
2213 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2214 operating environments where this matters.
2215
2216 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2217 PCRE options from within the C++ wrapper.
2218
2219 5. Named capturing subpatterns were not being correctly counted when a pattern
2220 was compiled. This caused two problems: (a) If there were more than 100
2221 such subpatterns, the calculation of the memory needed for the whole
2222 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2223 back references of the form \12, where the number was greater than 9, were
2224 not recognized as back references, even though there were sufficient
2225 previous subpatterns.
2226
2227 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2228 versions of gcc, e.g. 2.95.4.
2229
2230
2231 Version 6.1 21-Jun-05
2232 ---------------------
2233
2234 1. There was one reference to the variable "posix" in pcretest.c that was not
2235 surrounded by "#if !defined NOPOSIX".
2236
2237 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2238 the cross-check on the old pcre_info() function, for the benefit of the
2239 cut-down version of PCRE that is currently imported into Exim.
2240
2241 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2242 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2243 patterns that start (?i) but I don't think that matters. The use of (?i) is
2244 just an example; this all applies to the other options as well.
2245
2246 4. Since libtool seems to echo the compile commands it is issuing, the output
2247 from "make" can be reduced a bit by putting "@" in front of each libtool
2248 compile command.
2249
2250 5. Patch from the folks at Google for configure.in to be a bit more thorough
2251 in checking for a suitable C++ installation before trying to compile the
2252 C++ stuff. This should fix a reported problem when a compiler was present,
2253 but no suitable headers.
2254
2255 6. The man pages all had just "PCRE" as their title. I have changed them to
2256 be the relevant file name. I have also arranged that these names are
2257 retained in the file doc/pcre.txt, which is a concatenation in text format
2258 of all the man pages except the little individual ones for each function.
2259
2260 7. The NON-UNIX-USE file had not been updated for the different set of source
2261 files that come with release 6. I also added a few comments about the C++
2262 wrapper.
2263
2264
2265 Version 6.0 07-Jun-05
2266 ---------------------
2267
2268 1. Some minor internal re-organization to help with my DFA experiments.
2269
2270 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2271 didn't matter for the library itself when fully configured, but did matter
2272 when compiling without UCP support, or within Exim, where the ucp files are
2273 not imported.
2274
2275 3. Refactoring of the library code to split up the various functions into
2276 different source modules. The addition of the new DFA matching code (see
2277 below) to a single monolithic source would have made it really too
2278 unwieldy, quite apart from causing all the code to be include in a
2279 statically linked application, when only some functions are used. This is
2280 relevant even without the DFA addition now that patterns can be compiled in
2281 one application and matched in another.
2282
2283 The downside of splitting up is that there have to be some external
2284 functions and data tables that are used internally in different modules of
2285 the library but which are not part of the API. These have all had their
2286 names changed to start with "_pcre_" so that they are unlikely to clash
2287 with other external names.
2288
2289 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2290 a different (DFA) algorithm. Although it is slower than the original
2291 function, it does have some advantages for certain types of matching
2292 problem.
2293
2294 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2295 including restarting after a partial match.
2296
2297 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2298 defined when compiling for Windows was sent to me. I have put it into the
2299 code, though I have no means of testing or verifying it.
2300
2301 7. Added the pcre_refcount() auxiliary function.
2302
2303 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2304 match before or at the first newline in the subject string. In pcretest,
2305 the /f option on a pattern can be used to set this.
2306
2307 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2308 would behave wrongly. This has been present in PCRE since release 4.0.
2309
2310 10. A number of changes to the pcregrep command:
2311
2312 (a) Refactored how -x works; insert ^(...)$ instead of setting
2313 PCRE_ANCHORED and checking the length, in preparation for adding
2314 something similar for -w.
2315
2316 (b) Added the -w (match as a word) option.
2317
2318 (c) Refactored the way lines are read and buffered so as to have more
2319 than one at a time available.
2320
2321 (d) Implemented a pcregrep test script.
2322
2323 (e) Added the -M (multiline match) option. This allows patterns to match
2324 over several lines of the subject. The buffering ensures that at least
2325 8K, or the rest of the document (whichever is the shorter) is available
2326 for matching (and similarly the previous 8K for lookbehind assertions).
2327
2328 (f) Changed the --help output so that it now says
2329
2330 -w, --word-regex(p)
2331
2332 instead of two lines, one with "regex" and the other with "regexp"
2333 because that confused at least one person since the short forms are the
2334 same. (This required a bit of code, as the output is generated
2335 automatically from a table. It wasn't just a text change.)
2336
2337 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2338 option but starts with a hyphen. Could be a pattern or a path name
2339 starting with a hyphen, for instance.
2340
2341 (h) "-" can be given as a file name to represent stdin.
2342
2343 (i) When file names are being printed, "(standard input)" is used for
2344 the standard input, for compatibility with GNU grep. Previously
2345 "<stdin>" was used.
2346
2347 (j) The option --label=xxx can be used to supply a name to be used for
2348 stdin when file names are being printed. There is no short form.
2349
2350 (k) Re-factored the options decoding logic because we are going to add
2351 two more options that take data. Such options can now be given in four
2352 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2353
2354 (l) Added the -A, -B, and -C options for requesting that lines of context
2355 around matches be printed.
2356
2357 (m) Added the -L option to print the names of files that do not contain
2358 any matching lines, that is, the complement of -l.
2359
2360 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2361 continue to scan other files.
2362
2363 (o) The -s option was incorrectly implemented. For compatibility with other
2364 greps, it now suppresses the error message for a non-existent or non-
2365 accessible file (but not the return code). There is a new option called
2366 -q that suppresses the output of matching lines, which was what -s was
2367 previously doing.
2368
2369 (p) Added --include and --exclude options to specify files for inclusion
2370 and exclusion when recursing.
2371
2372 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2373 Hopefully, it now does.
2374
2375 12. Missing cast in pcre_study().
2376
2377 13. Added an "uninstall" target to the makefile.
2378
2379 14. Replaced "extern" in the function prototypes in Makefile.in with
2380 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2381 world, but is set differently for Windows.
2382
2383 15. Added a second compiling function called pcre_compile2(). The only
2384 difference is that it has an extra argument, which is a pointer to an
2385 integer error code. When there is a compile-time failure, this is set
2386 non-zero, in addition to the error test pointer being set to point to an
2387 error message. The new argument may be NULL if no error number is required
2388 (but then you may as well call pcre_compile(), which is now just a
2389 wrapper). This facility is provided because some applications need a
2390 numeric error indication, but it has also enabled me to tidy up the way
2391 compile-time errors are handled in the POSIX wrapper.
2392
2393 16. Added VPATH=.libs to the makefile; this should help when building with one
2394 prefix path and installing with another. (Or so I'm told by someone who
2395 knows more about this stuff than I do.)
2396
2397 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2398 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2399 match everything, including newlines. This is not POSIX-compatible, but
2400 somebody wanted the feature. From pcretest it can be activated by using
2401 both the P and the s flags.
2402
2403 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2404
2405 19. libpcre.pc was being incorrectly installed as executable.
2406
2407 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2408 it now also looks for '\r' so that it will work unmodified on Windows.
2409
2410 21. Added Google's contributed C++ wrapper to the distribution.
2411
2412 22. Added some untidy missing memory free() calls in pcretest, to keep
2413 Electric Fence happy when testing.
2414
2415
2416
2417 Version 5.0 13-Sep-04
2418 ---------------------
2419
2420 1. Internal change: literal characters are no longer packed up into items
2421 containing multiple characters in a single byte-string. Each character
2422 is now matched using a separate opcode. However, there may be more than one
2423 byte in the character in UTF-8 mode.
2424
2425 2. The pcre_callout_block structure has two new fields: pattern_position and
2426 next_item_length. These contain the offset in the pattern to the next match
2427 item, and its length, respectively.
2428
2429 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2430 insertion of callouts before each pattern item. Added the /C option to
2431 pcretest to make use of this.
2432
2433 4. On the advice of a Windows user, the lines
2434
2435 #if defined(_WIN32) || defined(WIN32)
2436 _setmode( _fileno( stdout ), 0x8000 );
2437 #endif /* defined(_WIN32) || defined(WIN32) */
2438
2439 have been added to the source of pcretest. This apparently does useful
2440 magic in relation to line terminators.
2441
2442 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2443 for the benefit of those environments where the "b" makes a difference.
2444
2445 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2446 to know about it. I have put a hack into configure.in that adds in code
2447 to set GCC=yes if CC=icc. This seems to end up at a point in the
2448 generated configure script that is early enough to affect the setting of
2449 compiler options, which is what is needed, but I have no means of testing
2450 whether it really works. (The user who reported this had patched the
2451 generated configure script, which of course I cannot do.)
2452
2453 LATER: After change 22 below (new libtool files), the configure script
2454 seems to know about icc (and also ecc). Therefore, I have commented out
2455 this hack in configure.in.
2456
2457 7. Added support for pkg-config (2 patches were sent in).
2458
2459 8. Negated POSIX character classes that used a combination of internal tables
2460 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2461 [[:^ascii]]. Typically, they would match almost any characters. The other
2462 POSIX classes were not broken in this way.
2463
2464 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2465 to find the match, as PCRE was deluded into thinking that the match had to
2466 start at the start point or following a newline. The same bug applied to
2467 patterns with negative forward assertions or any backward assertions
2468 preceding ".*" at the start, unless the pattern required a fixed first
2469 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2470
2471 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2472 starting at the last subject character, bytes beyond the end of the subject
2473 string were read.
2474
2475 11. Renamed the variable "class" as "classbits" to make life easier for C++
2476 users. (Previously there was a macro definition, but it apparently wasn't
2477 enough.)
2478
2479 12. Added the new field "tables" to the extra data so that tables can be passed
2480 in at exec time, or the internal tables can be re-selected. This allows
2481 a compiled regex to be saved and re-used at a later time by a different
2482 program that might have everything at different addresses.
2483
2484 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2485 -R library as well as a -L library.
2486
2487 14. The debugging options of pcretest (-d on the command line or D on a
2488 pattern) showed incorrect output for anything following an extended class
2489 that contained multibyte characters and which was followed by a quantifier.
2490
2491 15. Added optional support for general category Unicode character properties
2492 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2493 support. It adds about 90K to the size of the library. The meanings of the
2494 inbuilt class escapes such as \d and \s have NOT been changed.
2495
2496 16. Updated pcredemo.c to include calls to free() to release the memory for the
2497 compiled pattern.
2498
2499 17. The generated file chartables.c was being created in the source directory
2500 instead of in the building directory. This caused the build to fail if the
2501 source directory was different from the building directory, and was
2502 read-only.
2503
2504 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2505 file. No doubt somebody will tell me if they don't make sense... Also added
2506 Dan Mooney's comments about building on OpenVMS.
2507
2508 19. Added support for partial matching via the PCRE_PARTIAL option for
2509 pcre_exec() and the \P data escape in pcretest.
2510
2511 20. Extended pcretest with 3 new pattern features:
2512
2513 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2514 write the compiled pattern to the file whose name is "rest-of-line".
2515 This is a straight binary dump of the data, with the saved pointer to
2516 the character tables forced to be NULL. The study data, if any, is
2517 written too. After writing, pcretest reads a new pattern.
2518
2519 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2520 compiled pattern from the given file. There must not be any
2521 occurrences of "<" in the file name (pretty unlikely); if there are,
2522 pcretest will instead treat the initial "<" as a pattern delimiter.
2523 After reading in the pattern, pcretest goes on to read data lines as
2524 usual.
2525
2526 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2527 and 16-bit fields in a compiled pattern, to simulate a pattern that
2528 was compiled on a host of opposite endianness.
2529
2530 21. The pcre-exec() function can now cope with patterns that were compiled on
2531 hosts of opposite endianness, with this restriction:
2532
2533 As for any compiled expression that is saved and used later, the tables
2534 pointer field cannot be preserved; the extra_data field in the arguments
2535 to pcre_exec() should be used to pass in a tables address if a value
2536 other than the default internal tables were used at compile time.
2537
2538 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2539 now diagnosed as an error. Previously, most of the time, a negative number
2540 would have been treated as zero, but if in addition "ovector" was passed as
2541 NULL, a crash could occur.
2542
2543 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2544 new versions from the libtool 1.5 distribution (the last one is a copy of
2545 a file called libtool.m4). This seems to have fixed the need to patch
2546 "configure" to support Darwin 1.3 (which I used to do). However, I still
2547 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2548 workstation).
2549
2550 24. Changed the PCRE licence to be the more standard "BSD" licence.
2551
2552
2553 Version 4.5 01-Dec-03
2554 ---------------------
2555
2556 1. There has been some re-arrangement of the code for the match() function so
2557 that it can be compiled in a version that does not call itself recursively.
2558 Instead, it keeps those local variables that need separate instances for
2559 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2560 needs to "recurse". Keeping track of where control must go is done by means
2561 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2562 hide most of the details from the main code, and operates only if
2563 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2564 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2565 operating.
2566
2567 To make it easier for callers to provide specially tailored get/free
2568 functions for this usage, two new functions, pcre_stack_malloc, and
2569 pcre_stack_free, are used. They are always called in strict stacking order,
2570 and the size of block requested is always the same.
2571
2572 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2573 PCRE has been compiled to use the stack or the heap for recursion. The
2574 -C option of pcretest uses this to show which version is compiled.
2575
2576 A new data escape \S, is added to pcretest; it causes the amounts of store
2577 obtained and freed by both kinds of malloc/free at match time to be added
2578 to the output.
2579
2580 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2581 what's available on my current Linux desktop machine.
2582
2583 3. When matching a UTF-8 string, the test for a valid string at the start has
2584 been extended. If start_offset is not zero, PCRE now checks that it points
2585 to a byte that is the start of a UTF-8 character. If not, it returns
2586 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2587 this is necessary because there may be backward assertions in the pattern.
2588 When matching the same subject several times, it may save resources to use
2589 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2590
2591 4. The code for checking the validity of UTF-8 strings has been tightened so
2592 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2593 containing "overlong sequences".
2594
2595 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2596 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2597 should have been "&", but it just so happened that all the cases this let
2598 through by mistake were picked up later in the function.
2599
2600 6. I had used a variable called "isblank" - this is a C99 function, causing
2601 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2602
2603 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2604 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2605 so the version is shown; (c) stop "make test" echoing "./RunTest".
2606
2607 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2608
2609 9. The prototype for memmove() for systems that don't have it was using
2610 size_t, but the inclusion of the header that defines size_t was later. I've
2611 moved the #includes for the C headers earlier to avoid this.
2612
2613 10. Added some adjustments to the code to make it easier to compiler on certain
2614 special systems:
2615
2616 (a) Some "const" qualifiers were missing.
2617 (b) Added the macro EXPORT before all exported functions; by default this
2618 is defined to be empty.
2619 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2620 that it reads its output file name as an argument instead of writing
2621 to the standard output and assuming this can be redirected.
2622
2623 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2624 class containing characters with values greater than 255, PCRE compilation
2625 went into a loop.
2626
2627 12. A recursive reference to a subpattern that was within another subpattern
2628 that had a minimum quantifier of zero caused PCRE to crash. For example,
2629 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2630 recursion. If the recursively-called subpattern itself had a zero repeat,
2631 that was OK.
2632
2633 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2634 buffer into which it was copied (for escape processing) was still set at
2635 1024, so long lines caused crashes.
2636
2637 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2638 "internal error: code overflow...". This applied to any character class
2639 that was followed by a possessive quantifier.
2640
2641 15. Modified the Makefile to add libpcre.la as a prerequisite for
2642 libpcreposix.la because I was told this is needed for a parallel build to
2643 work.
2644
2645 16. If a pattern that contained .* following optional items at the start was
2646 studied, the wrong optimizing data was generated, leading to matching
2647 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2648 matching string must start with a or b or c. The correct conclusion for
2649 this pattern is that a match can start with any character.
2650
2651
2652 Version 4.4 13-Aug-03
2653 ---------------------
2654
2655 1. In UTF-8 mode, a character class containing characters with values between
2656 127 and 255 was not handled correctly if the compiled pattern was studied.
2657 In fixing this, I have also improved the studying algorithm for such
2658 classes (slightly).
2659
2660 2. Three internal functions had redundant arguments passed to them. Removal
2661 might give a very teeny performance improvement.
2662
2663 3. Documentation bug: the value of the capture_top field in a callout is *one
2664 more than* the number of the hightest numbered captured substring.
2665
2666 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2667 in incorrectly linking with a previously installed version. They now link
2668 explicitly with libpcre.la.
2669
2670 5. configure.in no longer needs to recognize Cygwin specially.
2671
2672 6. A problem in pcre.in for Windows platforms is fixed.
2673
2674 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2675 pcretest, it used to include the size of the study block as part of its
2676 output. Unfortunately, the structure contains a field that has a different
2677 size on different hardware architectures. This meant that the tests that
2678 showed this size failed. As the block is currently always of a fixed size,
2679 this information isn't actually particularly useful in pcretest output, so
2680 I have just removed it.
2681
2682 8. Three pre-processor statements accidentally did not start in column 1.
2683 Sadly, there are *still* compilers around that complain, even though
2684 standard C has not required this for well over a decade. Sigh.
2685
2686 9. In pcretest, the code for checking callouts passed small integers in the
2687 callout_data field, which is a void * field. However, some picky compilers
2688 complained about the casts involved for this on 64-bit systems. Now
2689 pcretest passes the address of the small integer instead, which should get
2690 rid of the warnings.
2691
2692 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2693 both compile and run time, and gives an error if an invalid UTF-8 sequence
2694 is found. There is a option for disabling this check in cases where the
2695 string is known to be correct and/or the maximum performance is wanted.
2696
2697 11. In response to a bug report, I changed one line in Makefile.in from
2698
2699 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2700 to
2701 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2702
2703 to look similar to other lines, but I have no way of telling whether this
2704 is the right thing to do, as I do not use Windows. No doubt I'll get told
2705 if it's wrong...
2706
2707
2708 Version 4.3 21-May-03
2709 ---------------------
2710
2711 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2712 Makefile.
2713
2714 2. Some refactoring to improve the quality of the code:
2715
2716 (i) The utf8_table... variables are now declared "const".
2717
2718 (ii) The code for \cx, which used the "case flipping" table to upper case
2719 lower case letters, now just substracts 32. This is ASCII-specific,
2720 but the whole concept of \cx is ASCII-specific, so it seems
2721 reasonable.
2722
2723 (iii) PCRE was using its character types table to recognize decimal and
2724 hexadecimal digits in the pattern. This is silly, because it handles
2725 only 0-9, a-f, and A-F, but the character types table is locale-
2726 specific, which means strange things might happen. A private
2727 table is now used for this - though it costs 256 bytes, a table is
2728 much faster than multiple explicit tests. Of course, the standard
2729 character types table is still used for matching digits in subject
2730 strings against \d.
2731
2732 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2733 ending in _t are). So I've renamed it as ESC_tee.
2734
2735 3. The first argument for regexec() in the POSIX wrapper should have been
2736 defined as "const".
2737
2738 4. Changed pcretest to use malloc() for its buffers so that they can be
2739 Electric Fenced for debugging.
2740
2741 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2742 to read one or more bytes before the start of the subject string. Often this
2743 had no effect on PCRE's behaviour, but in some circumstances it could
2744 provoke a segmentation fault.
2745
2746 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2747 to try to read one or more bytes before the start of the subject string.
2748
2749 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2750 UTF-8 support could misbehave in various ways if the subject string
2751 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2752 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2753 back over UTF-8 characters.)
2754
2755
2756 Version 4.2 14-Apr-03
2757 ---------------------
2758
2759 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2760
2761 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2762 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2763 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2764 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2765 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2766 and BUILD_EXEEXT
2767 Note: automatic setting of the BUILD variables is not yet working
2768 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2769 compile-time but not at link-time
2770 [LINK]: use for linking executables only
2771 make different versions for Windows and non-Windows
2772 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2773 libraries
2774 [LINK_FOR_BUILD]: new variable
2775 [OBJEXT]: use throughout
2776 [EXEEXT]: use throughout
2777 <winshared>: new target
2778 <wininstall>: new target
2779 <dftables.o>: use native compiler
2780 <dftables>: use native linker
2781 <install>: handle Windows platform correctly
2782 <clean>: ditto
2783 <check>: ditto
2784 copy DLL to top builddir before testing
2785
2786 As part of these changes, -no-undefined was removed again. This was reported
2787 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2788 in any case.
2789
2790 3. Some tidies to get rid of compiler warnings:
2791
2792 . In the match_data structure, match_limit was an unsigned long int, whereas
2793 match_call_count was an int. I've made them both unsigned long ints.
2794
2795 . In pcretest the fact that a const uschar * doesn't automatically cast to
2796 a void * provoked a warning.
2797
2798 . Turning on some more compiler warnings threw up some "shadow" variables
2799 and a few more missing casts.
2800
2801 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2802 option, a class that contained a single character with a value between 128
2803 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2804
2805 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2806 option, a class that contained several characters, but with at least one
2807 whose value was between 128 and 255 caused PCRE to crash.
2808
2809
2810 Version 4.1 12-Mar-03
2811 ---------------------
2812
2813 1. Compiling with gcc -pedantic found a couple of places where casts were
2814 needed, and a string in dftables.c that was longer than standard compilers are
2815 required to support.
2816
2817 2. Compiling with Sun's compiler found a few more places where the code could
2818 be tidied up in order to avoid warnings.
2819
2820 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2821 first of these names is deprecated in the latest Autoconf in favour of the name
2822 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2823 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2824 analogy I have changed it to CFLAGS_FOR_BUILD.
2825
2826 4. Added -no-undefined to the linking command in the Makefile, because this is
2827 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2828 linking step for the pcreposix library.
2829
2830 5. PCRE was failing to diagnose the case of two named groups with the same
2831 name.
2832
2833 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2834 literal character that is needed in the subject for a match, and scans along to
2835 ensure that it is present before embarking on the full matching process. This
2836 saves time in cases of nested unlimited repeats that are never going to match.
2837 Problem: the scan can take a lot of time if the subject is very long (e.g.
2838 megabytes), thus penalizing straightforward matches. It is now done only if the
2839 amount of subject to be scanned is less than 1000 bytes.
2840
2841 7. A lesser problem with the same optimization is that it was recording the
2842 first character of an anchored pattern as "needed", thus provoking a search
2843 right along the subject, even when the first match of the pattern was going to
2844 fail. The "needed" character is now not set for anchored patterns, unless it
2845 follows something in the pattern that is of non-fixed length. Thus, it still
2846 fulfils its original purpose of finding quick non-matches in cases of nested
2847 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2848
2849
2850 Version 4.0 17-Feb-03
2851 ---------------------
2852
2853 1. If a comment in an extended regex that started immediately after a meta-item
2854 extended to the end of string, PCRE compiled incorrect data. This could lead to
2855 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2856
2857 2. Moved to autoconf 2.53 and libtool 1.4.2.
2858
2859 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2860 the special perltest8 script is no longer needed - all the tests can be run
2861 from a single perltest script.
2862
2863 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2864 by \s. It has now been removed in PCRE. This means it isn't recognized as
2865 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2866 class [:space:] *does* include VT, thereby creating a mess.
2867
2868 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2869 space and tab.
2870
2871 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2872 its new features into the main test script, reducing the number of scripts.
2873
2874 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2875 were backward compatible, and made the (?i) apply to the whole pattern, as if
2876 /i were given. Now it behaves more logically, and applies the option setting
2877 only to what follows. PCRE has been changed to follow suit. However, if it
2878 finds options settings right at the start of the pattern, it extracts them into
2879 the global options, as before. Thus, they show up in the info data.
2880
2881 8. Added support for the \Q...\E escape sequence. Characters in between are
2882 treated as literals. This is slightly different from Perl in that $ and @ are
2883 also handled as literals inside the quotes. In Perl, they will cause variable
2884 interpolation. Note the following examples:
2885
2886 Pattern PCRE matches Perl matches
2887
2888 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2889 \Qabc\$xyz\E abc\$xyz abc\$xyz
2890 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2891
2892 For compatibility with Perl, \Q...\E sequences are recognized inside character
2893 classes as well as outside them.
2894
2895 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2896 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2897 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2898 signed/unsigned warnings.
2899
2900 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2901 option for pcretest, so I've replaced it by a simple function that does just
2902 that job.
2903
2904 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2905 "pcregrep -".
2906
2907 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2908 Java package. This provides some syntactic sugar for simple cases of what my
2909 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2910 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2911 item, you can use this simplified notation. Note that only makes sense with
2912 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2913 greediness, whatever the setting of the PCRE_UNGREEDY option.
2914
2915 13. A change of greediness default within a pattern was not taking effect at
2916 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2917 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2918 was abstracted outside.
2919
2920 14. PCRE now supports the \G assertion. It is true when the current matching
2921 position is at the start point of the match. This differs from \A when the
2922 starting offset is non-zero. Used with the /g option of pcretest (or similar
2923 code), it works in the same way as it does for Perl's /g option. If all
2924 alternatives of a regex begin with \G, the expression is anchored to the start
2925 match position, and the "anchored" flag is set in the compiled expression.
2926
2927 15. Some bugs concerning the handling of certain option changes within patterns
2928 have been fixed. These applied to options other than (?ims). For example,
2929 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2930 been the other way round. Some of this was related to change 7 above.
2931
2932 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2933 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2934 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2935 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2936
2937 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2938 mode. Unlike ".", it always matches newline, whatever the setting of
2939 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2940 assertions. Perl allows it, but it doesn't (in general) work because it can't
2941 calculate the length of the lookbehind. At least, that's the case for Perl
2942 5.8.0 - I've been told they are going to document that it doesn't work in
2943 future.
2944
2945 18. Added an error diagnosis for escapes that PCRE does not support: these are
2946 \L, \l, \N, \P, \p, \U, \u, and \X.
2947
2948 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2949 reading past the end of the pattern in cases such as /[abcd/.
2950
2951 20. PCRE was getting more memory than necessary for patterns with classes that
2952 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2953
2954 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2955 compiling PCRE for use with Virtual Pascal.
2956
2957 22. Small fix to the Makefile to make it work properly if the build is done
2958 outside the source tree.
2959
2960 23. Added a new extension: a condition to go with recursion. If a conditional
2961 subpattern starts with (?(R) the "true" branch is used if recursion has
2962 happened, whereas the "false" branch is used only at the top level.
2963
2964 24. When there was a very long string of literal characters (over 255 bytes
2965 without UTF support, over 250 bytes with UTF support), the computation of how
2966 much memory was required could be incorrect, leading to segfaults or other
2967 strange effects.
2968
2969 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2970 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2971 there was a subsequent back reference to those brackets. This meant that, for
2972 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2973 possible to check for precisely this case. All we can do is abandon the
2974 optimization if .* occurs inside capturing brackets when there are any back
2975 references whatsoever. (See below for a better fix that came later.)
2976
2977 26. The handling of the optimization for finding the first character of a
2978 non-anchored pattern, and for finding a character that is required later in the
2979 match were failing in some cases. This didn't break the matching; it just
2980 failed to optimize when it could. The way this is done has been re-implemented.
2981
2982 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2983
2984 28. Added a new feature that provides some of the functionality that Perl
2985 provides with (?{...}). The facility is termed a "callout". The way it is done
2986 in PCRE is for the caller to provide an optional function, by setting
2987 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2988 global variable. By default it is unset, which disables all calling out. To get
2989 the function called, the regex must include (?C) at appropriate points. This
2990 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2991 This provides a means of identifying different callout points. When PCRE
2992 reaches such a point in the regex, if pcre_callout has been set, the external
2993 function is called. It is provided with data in a structure called
2994 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2995 matching continues; if it returns a non-zero value, the match at the current
2996 point fails. However, backtracking will occur if possible. [This was changed
2997 later and other features added - see item 49 below.]
2998
2999 29. pcretest is upgraded to test the callout functionality. It provides a
3000 callout function that displays information. By default, it shows the start of
3001 the match and the current position in the text. There are some new data escapes
3002 to vary what happens:
3003
3004 \C+ in addition, show current contents of captured substrings
3005 \C- do not supply a callout function
3006 \C!n return 1 when callout number n is reached
3007 \C!n!m return 1 when callout number n is reached for the mth time
3008
3009 30. If pcregrep was called with the -l option and just a single file name, it
3010 output "<stdin>" if a match was found, instead of the file name.
3011
3012 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3013 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3014 pcre_exec(). This saves a malloc/free per call. The default value of
3015 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3016 when configuring.
3017
3018 32. The default maximum size of a compiled pattern is 64K. There have been a
3019 few cases of people hitting this limit. The code now uses macros to handle the
3020 storing of links as offsets within the compiled pattern. It defaults to 2-byte
3021 links, but this can be changed to 3 or 4 bytes by --with-link-size when
3022 configuring. Tests 2 and 5 work only with 2-byte links because they output
3023 debugging information about compiled patterns.
3024
3025 33. Internal code re-arrangements:
3026
3027 (a) Moved the debugging function for printing out a compiled regex into
3028 its own source file (printint.c) and used #include to pull it into
3029 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3030 separate copies.
3031
3032 (b) Defined the list of op-code names for debugging as a macro in
3033 internal.h so that it is next to the definition of the opcodes.
3034
3035 (c) Defined a table of op-code lengths for simpler skipping along compiled
3036 code. This is again a macro in internal.h so that it is next to the
3037 definition of the opcodes.
3038
3039 34. Added support for recursive calls to individual subpatterns, along the
3040 lines of Robin Houston's patch (but implemented somewhat differently).
3041
3042 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3043 allow it to read and process whole directories in Win32. This code was
3044 contributed by Lionel Fourquaux; it has not been tested by me.
3045
3046 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3047 used to name a group. Names consist of alphanumerics and underscores, and must
3048 be unique. Back references use the syntax (?P=name) and recursive calls use
3049 (?P>name) which is a PCRE extension to the Python extension. Groups still have
3050 numbers. The function pcre_fullinfo() can be used after compilation to extract
3051 a name/number map. There are three relevant calls:
3052
3053 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3054 PCRE_INFO_NAMECOUNT yields the number of entries
3055 PCRE_INFO_NAMETABLE yields a pointer to the map.
3056
3057 The map is a vector of fixed-size entries. The size of each entry depends on
3058 the length of the longest name used. The first two bytes of each entry are the
3059 group number, most significant byte first. There follows the corresponding
3060 name, zero terminated. The names are in alphabetical order.
3061
3062 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3063 case instead of 255. Making it the same both with and without UTF-8 support
3064 means that the same test output works with both.
3065
3066 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3067 calling malloc() with a zero argument.
3068
3069 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3070 optimization. I've improved things by keeping a bitmap of backreferences with
3071 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3072 fact referenced, the optimization can be applied. It is unlikely that a
3073 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3074 the match to follow \n) will appear inside brackets with a number greater than
3075 31, but if it does, any back reference > 31 suppresses the optimization.
3076
3077 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3078 of disabling numbered capturing parentheses. Any opening parenthesis that is
3079 not followed by ? behaves as if it were followed by ?: but named parentheses
3080 can still be used for capturing (and they will acquire numbers in the usual
3081 way).
3082
3083 41. Redesigned the return codes from the match() function into yes/no/error so
3084 that errors can be passed back from deep inside the nested calls. A malloc
3085 failure while inside a recursive subpattern call now causes the
3086 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3087
3088 42. It is now possible to set a limit on the number of times the match()
3089 function is called in a call to pcre_exec(). This facility makes it possible to
3090 limit the amount of recursion and backtracking, though not in a directly
3091 obvious way, because the match() function is used in a number of different
3092 circumstances. The count starts from zero for each position in the subject
3093 string (for non-anchored patterns). The default limit is, for compatibility, a
3094 large number, namely 10 000 000. You can change this in two ways:
3095
3096 (a) When configuring PCRE before making, you can use --with-match-limit=n
3097 to set a default value for the compiled library.
3098
3099 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3100 a different value is set. See 45 below.
3101
3102 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3103
3104 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3105 of things that can be changed at compile time. The first argument specifies
3106 what is wanted and the second points to where the information is to be placed.
3107 The current list of available information is:
3108
3109 PCRE_CONFIG_UTF8
3110
3111 The output is an integer that is set to one if UTF-8 support is available;
3112 otherwise it is set to zero.
3113
3114 PCRE_CONFIG_NEWLINE
3115
3116 The output is an integer that it set to the value of the code that is used for
3117 newline. It is either LF (10) or CR (13).
3118
3119 PCRE_CONFIG_LINK_SIZE
3120
3121 The output is an integer that contains the number of bytes used for internal
3122 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3123
3124 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3125
3126 The output is an integer that contains the threshold above which the POSIX
3127 interface uses malloc() for output vectors. See item 31 above.
3128
3129 PCRE_CONFIG_MATCH_LIMIT
3130
3131 The output is an unsigned integer that contains the default limit of the number
3132 of match() calls in a pcre_exec() execution. See 42 above.
3133
3134 44. pcretest has been upgraded by the addition of the -C option. This causes it
3135 to extract all the available output from the new pcre_config() function, and to
3136 output it. The program then exits immediately.
3137
3138 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3139 order to support additional features. One way would have been to define
3140 pcre_exec2() (for example) with extra arguments, but this would not have been
3141 extensible, and would also have required all calls to the original function to
3142 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3143 is used for passing in "extra" data from pcre_study().
3144
3145 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3146 contains the following fields:
3147
3148 flags a bitmap indicating which of the following fields are set
3149 study_data opaque data from pcre_study()
3150 match_limit a way of specifying a limit on match() calls for a specific
3151 call to pcre_exec()
3152 callout_data data for callouts (see 49 below)
3153
3154 The flag bits are also defined in pcre.h, and are
3155
3156 PCRE_EXTRA_STUDY_DATA
3157 PCRE_EXTRA_MATCH_LIMIT
3158 PCRE_EXTRA_CALLOUT_DATA
3159
3160 The pcre_study() function now returns one of these new pcre_extra blocks, with
3161 the actual study data pointed to by the study_data field, and the
3162 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3163 before. That is, this change is entirely upwards-compatible and requires no
3164 change to existing code.
3165
3166 If you want to pass in additional data to pcre_exec(), you can either place it
3167 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3168 block.
3169
3170 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3171 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3172 times with different match limits, until it finds the minimum value needed for
3173 pcre_exec() to complete. The value is then output. This can be instructive; for
3174 most simple matches the number is quite small, but for pathological cases it
3175 gets very large very quickly.
3176
3177 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3178 returns the size of the data block pointed to by the study_data field in a
3179 pcre_extra block, that is, the value that was passed as the argument to
3180 pcre_malloc() when PCRE was getting memory in which to place the information
3181 created by pcre_study(). The fourth argument should point to a size_t variable.
3182 pcretest has been extended so that this information is shown after a successful
3183 pcre_study() call when information about the compiled regex is being displayed.
3184
3185 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3186 because what follows is always an absolute path. (Later: it turns out that this
3187 is more than cosmetic for MinGW, because it doesn't like empty path
3188 components.)
3189
3190 49. Some changes have been made to the callout feature (see 28 above):
3191
3192 (i) A callout function now has three choices for what it returns:
3193
3194 0 => success, carry on matching
3195 > 0 => failure at this point, but backtrack if possible
3196 < 0 => serious error, return this value from pcre_exec()
3197
3198 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3199 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3200 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3201 use by callout functions. It will never be used by PCRE itself.
3202
3203 (ii) The pcre_extra structure (see 45 above) has a void * field called
3204 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3205 pcre_callout_block structure has a field of the same name. The contents of
3206 the field passed in the pcre_extra structure are passed to the callout
3207 function in the corresponding field in the callout block. This makes it
3208 easier to use the same callout-containing regex from multiple threads. For
3209 testing, the pcretest program has a new data escape
3210
3211 \C*n pass the number n (may be negative) as callout_data
3212
3213 If the callout function in pcretest receives a non-zero value as
3214 callout_data, it returns that value.
3215
3216 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3217 there were some redundant $(CFLAGS) in commands that are now specified as
3218 $(LINK), which already includes $(CFLAGS).
3219
3220 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3221 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3222 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3223 one-byte characters throughout. Note that case-insensitive matching applies
3224 only to characters whose values are less than 256. PCRE doesn't support the
3225 notion of cases for higher-valued characters.
3226
3227 (i) A character class whose characters are all within 0-255 is handled as
3228 a bit map, and the map is inverted for negative classes. Previously, a
3229 character > 255 always failed to match such a class; however it should
3230 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3231
3232 (ii) A negated character class with a single character < 255 is coded as
3233 "not this character" (OP_NOT). This wasn't working properly when the test
3234 character was multibyte, either singly or repeated.
3235
3236 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3237 mode, for example: \x{100}{2,3}.
3238
3239 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3240 singly or repeated) now correctly test multibyte characters. However,
3241 PCRE doesn't recognize any characters with values greater than 255 as
3242 digits, spaces, or word characters. Such characters always match \D, \S,
3243 and \W, and never match \d, \s, or \w.
3244
3245 (v) Classes may now contain characters and character ranges with values
3246 greater than 255. For example: [ab\x{100}-\x{400}].
3247
3248 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3249 PCRE in UTF-8 mode.
3250
3251 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3252 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3253 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3254 value.)
3255
3256 53. The single man page has become too large. I have therefore split it up into
3257 a number of separate man pages. These also give rise to individual HTML pages;
3258 these are now put in a separate directory, and there is an index.html page that
3259 lists them all. Some hyperlinking between the pages has been installed.
3260
3261 54. Added convenience functions for handling named capturing parentheses.
3262
3263 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3264 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3265 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3266 are faulted.
3267
3268 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3269 calling configure. These values are used when compiling the dftables.c program
3270 which is run to generate the source of the default character tables. They
3271 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3272 you will need to set these values.
3273
3274 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3275
3276
3277 Version 3.9 02-Jan-02
3278 ---------------------
3279
3280 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3281
3282 2. If --disable-static was given, the building process failed when trying to
3283 build pcretest and pcregrep. (For some reason it was using libtool to compile
3284 them, which is not right, as they aren't part of the library.)
3285
3286
3287 Version 3.8 18-Dec-01
3288 ---------------------
3289
3290 1. The experimental UTF-8 code was completely screwed up. It was packing the
3291 bytes in the wrong order. How dumb can you get?
3292
3293
3294 Version 3.7 29-Oct-01
3295 ---------------------
3296
3297 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3298 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3299 this didn't happen under Solaris 8, where I normally test things.
3300
3301 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3302 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3303 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3304 this fix an hour or so after the initial 3.7 release.)
3305
3306
3307 Version 3.6 23-Oct-01
3308 ---------------------
3309
3310 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3311 offsets passed as NULL with zero offset count.
3312
3313 2. The config.guess and config.sub files had not been updated when I moved to
3314 the latest autoconf.
3315
3316
3317 Version 3.5 15-Aug-01
3318 ---------------------
3319
3320 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3321 had been forgotten.
3322
3323 2. By using declared but undefined structures, we can avoid using "void"
3324 definitions in pcre.h while keeping the internal definitions of the structures
3325 private.
3326
3327 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3328 user point of view, this means that both static and shared libraries are built
3329 by default, but this can be individually controlled. More of the work of
3330 handling this static/shared cases is now inside libtool instead of PCRE's make
3331 file.
3332
3333 4. The pcretest utility is now installed along with pcregrep because it is
3334 useful for users (to test regexs) and by doing this, it automatically gets
3335 relinked by libtool. The documentation has been turned into a man page, so
3336 there are now .1, .txt, and .html versions in /doc.
3337
3338 5. Upgrades to pcregrep:
3339 (i) Added long-form option names like gnu grep.
3340 (ii) Added --help to list all options with an explanatory phrase.
3341 (iii) Added -r, --recursive to recurse into sub-directories.
3342 (iv) Added -f, --file to read patterns from a file.
3343
3344 6. pcre_exec() was referring to its "code" argument before testing that
3345 argument for NULL (and giving an error if it was NULL).
3346
3347 7. Upgraded Makefile.in to allow for compiling in a different directory from
3348 the source directory.
3349
3350 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3351 options bits, the pointer it was passed was to an int instead of to an unsigned
3352 long int. This mattered only on 64-bit systems.
3353
3354 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3355 generated) instead of pcre.in, which it its source. Also made the same change
3356 in several of the .c files.
3357
3358 10. A new release of gcc defines printf() as a macro, which broke pcretest
3359 because it had an ifdef in the middle of a string argument for printf(). Fixed
3360 by using separate calls to printf().
3361
3362 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3363 script, to force use of CR or LF instead of \n in the source. On non-Unix
3364 systems, the value can be set in config.h.
3365
3366 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3367 absolute limit. Changed the text of the error message to make this clear, and
3368 likewise updated the man page.
3369
3370 13. The limit of 99 on the number of capturing subpatterns has been removed.
3371 The new limit is 65535, which I hope will not be a "real" limit.
3372
3373
3374 Version 3.4 22-Aug-00
3375 ---------------------
3376
3377 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3378
3379 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3380
3381
3382 Version 3.3 01-Aug-00
3383 ---------------------
3384
3385 1. If an octal character was given, but the value was greater than \377, it
3386 was not getting masked to the least significant bits, as documented. This could
3387 lead to crashes in some systems.
3388
3389 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3390 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3391
3392 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3393 These just pass their arguments on to (pcre_free)(), but they are provided
3394 because some uses of PCRE bind it to non-C systems that can call its functions,
3395 but cannot call free() or pcre_free() directly.
3396
3397 4. Add "make test" as a synonym for "make check". Corrected some comments in
3398 the Makefile.
3399
3400 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3401 Makefile.
3402
3403 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3404 command called pgrep for grepping around the active processes.
3405
3406 7. Added the beginnings of support for UTF-8 character strings.
3407
3408 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3409 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3410 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3411 out for the ar command.)
3412
3413
3414 Version 3.2 12-May-00
3415 ---------------------
3416
3417 This is purely a bug fixing release.
3418
3419 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3420 of ZA. This was just one example of several cases that could provoke this bug,
3421 which was introduced by change 9 of version 2.00. The code for breaking
3422 infinite loops after an iteration that matches an empty string was't working
3423 correctly.
3424
3425 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3426 when matched against abbab (for example). After matching an empty string, it
3427 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3428 caused it to match further down the string than it should.
3429
3430 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3431 was there because it doesn't seem to be needed, and it causes trouble on some
3432 systems, as it is not a Standard C header. It has been removed.
3433
3434 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3435 were reported on the Macintosh. The changes were from
3436
3437 while ((c = *(++ptr)) != 0 && c != '\n');
3438 to
3439 while ((c = *(++ptr)) != 0 && c != '\n') ;
3440
3441 Totally extraordinary, but if that's what it takes...
3442
3443 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3444 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3445 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3446 assumes the way PCRE uses memmove() (always moving upwards).
3447
3448 6. PCRE is being used in one environment where strchr() is not available. There
3449 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3450 faster code anyway.
3451
3452
3453 Version 3.1 09-Feb-00
3454 ---------------------
3455
3456 The only change in this release is the fixing of some bugs in Makefile.in for
3457 the "install" target:
3458
3459 (1) It was failing to install pcreposix.h.
3460
3461 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3462
3463
3464 Version 3.0 01-Feb-00
3465 ---------------------
3466
3467 1. Add support for the /+ modifier to perltest (to output $` like it does in
3468 pcretest).
3469
3470 2. Add support for the /g modifier to perltest.
3471
3472 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3473 matches null strings.
3474
3475 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3476 pattern. Perl treats empty patterns specially - it reuses the most recent
3477 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3478 effect.
3479
3480 5. The POSIX interface was broken in that it was just handing over the POSIX
3481 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3482 required a bigger vector, with some working space on the end. This means that
3483 the POSIX wrapper now has to get and free some memory, and copy the results.
3484
3485 6. Added some simple autoconf support, placing the test data and the
3486 documentation in separate directories, re-organizing some of the
3487 information files, and making it build pcre-config (a GNU standard). Also added
3488 libtool support for building PCRE as a shared library, which is now the
3489 default.
3490
3491 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3492 09 are not valid octal constants. Single digits will be used for minor values
3493 less than 10.
3494
3495 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3496 existing programs that set these in the POSIX interface can use PCRE without
3497 modification.
3498
3499 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3500 return all that pcre_info() returns, plus additional data. The pcre_info()
3501 function is retained for compatibility, but is considered to be obsolete.
3502
3503 10. Added experimental recursion feature (?R) to handle one common case that
3504 Perl 5.6 will be able to do with (?p{...}).
3505
3506 11. Added support for POSIX character classes like [:alpha:], which Perl is
3507 adopting.
3508
3509
3510 Version 2.08 31-Aug-99
3511 ----------------------
3512
3513 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3514 trying to match at the startoffset position, but instead was moving forward to
3515 the next newline as if a previous match had failed.
3516
3517 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3518 and could get into a loop if a null string was matched other than at the start
3519 of the subject.
3520
3521 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3522 be distinguished at compile time, and for completeness also added PCRE_DATE.
3523
3524 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3525 in GnuWin32 environments.
3526
3527
3528 Version 2.07 29-Jul-99
3529 ----------------------
3530
3531 1. The documentation is now supplied in plain text form and HTML as well as in
3532 the form of man page sources.
3533
3534 2. C++ compilers don't like assigning (void *) values to other pointer types.
3535 In particular this affects malloc(). Although there is no problem in Standard
3536 C, I've put in casts to keep C++ compilers happy.
3537
3538 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3539 should be (const char *).
3540
3541 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3542 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3543 However, I haven't made this a standard facility. The documentation doesn't
3544 mention it, and the Makefile doesn't support it.
3545
3546 5. The Makefile now contains an "install" target, with editable destinations at
3547 the top of the file. The pcretest program is not installed.
3548
3549 6. pgrep -V now gives the PCRE version number and date.
3550
3551 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3552 causing the entire string to be ignored, instead of just the last character.
3553
3554 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3555 non-matching string, it can take a very, very long time, even for strings of
3556 quite modest length, because of the nested recursion. PCRE now does better in
3557 some of these cases. It does this by remembering the last required literal
3558 character in the pattern, and pre-searching the subject to ensure it is present
3559 before running the real match. In other words, it applies a heuristic to detect
3560 some types of certain failure quickly, and in the above example, if presented
3561 with a string that has no trailing " it gives "no match" very quickly.
3562
3563 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3564 other alternatives are tried instead.
3565
3566
3567 Version 2.06 09-Jun-99
3568 ----------------------
3569
3570 1. Change pcretest's output for amount of store used to show just the code
3571 space, because the remainder (the data block) varies in size between 32-bit and
3572 64-bit systems.
3573
3574 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3575 start matching at. This allows lookbehinds to work when searching for multiple
3576 occurrences in a string.
3577
3578 3. Added additional options to pcretest for testing multiple occurrences:
3579
3580 /+ outputs the rest of the string that follows a match
3581 /g loops for multiple occurrences, using the new startoffset argument
3582 /G loops for multiple occurrences by passing an incremented pointer
3583
3584 4. PCRE wasn't doing the "first character" optimization for patterns starting
3585 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3586 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3587 the letter 'x'. On long subject strings, this gives a significant speed-up.
3588
3589
3590 Version 2.05 21-Apr-99
3591 ----------------------
3592
3593 1. Changed the type of magic_number from int to long int so that it works
3594 properly on 16-bit systems.
3595
3596 2. Fixed a bug which caused patterns starting with .* not to work correctly
3597 when the subject string contained newline characters. PCRE was assuming
3598 anchoring for such patterns in all cases, which is not correct because .* will
3599 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3600 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3601 must be retried after every newline in the subject.
3602
3603
3604 Version 2.04 18-Feb-99
3605 ----------------------
3606
3607 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3608 computation of the store needed to hold the pattern was incorrect (too large).
3609 If such patterns were nested a few deep, this could multiply and become a real
3610 problem.
3611
3612 2. Added /M option to pcretest to show the memory requirement of a specific
3613 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3614
3615 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3616 compiled in such a way that the backtracking after subsequent failure was
3617 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3618 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3619
3620
3621 Version 2.03 02-Feb-99
3622 ----------------------
3623
3624 1. Fixed typo and small mistake in man page.
3625
3626 2. Added 4th condition (GPL supersedes if conflict) and created separate
3627 LICENCE file containing the conditions.
3628
3629 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3630 Perl, that is the internal \ allows the delimiter to be included in the
3631 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3632 the final delimiter, add \ to the end of the pattern (to test the error).
3633
3634 4. Added the convenience functions for extracting substrings after a successful
3635 match. Updated pcretest to make it able to test these functions.
3636
3637
3638 Version 2.02 14-Jan-99
3639 ----------------------
3640
3641 1. Initialized the working variables associated with each extraction so that
3642 their saving and restoring doesn't refer to uninitialized store.
3643
3644 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3645 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3646 fix the problem.
3647
3648 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3649 calls, and wasn't printing the correct value for compiling calls. Increased the
3650 default value of LOOPREPEAT, and the number of significant figures in the
3651 times.
3652
3653 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3654
3655 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3656 a building problem on Windows NT with a FAT file system.
3657
3658
3659 Version 2.01 21-Oct-98
3660 ----------------------
3661
3662 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3663 to character tables built by pcre_maketables() in the current locale. If NULL
3664 is passed, the default tables are used.
3665
3666
3667 Version 2.00 24-Sep-98
3668 ----------------------
3669
3670 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3671 it any more.
3672
3673 2. Allow quantification of (?>) groups, and make it work correctly.
3674
3675 3. The first character computation wasn't working for (?>) groups.
3676
3677 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3678 end of the subject) and add 5.005's \z, which really does match only at the
3679 very end of the subject.
3680
3681 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3682
3683 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3684 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3685 localized options. All options to pcre_study() were also removed.
3686
3687 7. Add other new features from 5.005:
3688
3689 $(?<= positive lookbehind
3690 $(?<! negative lookbehind
3691 (?imsx-imsx) added the unsetting capability
3692 such a setting is global if at outer level; local otherwise
3693 (?imsx-imsx:) non-capturing groups with option setting
3694 (?(cond)re|re) conditional pattern matching
3695
3696 A backreference to itself in a repeated group matches the previous
3697 captured string.
3698
3699 8. General tidying up of studying (both automatic and via "study")
3700 consequential on the addition of new assertions.
3701
3702 9. As in 5.005, unlimited repeated groups that could match an empty substring
3703 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3704 runtime if any iteration does actually match an empty substring.
3705
3706 10. Include the RunTest script in the distribution.
3707
3708 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3709 discrepancies, some of which were old and were also with respect to 5.004. They
3710 have now been fixed.
3711
3712
3713 Version 1.09 28-Apr-98
3714 ----------------------
3715
3716 1. A negated single character class followed by a quantifier with a minimum
3717 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3718 program crashes, or just wrong answers. This did not apply to negated classes
3719 containing more than one character, or to minima other than one.
3720
3721
3722 Version 1.08 27-Mar-98
3723 ----------------------
3724
3725 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3726
3727 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3728 latter must appear before anything that relies on it in the pattern.
3729
3730
3731 Version 1.07 16-Feb-98
3732 ----------------------
3733
3734 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3735 repeat of a potentially empty string).
3736
3737
3738 Version 1.06 23-Jan-98
3739 ----------------------
3740
3741 1. Added Markus Oberhumer's little patches for C++.
3742
3743 2. Literal strings longer than 255 characters were broken.
3744
3745
3746 Version 1.05 23-Dec-97
3747 ----------------------
3748
3749 1. Negated character classes containing more than one character were failing if
3750 PCRE_CASELESS was set at run time.
3751
3752
3753 Version 1.04 19-Dec-97
3754 ----------------------
3755
3756 1. Corrected the man page, where some "const" qualifiers had been omitted.
3757
3758 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3759 input syntax.
3760
3761 3. Fixed memory leak which occurred when a regex with back references was
3762 matched with an offsets vector that wasn't big enough. The temporary memory
3763 that is used in this case wasn't being freed if the match failed.
3764
3765 4. Tidied pcretest to ensure it frees memory that it gets.
3766
3767 5. Temporary memory was being obtained in the case where the passed offsets
3768 vector was exactly big enough.
3769
3770 6. Corrected definition of offsetof() from change 5 below.
3771
3772 7. I had screwed up change 6 below and broken the rules for the use of
3773 setjmp(). Now fixed.
3774
3775
3776 Version 1.03 18-Dec-97
3777 ----------------------
3778
3779 1. A erroneous regex with a missing opening parenthesis was correctly
3780 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3781 on some systems.
3782
3783 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3784 it was reported that one broken compiler failed on the former because "code" is
3785 also an independent variable.
3786
3787 3. The erroneous regex a[]b caused an array overrun reference.
3788
3789 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3790 fail on data ending with that character. (It was going on too far, and checking
3791 the next character, typically a binary zero.) This was specific to the
3792 optimized code for single-character negative classes.
3793
3794 5. Added a contributed patch from the TIN world which does the following:
3795
3796 + Add an undef for memmove, in case the the system defines a macro for it.
3797
3798 + Add a definition of offsetof(), in case there isn't one. (I don't know
3799 the reason behind this - offsetof() is part of the ANSI standard - but
3800 it does no harm).
3801
3802 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3803 most of the places where whitespace preceded '#'. I have given up and
3804 allowed the remaining 2 cases to be at the margin.
3805
3806 + Rename some variables in pcre to eliminate shadowing. This seems very
3807 pedantic, but does no harm, of course.
3808
3809 6. Moved the call to setjmp() into its own function, to get rid of warnings
3810 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3811
3812 7. Constructs such as \d{8,} were compiling into the equivalent of
3813 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3814 outcome, but in this particular case used more store than had been allocated,
3815 which caused the bug to be discovered because it threw up an internal error.
3816
3817 8. The debugging code in both pcre and pcretest for outputting the compiled
3818 form of a regex was going wrong in the case of back references followed by
3819 curly-bracketed repeats.
3820
3821
3822 Version 1.02 12-Dec-97
3823 ----------------------
3824
3825 1. Typos in pcre.3 and comments in the source fixed.
3826
3827 2. Applied a contributed patch to get rid of places where it used to remove
3828 'const' from variables, and fixed some signed/unsigned and uninitialized
3829 variable warnings.
3830
3831 3. Added the "runtest" target to Makefile.
3832
3833 4. Set default compiler flag to -O2 rather than just -O.
3834
3835
3836 Version 1.01 19-Nov-97
3837 ----------------------
3838
3839 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3840 like /([ab]*)*/, that is, for classes with more than one character in them.
3841
3842 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3843 as /((?>a*))*/ (a PCRE_EXTRA facility).
3844
3845
3846 Version 1.00 18-Nov-97
3847 ----------------------
3848
3849 1. Added compile-time macros to support systems such as SunOS4 which don't have
3850 memmove() or strerror() but have other things that can be used instead.
3851
3852 2. Arranged that "make clean" removes the executables.
3853
3854
3855 Version 0.99 27-Oct-97
3856 ----------------------
3857
3858 1. Fixed bug in code for optimizing classes with only one character. It was
3859 initializing a 32-byte map regardless, which could cause it to run off the end
3860 of the memory it had got.
3861
3862 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3863
3864
3865 Version 0.98 22-Oct-97
3866 ----------------------
3867
3868 1. Fixed bug in code for handling temporary memory usage when there are more
3869 back references than supplied space in the ovector. This could cause segfaults.
3870
3871
3872 Version 0.97 21-Oct-97
3873 ----------------------
3874
3875 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3876
3877 2. Optimized negated single characters not to use a bit map.
3878
3879 3. Brought error texts together as macro definitions; clarified some of them;
3880 fixed one that was wrong - it said "range out of order" when it meant "invalid
3881 escape sequence".
3882
3883 4. Changed some char * arguments to const char *.
3884
3885 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3886
3887 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3888 pcretest.
3889
3890
3891 Version 0.96 16-Oct-97
3892 ----------------------
3893
3894 1. Added a simple "pgrep" utility to the distribution.
3895
3896 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3897 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3898 where "ddd" means "one or more decimal digits".
3899
3900 3. Fixed serious bug. If a pattern had a back reference, but the call to
3901 pcre_exec() didn't supply a large enough ovector to record the related
3902 identifying subpattern, the match always failed. PCRE now remembers the number
3903 of the largest back reference, and gets some temporary memory in which to save
3904 the offsets during matching if necessary, in order to ensure that
3905 backreferences always work.
3906
3907 4. Increased the compatibility with Perl in a number of ways:
3908
3909 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3910 to request this handling. The option can be set at compile or exec time.
3911
3912 (b) $ matches before a terminating newline by default; an option
3913 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3914 mode). The option can be set at compile or exec time.
3915
3916 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3917 the same as Perl's. If the decimal number it represents is less than 10
3918 or there aren't that many previous left capturing parentheses, an octal
3919 escape is read. Inside a character class, it's always an octal escape,
3920 even if it is a single digit.
3921
3922 (d) An escaped but undefined alphabetic character is taken as a literal,
3923 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3924 escapes.
3925
3926 (e) {0} is now permitted. (The previous item is removed from the compiled
3927 pattern).
3928
3929 5. Changed all the names of code files so that the basic parts are no longer
3930 than 10 characters, and abolished the teeny "globals.c" file.
3931
3932 6. Changed the handling of character classes; they are now done with a 32-byte
3933 bit map always.
3934
3935 7. Added the -d and /D options to pcretest to make it possible to look at the
3936 internals of compilation without having to recompile pcre.
3937
3938
3939 Version 0.95 23-Sep-97
3940 ----------------------
3941
3942 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3943 \x20 at the start of a run of normal characters. These were being treated as
3944 real characters, instead of the source characters being re-checked.
3945
3946
3947 Version 0.94 18-Sep-97
3948 ----------------------
3949
3950 1. The functions are now thread-safe, with the caveat that the global variables
3951 containing pointers to malloc() and free() or alternative functions are the
3952 same for all threads.
3953
3954 2. Get pcre_study() to generate a bitmap of initial characters for non-
3955 anchored patterns when this is possible, and use it if passed to pcre_exec().
3956
3957
3958 Version 0.93 15-Sep-97
3959 ----------------------
3960
3961 1. /(b)|(:+)/ was computing an incorrect first character.
3962
3963 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3964 but not actually doing anything yet.
3965
3966 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3967 as Perl does (e.g. [-az] or [az-]).
3968
3969 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3970 all possible positions.
3971
3972 5. Split up into different modules to avoid including unneeded functions in a
3973 compiled binary. However, compile and exec are still in one module. The "study"
3974 function is split off.
3975
3976 6. The character tables are now in a separate module whose source is generated
3977 by an auxiliary program - but can then be edited by hand if required. There are
3978 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3979 toupper() in the code.
3980
3981 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3982 make them global. Abolish the function for setting them, as the caller can now
3983 set them directly.
3984
3985
3986 Version 0.92 11-Sep-97
3987 ----------------------
3988
3989 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3990 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3991
3992 2. Caseless matching was not working in character classes if the characters in
3993 the pattern were in upper case.
3994
3995 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3996
3997 4. Make PCRE_ANCHORED public and accept as a compile option.
3998
3999 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
4000 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
4001 pass them.
4002
4003 6. Give an error if bad option bits passed at compile or run time.
4004
4005 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
4006 pcretest to cause it to pass that flag.
4007
4008 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
4009 options, and the first character, if set.
4010
4011 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
4012
4013
4014 Version 0.91 10-Sep-97
4015 ----------------------
4016
4017 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
4018 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
4019
4020 2. PCRE was looping on encountering an indefinitely repeated back reference to
4021 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
4022 Perl does - treats the match as successful.
4023
4024 ****

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